Programmed_Statistics_(Question-Answers), programmed statistics, statistics, b l agrawal, bl agrawal, buntu, niteshkewat, nitesh kewat, kewat

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

9,8K Aufrufe

Programmed_Statistics_(Question-Answers), programmed statistics, statistics, b l agrawal, bl agrawal, buntu, niteshkewat, nitesh kewat, kewat

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

- Business Statistics by [S P Gupta]
- MCQs Basic Statistics 1
- 40 solved MCQs of STA630- Research Methods
- Programmed Statistics by b l Agarwal PDF Free
- Mcq Testing of Hypothesis
- Statistics and Data With R
- Fundamental of applied statistics by SC Gupta
- Methods of Time Series
- Business Statistics
- Sampling Theory and Methods
- MCQs on Correlation and Regression Analysis
- Statistics Mcq'S I.Com Part 2
- Statistic Exam (MCQ)
- Operations Research - JK Sharma
- Fundamental of Stat by S C GUPTA
- Index Numbers
- Differential Equations (Golden Maths Series) - N. P. Bali
- MCQ stats
- William G. Cochran Sampling Techniques
- An Introduction to Multivariate Statistical Analysis.pdf

Sie sind auf Seite 1von 551

-."S. •

, : I."~

• 0-

j-' .~..-_..

:.~.

~ .. '

C<Jpyrigbt 0 2003 New Age International (P).W., .Publiehera Second Edition : 2003

Reprint: 2{)05

NEW AGE INTEBNATIONAL (PI LIMITED, PUBLISBERS 4835f24, Anaari Road, Daryagarl,j,

New Delhi· 110 002

Visit WI at : www.newagepublishenl.tom

Offlcuot:

BanraJ"re, ChaMai, Cochin, Guwabatl, Hyderabad. JaJandbar.

Kolbta, Luclmuw, Mwnhai andRanchi

This book or any part- thereof may not be "l~uc:ed in any form. without the ...-itten ""nniallion of the publisher.

Thl. book """"01. be IIOW ou!.oide the COUDIl'y to which it is coMi&ne<i by the plitilloher witho"t th<o prior pormillOdon of tha puhliaur.

n..275.oo

ISBN ,81·224-1415&.3

3'" 5 6 7 8 D 10

Publis_hed by N ...... Age International (P) Ltd., 4885124.AI!eatt ROaiI"J)aryaa!I!lj •. N_ Delhi-110 002 and printed in India at T.u Preu, New OeJhi.110 064 typeaetted at p..,..tek, o..lhi-110 092,

U rheberrec htl lch ge5c h utztes Iv! ateri"

Contents

P'I':/uce 10 rh. Second EJI,um Preface 10 ,I" Fin, £Jillml

Chapb'r 1

Stntistl""

Shari =r type question, On:

Defi nitl Of! S of '*' tjs'i cs I Srm em en's nbOlu STot i ::it; os I StOli:stical prospect;"" 2

o ivi ,j on and 1i m ifill' pn!jj Func:ti OM of stall sHes i Mcrb ods of collecl'ion Q f d;na 1)<pe" of .latl'tical dllln 3 S'ali.,ic.1 ,."rver 3

Que,li"n nnlp> nod schod"le 3 Kind. ,,($1.(1(1."001 In_"go.Uon 4 ~

~

ApprO?d"",u on of .. l"es 5 PrltIlorynnd $et<lndary dolO 5 Editing of d.ln 5

Fill in rbtl; bl:mks 6

Mull; piech";",, que,hons 7 An,swrrs In

So gg,e5lcd ncodl n S II

ClasslfiClltion. Thbula.tion DIlII FrequencY Distribution

Chapter 1

12

,s.h on es .. y type guestl ".as on:

ClflsslOeOlian 12 Groupl" sof dall> 13 Group; fig '''fllI 13 ThbuJil'inD 14 Frog,,:" or /5

U rheberrec htl ich gesc h utztes M a\eria

Chapter 3

Cu on" I 'Ii lie f'."'Iuen oy J 5 Types of "'ri~, 15

Slem and 1",,[ d i.yl .• y of cl~ta 15 Fill in Ihe blanks 15

Mul!iple "hoi"" '1"o>1ion, 17 Answcrs 20

S" gge>led rend jng 2Q

DiagrumaUc lind GruphiCIII Represenlation

21

Chap!~r4

Short essay type q klCS" l)n.'ii ("I n:

Ad va"'''R''' of d'"srams 2 J

Bar diagram 21

Su[>.div'dod b." d iapJ'3m 21

M u ItipJeb<tr J jay"", 22

Dey; 'Ii on bar di,srom 22 D"lKli""'lio"~1 bur di'Brrut1 22 Pnired bn, d io grom 21

Slid iog bnr d j. gram 21

lI'o",,~ b .. d;"gnm 23

Li ne d iugram 24

PTc-_cbtJrl 24

Hi"",gmm 25

Fre<] .oncy polygon 25 Freq lJ ency c utv~ 15 SConil<Jgo,ilhmic gnph 26 Ogi oc cu rve 26

1_0 rcn? eiJ rye 27

Pic 109", 'no 27

0,1'1.1 m n ch.an wi In ci fCilbr base 28 SlOP b.tr chon 28

Oocd.pp iOIl bar < rum 28

FiU in Ibr' bIimkr'. 29

M" !tip Ie ch oice gu ""I ion. 30 Answers 15

Su "CSled trod ins 35

MeasuJ;'l'll of Ceolr.l\ Timdeng

36

Com:cpl and deli"; lioll3 36 Proponies of cen Iml tendency .16 CI" .. ificalion of avera!!"" 36 Ftmclj no Ii 16

Li mjtnJ i.Qn:'O JZ

Arilhmelic· mert" 37

Weighlod m""" .Nl

Movi fIg avcr4lg,e 38

Pragre$1:ive ~'LIeragc 38

U rheberrec htl ich gesc h utztes Materia

xl

Com po, i.le ",v.rage 38 Pllul.~d meal) 19 GeQO) e! tic' IDeM 12

H armca jc mf· .. 1D 40

Relmion belw.en A.M., a.M .• nd H.M. 41 Qu odmlic me.n 4/

M(~d"'u 42

.MllWl......n

Qu.I'l;[o.' 43

nrejl'~:;; 4~

Pt-rcen1il¢s 4·1

Fill in l'bc~ bbnks 44

Mulli pie choice. quest; on, 47 An$wcrr;, 58

Sugge<led reading 59

Measures 0 r Dispersion, Skewness and Ku rtesls

Clmp[e~5

60

Concept and ,tatement' 00

Reg" i, ites of " good ",e:!Sure 0/ llir.i......1ll

Rang" 61

Cocffici cnt of ""SC 61 1 n lerquanU [: range 61 Perce" tile '<In ge 6.1

Qu arli Ie deviation 61 MC)]" devj Olio n 67 V!!!ri3no~ 62

SI:mdard deyi~lij1" 6] Men" [)bsnhlte,_dj~ ... i;;r'inll 6j Coe OJ ci eut .,r Y"O ill'io [! 64 PoQled' ygrin nee 64 Moments 65"

Pmh_ilblt; !'·rmr 65

Empirical ,.Ifition I>o!ween SD" Q.D.,nd M.D. 65 rl c I j'! I ions brtWN"n n"" 'm I j'! nil til VI mo n]""nts 6 5 Skewness 66

Sheppard '., correct;" ns 68 Fill in !he bi'lDlqit 62

Multiple choke question, 72 Answers 81

Sugge.'le<i read; ng 8l

Etemcnlory Probability

Clmptcr6

83

Short essay tyn~' q ue,~;;ti on 5 t1 n:

C<lnCCpl 83

U rheberrcc htl ich gesc h utztes M a\e,ia

Chapl .... '

~

comp Ie rnM!ary 83 simple ,84

~

oompound 84 mutually '" du.ive 84 rrim.ry 84

deriYr.d ,84

Un .sinn and intenccljoo of eve pIS 84 Imp=ible """nl 85

5)j,b£J 11 'itt_Wi _85

Pairwi.., i.n dependcn I evenlS 85 Dcfin'ti.on, of pmbob ilily;

dns.-si¢.,)! 85 ~!a;risrk';I11_ 86

Add,t; '" I ow of proll.bi I ity lJ6 Mllilip I j ".t've low of probttbilily lJ6 Condit; on.1 prob.bi I ily 86

Baye.' probnbililY 87

D' Alernbcrt', parad".. 87 Fi II in !he blanh 1!8

Mul!iple ehoi ee question. 90 Answ ers 10/

Sugges,"" ... ading 105

Random Variable, Mlllhematical Expedaito.n

ana Pro babili Iy Dlsln butions

CONTENTS

106

Short .<oay lypeguo'l;ons on, RI1 uti om vnrj;;ble 1M djS!;rele 107 continuous 107

Dj:;1 ributj on fiJ 11 chon 107

Ci..utJ u lalive d isn-ibuliQn f" nclio n 107 I'rob.bility on .. , f"neUen 107

Step fun eli on /07

Pm babi lity den.<:ily functio n 107 Prob"" illty dellOi IV co rw 107 M"thcm"u c.l e:<pec~'(iO n 108 Cauchy-Schwartz inequality /09 Jensen's i neq ""I ity 109

Gurland's inequality on expectation 109 Moment ge""'llting function 109 Characterisuc function 109

Probability I1"nerating function /10

PI scm.; di srdhut i nnt_"

uoirpnn _I/Q

U rheberrcc htl ich gesc h utztes M a\eria

CONTENTS

ChapterS

Bernoulli 110 Binemia! I/O Poisson /12 Neg."live 113 Gm-'tUe1rjc 114 Poly,', 115 HypergoomcIrio 115 Multinomlrd 11'6

Continuous distributions;

uniform or rectangular 117 exponentlal ns

Lap] "00 Or dou b 10 "" pGnonti(ll 119 gOUDa' 1t2

lognormal 123

Cauchy 123

beta di sui berions of I and U kind 121 gamma 125

logistic 126

Pareto 126

WcibuU 126

Circular 126

Pearson ion 126

Sampling distributions:

Chi-square 127 St.ud"nt's-I 128

Pearson', Chi-square /2S Chi 128

Fisher's-z /3/

S netl CCQ{Sr F '-' , Fill in dm blqnks In

Mu Iii pi" "ho i"e 9 "esli ons J 38 AnllWC!); 160

SU!!8"'lcd reading 162

BlVlniat~ Random VariablQ and Distributions

xIII

163

S hert cs'"y ty pe gu .slions 0 n:

B.ivvriate, randgm yaric.ble" JOl ,1jsrttte 161

cnnrin"Q", IrB

JQinl d isrributi pn fy nc lio n 164-

lei n I pro bnhiJity m",s function 164

J <>j" I pro 1mb, lily d. n$,ly funclion 164 Mar;ginal probability functions /64 Conditional yariable 165

Conditional probability density funclion 165 I ndeP<'n d once or rrutdnm ""riab I.. 165

U rheberrec htl ich gesc h utztes Materia

xlv

CONTEN'lS

Conditional expectation 16J OU! r! j I 100uil YMi Of! ct: 166 Condi I ion.,! (oymj Flne<: 167

Joint mom!!!nl ge_J'Ier~:u:illg, [uncl inn 167 -!oi n1 fa IN 100 nt,·MS 167

lui a! central m pmems 167

n jynri{J1 ~ n Qrm!.'ll d iSirihu I'io II J,67 Fill in. the bonks 172

Mubiplc "hoi". questions 174 Angwers IS?

SuggCSled reoding 183

Chaptcr9

184

ShOI1 ,,=t lyP" 9" .... Ii On' "n ,

Same ling vs. Cornu lele enu",enuion 184 POP" Intia n 184

U_IDJ~/l1J jnfinjle Uij

n:nl.......Lti 1111'0111"1"01 185

Sampling f'~n,e 185 Rondom """,piing 185

S i",p [e,ando m .ompl i og 186

Princip los of =plj ng method.> 186 Samp I i og nn d non· • .,m pi i ng errors 186 Po 'PO'; ye ""In pli n ~ 186

Parnrm:ler un

EM i rna to [ :' uti cS1j m ah~ 187 St;1!islir 187

Simp Ie. random sam pi i nt, with al1d whhout rep! ace mc:nt 187 $nmp ie ,i'e 11i7

Sa'" pill m""" IlJlI,

Sam pl~ vn,inn"" ,188

Slnnd'~[ji' emlf 189

Fi n ite pop"inti on correction /89 Sampling fraction 189

Suatified ,,,ndomsamplin& 190 Ex igencies:

cri I L'"rjn for $Irtlt i (jcflljon '9 , Dllmbr~r of Slrj\! it 191

£CliVI'S .,f '!t\~[!Jmr;""on _l'2_l ""nlpl~ 81',.<' 191

s.nmpling proc ... ilure /9'1

illIili:al.lnn:

"'ILJ~I 191 propo"ion~1 /9/

U rheberrec htl lch gcsc h utztes M ate,ia

urlimum 191 N~·ymll" 191

F'unmlltl1: for mc!) nand YO 0;,"" 192 Two way ""Iifocalion 193

Cun! m Bed sew'! no_ L93

Sy~i~m~Hic ,"i1mplillg 193

lin=_____1_jM

circular 194

nlfml1l!jle fur menn and yarijJuce 194 Dlmhle "'m~ling 195

Clwucr 0, ",oa ... m ph os 195 MulliS!.! ... ,mpli"! 196

TWo SI,a'ge ~runp ling 196

lnve". sam pi i n £ 196

Slunpl ""~ w.lth prob.b II I 1'1' propo'llo",,1 to.1 '0 198 Non~rcsponsc 198

Fill in Tbe blrLl1k:i 1M

Mull;plc choice <,".stions 202

An5W'-'rs 212

Sug,s.,,"d re.:ld in g 2 J J

Chapter 10

Thoory .or Estimotlob

214

Shon e, ... y type questions on: pgrn mCler ? 14

mfim[\!e 214

Est! m tUtU ? 14

Pro""" ios of .,1 imotol'S 215 cons; s'.ney 215 unbi;tsedness 2/.5

me. n. squ.red err", 21 S

mea" squored consistency 216 CfllJnmer·R.oo in.quality 216

B~'I osymplolicolly norm;lle,limOlor 217 Ellie;"" oy 217'

Mjnir01iITi vannore llnhi,nosM rSlttn;'Uor 217 Sullie; eney :1 J7

F"her-Neymfl" fncloriZ<1lion theorem 218 Complete""", 218

Bao .. B!3Ckw.·U tbctm:m 218

Illvarip nee prOf"'"ny of e5li m<:)Jo no 2 J 9 Adn, I "'ill ilily of an , .. ti",OIO' 219

MI'[hods of estjmatjov:

ma:t"r,nIJID likelihood wimj'lhnn 219 least square estimal;on 219

ID!:\hOiJ of m pmI'm ts _2'9

8 ... .,1 lin ear !!Dh i ,,'Sed est1 m.aI or 2M

,'i(V

U rheberrec htl lch ge5c h utztes M aleri"

xvi

Chapter Il

Minimum "ori.n"" gll"drnl;cesl1mol0r 224 G'nus5- Markov theorem 224

M. <hod of M; n; mum Cui"quares 225 B"y~ .. · ",lim'IOts 221

M:Lnjmn' P5ljm$1_1gr; 227

Pitman ~i'ii estimator 228

1utftr'f'Dl ('$'111),,110" 229

Co n fide net~ c'oefflc lent 229 Cn nfiden£i'_ l1mi IS_ !'or·

~ Yllrj (,"ce, Z 10

diffure nre to two mertn s 2 in [ali 0 Of I uro varin ncr" 2' Q

Shoves' rm,fidcncr intern! 21 , Confiden 00 ,cSioo 23 I

nne it' [roo to n fj di·m:g, j ntctVr'd 211 Eill in 'he blanks 2 n

Mu I!ip 10' clio;"" g",,""O n. 236 Answers 247

Suggested reading 248

TestingParametrk .Hypotheses

Short es.'"y type q uestinn on:

Hypo.ncsi. 249

null ~Ih'm"ti"" 250 simple 250 composite 250

Type. of errors:

'f!'" I error' 250 Iypc II error 250 Power of a lest 250 Level of .igllif'c;lncc 250 Size of "te" 25 I

Best critical region 251 One oM two tailed lest 251

Ncnmsdem iud lind randomized Ie" 2J I Degrees of freedom 251

Critical funelion 251

Optimum 1':$1 252

Mo.t powerful te.t 252

N"ym.;!I\-Pearson lemma 252

Minimu teo;l 254

Unbiased "'~t 254

Uniformly moot powerful unbiued test 255 Admissible lest 255

I j\:e'libood (Min ''''''1 25'6

U rheberrcc htl ich gesc h utztes Materia

CO>lTENTS

Chapter 12

S tudem's z-iesi 257

Cochran's approximate r-rest 258 B,~hn"u~:;-Ftsb¢[ [cst 258

Fa imd ,. te::st .260

Z-Ie'l for properties

Chi->guOJe le'l for" variance

TCSI (I)r stllildard_ deyi .. t] on. .258

Cit i· >guo", te>1 in mulli nomi"! dimi bUlLa" 264 Con li"gency IIIble 265

Incomp let. con I i "geney IIIhle 265 Stru 'ill Ira' and modem l.croS 265 E, pInnal<> ry vonoble 266

Te't of independ,,"ce of "Iribule, 266 Yo, es' correct inn 266

Fisher's exact test 268

Coefficienl of contingency 268

B anICle,' test 269

1"-1"'1. 269

Te'l of equulity of two population variances 269 test of equality of several population mean, 269 Analysi, of variance 27Q

Bayes' tesr 270

Sequential probability ratio le'l 270 Fill in Ihe blank. 271

Multiple c[,o;eo que'lion, 277 Answers 289

Suggesled ",ad; ng 290

No"parametrlc Statistical Methods Short essay type qu .. suons on:

General discussion 292 A.ymplolic relative efficiency 293 Power efficiency 293

Tied observations 293

Ordered ;;;t.i.lis! ic

Koltn"Bcrov.Smimov (o,\Sl for one ,ample 295 (mJinanr sign lest 295

Wilcoxon'. ~.;gn"d rank lest 296

Kohnog,onrv·Sm.rno v ' two sample lest 298 Wi lc 0"""'" lost for matched pairs 298 Median teat 298

Wald.WblfowilZ run test 301

Man n- Wh ilney U·teO! 301

Mcnemar', test 303

Mood· s test 304

xvII

292

U rhcberrec htl lch gcsc h utztes M aleri a

""III

Mose,' test 30' Cochran's Q,Ie.<1 306

Krusknl- W~II is one way ~""I y.i, 307 Friedman's IW<I woy .nalysi. or vnriance J09 Mortiple comparisons in Priedmuns' lest 309 J one k hcere-Terpstrn lesl 3/0

Page's test 311

Spearm on', ran k corre lotion j 12 ,",,"d.Ws rank correlation 3/2 Coefficient of concordance 314

Nonp."m,,!ric approach in regression analysis 316 Urown and Mocxfs lesl 1/7

Mood'S If'S! il7

'Ow i Is" tf"s1 118

Cnnfjd!~ na~ _n "T!!r ".t! _ ~_l~

Confidence hand 319

Fill in !be blanks 319

Mu Itil'10 ch oic. questi ons 323 Answers 335

S'ugge'tc-d ..... ding 336

Chapter 13

Regression. and Corrcla lion Melbods

338

Shonel..,.y Iy,pe que"tic os 00:

Oi' net.' I di Sj'IISsfO n 1 18 Rc g "' .. ion model. 338 Scatter d iagrnm 339

E.'ti m ,ti on (I r po" mel"", of , regreos.ion Ii nc 340 Properties of regr ... ion cne mel.Qt' 340

Two rc:gre,ssjon lines:

poinl of i nlCrsection 341

<:In :gl~' bel \veen ~wo regress)",," Ii nes 34 J O><ling 342

Te.,1 of~ignific"nce of regression coefficient 34,2 Te.l of ,oi go ifi.""e. of tile i" =pl 342

Rcgre .. ion funotion 343

c~nnden"" limi!' "I'.~ and (l 343

To",. of lineo.ril.y of regression 344

Weighled rog",,,,ion 345

Curvi lin oar' regressi"n 346

Filling. of orthogonal polynomiol 347

earwlglinn HZ Crn' fUI'j •• nb 0 (-

dM~nn ina 'to n 150 n('od1·lentj jntJjjoD i ')'0 "Iiemujon Jill

"" milS of fgrm! HI ion em: rti ,oj 1m!

JW

CON.TENTS

U rheberrcc htlich gesc h uIZI"S M a\eria

CO><Ta<TS

Chllpter 14

xht

'Pro"""i"" of COu"IMion ooeflkienl .15Q Probabl e error 351

0, n fjdcncjl, I jill \1 Ii for P 152 T~~t Or s.igllific;:U].(i~: of;

i'orr_elat; OR for OJ ci '-""'- J P_

., specified "~Iue of P J52 Ego.lily of l\Yo COrt. <"ofr.. 352

Eg y.1 i Iy of more Ihon Iwo COrt; """rr. .152 Coefficient of coneu rre nl deviation 353

Correl 'li on coo mci enl uSing vorian co 01 th_ d i If_"" nCe 355 Correlation coefficient by the method of lema squares 356 Correlation ratlo 357

lntrsclass enrrelation 357

Biserial correlation 358

Tetrachoric cnrrclauon J58

Multiple linear fC~ression J.~9

Test or Sill"inc'ncc of pM;,1 regn.\Ssion coeffidcnt(s) 361 Cqrrtl;·unrm ind,~l' 162

Sl"'cificai ion of pOlen I iol pammelers 362 M "11,ir Ie, en"", [alion J(jJ

Reg .... .,;,; on in If; yon Ole POpul'lion 363 Partial correlation 364

PArt COWl.., I io n J65

Spuriou, correlation 366

Fill in the blanks 366

Moll ip Ie choic~ guesli" n, 372 A nswers 387

Suggc"ed reading J88

M",,~urcs of Associntion of Attributes

Shorr eMily type qucstlon5. On:

General discussion and notations 390 Order of a cl"" 390

Class frequency 392

Inconsistency of dOI~ 392

Kind. of o.~,odalj<>n 392

Methods of'mc.asurin{!: associariotu

Proponion method 39J Method of probability 393

Y"I.,,·, coeffldenl of association 394 Coefficient of colligo.lion 394

Partial essoci ation 394

Hlusory :I.'I,ociatlon 394 Coefficienl of contingency 395 Tschuprow' coefficient 395 Fill in the bank. 395

390

U rheberrec htl ich gesc h utztes M a\eria

CONTENTS

Mul1.ipl~ chnice questions J!l8 Answers 404

Suggested reading 404

Inferpclatlnn and EJ:lrap!llatioll

Chapler IS

405

Sil crl .... y type gu .OI;ons on: crt! eml disc]! Sision 4(}5 Assumption, and uses 405 [01"'P0l.1 ion m",bod;:

gr~phical m~lhod 4Dff

bi n om i.1 expansion me thod 406 ron-nbc [i.t:: clitvl! method ,406

Fin ue di fferences 408 Divided differences 408 Diagonal difference table 408 C~n'nd d ifference table 408 Divlded difference table 4D9

NeWlon's formulu of advancing differences 410 NeWlOn'S backword formula 4//

Ne,wlon-Gauu r,,~ formul. 41/ Newwn-Gauss ba.:I<wani formula 41/ Newton', method of b.e'ward differences 4/2 Newton', method of divided differences 412 Lagrange's interpolation formul. 4/2

Inverse interpolation:

i."smnge 'sinterpoletlcn formul. 41.2 Central interpol,at;on 413

Sterling,'s rom,.I.. 4/J

1I.,ul'. formula 414

Inverse in terpolatio n:

Lagronge's method 415 Itcrnti ve method 4/5

Suee ess ive approximatian method 4/5 Fill in the blanks 416

Multiple choice questions 418

Answers 414

Suggested readins 424

Clilipler 16

Time Series A.nalysl$

425

Sho" e'iSl) Y lype queslions On:

Drfj'!1;lions 4'):1)

Editing of dOla 426 S("(~It! cr I rcn d 426 Scason jll ytJ ";UiOD 427 Cyclic , .. i"lion 427 Irregular vlrinlio 0' 427

U rhcberrec htl lch gcsc h utztes M ateri"

Chapter 17

MCKie! s [or ti me seri eLI! 427

Moth odl 0 f mc".ilr'ns I inea, trend: gwaphic meChod 428

.emi ·QV""'go m. thod 428 Movinl! '''''''''ge method 428 Leasl .'9uar. melhod 429

Shifl of origin 430 CtnyjljuMt rrend 4JI

M~thoo~ of measuring seusonal variations: simple average method 433

rotio to trend method 433

ratio 10 moving average method 4J3 specific nnd typical seasonals 434 link relative method 434

Me~\ods of measuring cyclic variations: resi dual method 435

first difference method 415 percentage mtio method 435 direct method 435

reference cycle nnalysis 415 harmonic, analy.is 435

M"""urcrncnt of irregular variations 4,)6 Fill in the blank. 437

Multiple <hoit(: 9[''''lio", 440

An.we rs 446

Suggested read i nil 447

Index Numbus

448

Short essay type questions on:

Definitions 448 U"'" 449

Llmhations and tecunoe 449 Price index numbers 450 V'.lue and diffusion index 450

Problem. involved in the eonseucrion of index numbers 450 Laspeyre's index number 451

Paasche's index number 452

DrOOi~h·Bowlcy lndex number 452

Walsh (Fisher', ideal) ind"" number 452

Marshall and ,Edgewcnh mdex number 452'

Kelly" noed weight fOJmnl~ 453

Form' lla_ error 1lJ

Homogeneity error 454

TIme reversal tcst454

Factor reversal test 454

eiff L1]3 r '!eSi! 451

U rheberrec htl ich gesc h utztes Materia

xxii C<lNTENTS

Ch"jn b3Se fll,.almd 455 Base shining 456 Splicing 457

Consumer price index 459 Explicit nnd implici1. weiShls 46f Whol "-",Ie price i pd." 461'

lnde. crindustriol producticn 462 Oro,. n.ti 0.0..1 product. 462

611 in Ihe binD);. 464

Mull ip Ie. eh cke questi 0 n. 467

A m;;wers_ 474

Suggested reading 475

Chapler 18

n usiness ForeCllStin g

476

Shon essay .Iyre quesl i(m:!: 011:

DeGni'Iions 47.,."

H i.,wricol .,..Iy.is 476 .

Analysis of current economic conditions 477 Method, of (ore<:3<linS'

Naive ,nethod 4n

specific historical analogy method 477 lead-log rcl.tionship 477

Diffusion index 478

action-reectiun theory 47,~

[,,"(Or Ii.sting method 479

cross-cut analysis theory 479

opinion polling 480

exponential smoothing 480 econometric method 481

R>rI."".&ting age"ci"" 481 Fill in the blank. 482

M u I tip Ic eh oie<: questio.ns 483 An"wers 486

Suggested reading 487

Statistical Tccllniquc1i in Quality Control

488

Short es~ay ~ roe questions on:

Cb ['In('(1 file!(] r 488'

A." ig nable couse> 488 Sbewh.,n rODI wI chans;

X-chon 490 ,,- chon 490 fl- ehrU1 49'

Control chort. for fraction dcf"",iv,,", (p·chartl 495 Control chpniil (or numAAr of defcc!s (e-chart) 4%

U rheberrcc htl ich gesc h utztes Materia

xxiII

Ad VII" .... W'. 0 r .totj,ti COIl quality conlrol 499 Nit1urn I COn I ml 1 i mill; .499

Sp<eifiCOllion limits 499

Mndjfird CQn!ro' limhs 5t!O

AceoplOncc, sampling pl." SOO

I nspecI; 'm h¥ .uri bul,e.., 5o.l

I ",peeli On by vo nables 50.l

Delio'lion_Iii and exp'anations for:

roo jcr de recl 50.1 minoT'defeet 5'01 Prodllcer~ rkk 50' ColViumcr's risk 502 Acceptance qunHly I "vel 50.2

Lot tol.",n"" !?Cree"I'ge der.cI; vo 50.2 A •• ",g. 0" ISO ins qu.li Iy 50.2

'8 I ind So", pi iog 50.2

Average, sample number nnd cUI'Ve 502 Opcmting ch=,",i,5Iic f"nOlion ODd curve 50..1 Slogle '" mplio g i 0 ,peCI i on pi on 50.3

Dodge ond Rornin8 OC curves 504

Double sampling inspection pl.n 50.5 Sequeatla] sampling plan 50.5

Sequ~"lja] probabl I ;Iy ratio test 50.5

Ell _iv_lbg blanks roB

M ullip ie, choice q "esli 0.. 51 J

Answers HZ

SUMeSled re.ding 518

ChapterW.

Vital and Popu 131 I au S Inlistics

519

Short ",say type qU.'lion, on:

D!~nnitions 512

~

eoUCC1 in n or vita! ';inti stl c-s 52,0 rcgi stration method 5U

census e nu merati on, method 522 ""IV"Y method 522

""mpiing registration system 522 !'onn"! •• for .slim ali on. of po puJalion 523

Crude d,'.:l' h rn r.~ • 'S? 5

Spec ifie death rate 525 Standardi sed deatl, ml" 525 Crude birth .. rc 527

Age specific fertil i Iy rate J28 General marital fenility mto 528

U rheberrec hti lch ge5c h utztes M aleri a

xxiv C!lN1'alTS

Age specific mari~1 fertility rate 529 T<)tal !n'rilal fer1ilily rate 529

Total ferti I ity rate 529

MI!IIS= of population growth:

crude rate of natural i no re ase 529 vital index 529

gross reproduction rate 530 net reprodecnon rate 531

Replacement index 533 Life table 5JJ

Construclion of life table 534 U nemployment rate 535

Stable, and stationary popul.lion 535 Abridged life table 516

Cen 11'111 mo "alit)' rale 537

Fure" of ,,'oMli'y 537

<'ill in the blanks 539

Mull.iple choice questions 542 Answer;;, 5'49

Suggested ;re.,din~ 549

Basi e Experimental Design,.

Chapter 21

550

ShO',1 .".f Iype questions on:

Dct1njlion!ji'

ex perimcnt:11 unit 55 Q lre;Hmen. 550

Reg ui ,",men Is of " sood de.; gn 550 mndo m nWjgn ,. 5 ,

replication 55}

lorn! ,control }'!i"'

E. peri""", 1.1 mo, 552

DeICmlinntion of number of replication, 552 Si~ nnd shape of experimental units 553 ReIOli ve efficiencY of design, 554

An.1 flis of variance 554 '

SlaliSlionl model. for ex perimenW designs 555 Tcsu; [or Qrdered lret! !m,'n! mMn 9j'

le<l$! ,igo; ficant differe""" 557 StudeTI!~N ewmilll Ku"'. (eM 5 57 Duncao"s mul~ple mnlle lesl 558 Tukey.' lCSI 558

Contrasts {compuri'''""j 558

Pairwi,e and no np';rwi'~ "<lntmslS 55P Priori "TId posteriori con""," $59

Sf" """,hc d.,ign. 562

RnndomiUld de.>lgru; 562

U rhcberrec htl lch gcsc h utztes M aleri" I

xxv

C"m pletel y '0" dOITl ;l.ed do,; go 562 Cms:s ... d dnnifijtil'ioiJ 567

N.~S' Cd or h L(:rurthj ral r 1112'; fj CfI I' ron 567 RnndQmilCd block design (RBD) 56?

M; &.i og plol tech n ique in RB D 571 . 'Ptt:Ji'rninary lest of signifr(:il.nce (P'TS) 573 Latin "'l unre design (LSD) 574

MutuoUy orthoeonal win '9"= (MOLS) 576 G reXo· L~ti n oq llore 576

Mi ",i'8 plot tech n iq lie in LSD 577 Cm', 0'", d •• i~ n 580

FoCI.'; nJ "'pe rim en l' 581

Moin effect 58'

lnl'emrtion _58?

Method, or esl;",.lion ~nd .".iy,i. IbJOugh 583

con trusts 58";

modulo technique 584

CIj).;;OS Wb Its 581j"

Yuh'S~ mNbod i86

Expc"CtOO rue,n '9 uore< ;tnd their role 590 Compl.ele confounding 591

Parti.1 """roonding 59/

Confounding through;

{OilMen SIS 592

modo 10 techniq 00 592

Confuunding in "symmetrical (.1clorial experimcn15 591 F,ru.'l;O n81 replical ion 595

Re.<o1 u,in n or fr.cl; onal repl ;calio" 596 Spl it p lot des; go 596

Spl;I •• plil plOI des;sn602

Sill; I block. or sir; p plot de,; go 604 Coni onndi ng j n ~ (lli! pJ ot d~si gn 605 Fill in the blanks 605

Mu Itip] ~ choice questions (; J 2

A ns_wr rl_ 627

Su ggcs_tM rend i n ~ 629.

U rhcberrec htl lch gcsc h utztes M ateri"

U rheberrcc htl lch gesc h (jutes M ateri"

Chapter 1

Statistics in General

SECTION-A

Short Essay Type Questions

Q. 1 Define SIOlw.ies D$ given be Sir R.A. Fisher. \\1.1\.. \\I1lllis and ILV. Roberts: SI.ti'tics is not a

Ails. Si r R.A. Fish<:r defined siausues as. '1110 science of statistics is essenuelty a branch of applled mathematics and may beregarded as mathematics appl ied 10 Q bse rv ationnl data."

Q. 2 Give the definitions of staristlcs given by A.L. Bowley. Lovitt. W.A. Wallis and H.V. Roberts and en. Meyer".

Ans. The definition, given by statisticians named in the question .'" quoted below:

A.I.. Bowley,

(i) Stalislie. is tbe device for abbreviating and classi fy ing the stetements and milking dell!" the relations.

(ii) Suiti~I_ic:;:5. iiii the science of measurement of social phenomenon regarded: as a Whole in nil it. manifeSlalions.

(iii) Sioti stiesis thenu merical staternen I of facts in any departmenl of enquiry .. placed in

relation 10 -.ach other. .

1.6,'111: Statistics is the science wltich deal. with Ihe collecling., classifying, presenting, comparinS and interpreting numerical dala collected to throw light on any sphere of enquiry.

body of substantive knowledge, bUI a body of methods obl;l;ning knowledge.

Codl H. Meyers, Statlstics may be defined as a science of numerical infortn3tiQn which employs the processes of measurement and ccllection, cla ss [; fication. ,naly.,i s, decision-makmg and communi· cation of res u II, in • manner understandable and veri liable by other.

Q. 3 Gi ve th e statements given by A.L. Bowley and WiUiam A. Spurr and Cherles P. Bonini.

An.... A.L. Buwley: Great numbers an: notceunted correctly 10 n unit, tl.,ey are estimated,

William A. Spurr .nd Cbarles P. Bon[P]: Not.11 numbers are 5tatisticat Logarithms, Ior Instance, ate mere abs: mct numbers, Statistlcn \ data rue concrete: numbers, which represent objects.

Q. 4. Quote S1alements about 'lotist ics made by A.S, W'ogh, A.L. Boddington, Whipple, Tippet, W.l. King. M.rshall, Yule and Kendall, RA Fisher,A.M. Mood, Disraeli an~ Darrel Huf.

A.E. Waugh: The purpose of statistical method. is 10 'impury greot bodi es of numerical data.

A.L. Doddington: 11,1. essence of staristics is not men: counting but comparison.

U rheberrec htl lch ge5c h utztes Iv! ate,i"

2

Whipple, Statlstics enables one 10 enlarge his Horizon.

L.H.c.nppel,

(i) Planning is the order of the' doy and whhont statistics planni,ng i.5 inconceivable.

(iiI Statislics is both a science and on on.

W.I.King,

(i) The science of statistics is a most useful servant, bul only of Kreal .alue to lhose who understand i IS proper use.

(ll) The science of staliot;"" i. the method of judging cuUeclive,. natural, or soelalphenomena from Lhe results obtained by the analysi, of en umeration or collection of estimates.

Manhd, St.uil;ti", are the, straw out of which I like every other economist have to make bricks.

Yule aDd IUJMlall' Statistic. is not. science, it is n scienti fie method,

R-A. y ...... : Statistics is a branch of applied mathematics which .pecj;ilis.s in data,

A:M. Mood: Statistics provides tool, aod techniques for research workers.

Dis.aell: There are 1hree kinds of lies; Ii es , damned lies and stansties.

Darrel ,HPI: .A weD wrapped s!;ltistiC$ i, better !h,an Hiders biglie it misleads; yet it cannot be pinned on you.

Q. S. Wh ieh definition of statistics is considered to be the best.

Ans. The definition of !laliSlics given by R.A, Fisher is considered 10 be the best nnd must exact. Q. 6. Give in I rew woods the, 'latistical perspective, AIlS. Statisticnl perspective i$ the invaluable compendium which gather' .11 the facts. ligures" objective survey and fascin'ling remembrances to. assimilable recosd,

Q. 7_ Mention main divisions of ststlstics,

A.... Following are the main divisions of statisncs: (;) MalhmU!t;cal or th40reiical statisucs: lt covers deve1opmeol of Slall!lical distributions" experimental dosips, .sampling designs, ole.

PR~.ED SrATlSTtCS

(ii) Stolislical melhod. 0' /unctions: It covers collection. tabulation, analysis and ;nlerpretation or data .• etc.

(iii) Descriptive tlari<lks: Cla .. lficatlon and di •. gramaiic representation of data.

(Iv) Intef1!"lia/ statistics: 1b draw conclusion abou I population On the basis of sample drawn ~m~ •

(V) Applied stasisucs: [1 mainly covers population, census, n.tional income. production, oo..iness .tatiSlics. industrial .tati..tics. quality control. bioSllltl,tics, etc.

Q. 8. What are the limitatlo!l$ of SIllliSlics? A.II$. S!OQdly the limi tations of statistics are"

(i) Stetistics deals with quantitative d'll! only.

Even qualitative information is converted into nu merieal dlWl by the method of ranking, seonng or scaling.

(ii) Stati.tie.;" true on an A.erogo only.

Wi) St.ti,hc, deals with' the masses, not aD i ndiv.idual. No .talislic~ is API'I icable for a single observetlon,

(iv) Stall'tical resuhs "'" """"'" in • general sense. They ore alway. subject to certain .mount of error.

(v) Slatistics i. only a means to d~l'i conclusions about rnllUCll or population but not • panacea 10 111.1 son of problems.

(vi) SI:Ili.tics cat! be misused in many ways.

Q. 9 What ore different types of investigations? A..... There ore IWO Iypes ofinvesligalions,'narnely: (i) in"e$lig'li.on through census method

(ii) investigation through sample methods.

Q. 10 What docs census method imply?

A.... Cen.,l!S method means t.o i nclude each ~Qd eve,y unit or objecl of the population under reference for enquiry Or ob.ervnti on. For example. 10 know fhe national income, we have to include every individual or unit which contribmes towards Ih. national lneome,

Q. 11 What i. meant I»' investigation Ihro"gh sample method?

U rheberrec htl lch ge5c h utztes Mater; a

ST ... TlSTIC$ IN GENERAl

ADS. In sample methodan investigator has to select 'lome units from the population about which eoncluslons have to be, dm'Wn and take observmions on the selected units, The results obtained from sample values ere applicable to the population as • whole. For i nstance, to know the average age at marriage, an investigator selects 3" oOequote number of married couples and arrive at an average oge of marriage which is considered to be the ev erage age of marriage for the whole popularion.

Q. 12. Whot a tre four main. functions of stausucs? Ions. Four functions of s!alisties are:

(i) Collection of d.UI~

(ii) Presemation of <lat.: (iii) Anoly,i, of dOll>; and (Iv) Interpretation of results,

Q. 13 Give different method. of collection of data Ans. Following are the methods of colleeucn of data:

(i) Direct personal enquiry method: (il) I nd irect oral investigation;

(iii) By filling of schedules;

(iv) Dy moiled questionnaires:

(v) Information from local agcol.$ ond correspondenis;

(vi) By old records; and

(vii) By direct ebservatinna! method,

Q.14 Name two kinds of statistical dot a u nd describe them In brier.

Ions. ·lWo kind! of $latisticol dala are:

(i) Primary data: Primary data are those which are collected frnrn the units Or individual, di""Uy and these data hove never been uoed for any purpose ea rlier,

(i i) Second"ry data: The data, Which hod been collected by some individual or ngcnq" and statistically treated to draw certai n conclusian s. Again the same data are used and anal ysed 10 ex rract some other lnformatlon, are termed as ,econdary d.ta.

Q. IS Wltat are the requisites of. reliable dOl.?

3

Ans. The rcqujsltes of a reliable data ore: (I) It should be con1pltle~

(ii) 11 should be consistent;

(iii) It should be accurate: and

(iv) I t should be homogeneous in respect of unit

of information.

Q. 16 What precautions should 00 taken in the planning of" statistical .u,\'\.')'?

Ions. Following precautions are to be tak .• nin the plan oi ng of • survey:

(i) PU'l"'se: FiT<! a clear-cut objective of the ,urvey should 00 spelled out,

(ii) Scope of SII''''Y' Different •• pects to 00 covered to achieve the fixed objecti ves s nould dearly be explained,

(iii) Dqini)/on of terms: All the terms involved in a sur<ey should be defined without ambiguity sn lhat. no unit is likely to fall in more than OnC category,

(iv) Slatillg 1M hypolh~.ir. Hypot.hc.," to be tested (rom 'he data collected by wny of su rvey should 00 lnid down in accordance wi Ih tbe objective, of the survey,

Q. 17 Give briefly the ch ar acteristlcs of a good qu estiennaire or a schedule,

Ans. Charactenstlca of :1 good questionnaire 'Or 8 "'hodule are:

(l) Number of quesdons should be such that it e.tmel<.n information required for the report, (ii) ElIoh question should hove ;:dmo", 011 alter· native an_'5WCJS,

(iii) 11,e question should be. cl ear and without ony ambiguity.

(iv) All questions should be mutually exclusive in nature.

(,.) Some vet)' personal questions be avoided. (vi) Question.n';", or schedule should DOL be very

I.engthy and t imcc<:on,umi ng,

Q. 1 S Nome fi ve fields where st.tisti.cs is inevi table. A~ Broadly five fields when! 5lati11ic;!' is me ... i· tab Ie can be named as f 0 llews:

U rhcberrec htl lch gesc h utztes M ateri a

4

(i) Scieruiflc research; (ii) E<;o!lOmic ~n.ly.if; (iii) Planning;

(iv) Business and commerce; and (v) For.ca.ning and projection.

Q. 19 Mo.nrion diffetcnt kind. of,tati.tica( ifl\'estigations,

ADS. Different kinds of !<I.listie.1 inveslig a tiens arc:

(i) SUI\"Y' or experiments:

(ii) Surveys through census or sample enquiry; (iii) Confidential or ope" enquiry;

(iv) Direct or indirect enquiry;

\V') Origin.l or repetitive enquiry; (v i) Regular 0' ad hoc enquiry; and (vii) Limited or extenslve enquiry.

Q. 20 Wh al is an absolute biased error?

AIlS.. When the figures orc rounded straight way to the nearest lowest unit of rounding or to rhe nearest highest unit of nlunding, lite difference between acjual and estimated (rounded) values in the two cases ore ",,11.d· biased errors. For example, if we round the value 357 10 the nearest lOO~ the nearest lower value is 300 and nearest higher value is 400. In, ca. se I, Absolule biased error", 351 - 300 = 51. In case n,

Absolule biased error = 357 - 400 = - 43.

If there are I,"" or more values in 0 set, the sum of absolute errors is taken,

Q. 21 What is an nbsolute unbiased error?

AItS. If rhe va lues are rounded as per !be rule. of rounding! i.e._ •... given value is rounded to nearest lower va Iue of the unit of rounding: in ease it is less than half of the unit of rounding and 10 next higher value of the unit ofronnding if it is more than halfof Ihe unit of rounding, il is celled unbiased error. The difference between the actual value and the estimated (rounded) value is called absnluie unbiased error. !C, there are two Or more values in a set, then the sum of the absol Ute, unb iased error is taken.

Q, n How do you estimate an average unbiased ahsolute error (A,B_).

PROORAMUED STA T1ST1CS

AII$.ThcFormula i.,

Avcrngc A.B. (unbiased) "AVClage error" .r,; where, n = uum ber of items in the set and Average error » The mean of die minimum and maximum va Iue. Which are likely to be left over or increased in lite process of rounding, For example, if we are rouoding a value 10 lit. OOlU"C$t I OO,lite chances are th.1 the lowest value which may be left out is I} and maximum value which may be, added is 50" Hence. the ",'=go =0' is (0 ... S{I)/2. .. 2.1.

Q. 23 Enunciate lite law of 'tatlstical re guIMity. Aali- TIle law states that a reasonably Jorge number of items selected al "'MOIn from a I"'g" group of item, will. on. the average, be representative of the lorge group or population. This law is governed by the theory of probability.

Q. 24 State the law of io..,i a of large nunibers. AIlS. The law ofineni. states th.t lite large, aggregates = more stable than small ones, According to . Professor A.L Bowley, great numbers and "vernges

resulting from ilicm",uch as we always obtain measuring soc ial ph. nomena have a great i nerti a, Q. 2S What is the law of persistence or ,small number?

Aali- The law of pe rs iste nee ohm.1I numbers states that the ratio of the small number of items having some distinguished characteristics to lite total number of units in lite population remains constant even through the population size I$. immensely increased.

Q, Ui Give an e •• mple of the law of persistence of small numbers.

Aas. Suppose a school admits all 80M students, Even the!l somestudents wi II be of poor intelligence. The ratio of such students 1.0 the rotal number of ,tudent. will remain the same even if lite number of students in the, school. is doubled, trebled, ere,

Q, 27 SUlle the law of decreasing variation?

AIlS. Law of decreasing variarien indioates thai the variation In D sample lends tn reduce lIS the sample size i ncreases.

Q.18 How is the law nf deereasing variation helpful in sample 'UIVey,?

U rhcberrec htlich gesc h utztes M ate,i"

Ans. II is the law of decreasing variation wh ich pul$ an inV<!$ligalOr on sound footing!O decide about the adequare ,amplc size which is a eucrepresentat;", of thc population.

Q. 29 What db you understand by approxlmatlon of values or figures?

A ns. To express a value, or fig"", to a rou nd figure which is "'I5Y to writ" and understand is called appro><im;UiOJl. 'This is mo>~y dcce from eenvenicnco point of view. It helps in comparison of values tremendously,

Q. 30 Give different method, of approximcrlon wi th ., brief description.

Ans. OilY"'''"t methods of approxlmation are,

(I) BY' adt/ing jigurtl: In thi' methods the given value of figure is always i ncreased to nex l h igher val ue of unit of rounding. r"Or i nsiance, a value 21. ~S7.4 will be appmximated to 21,3.58 up 10 unil place, 21, 360 up !(I tenth place, 210400 up to hU.ndr.dth place and 21,000 up 10 thousandth ploce.

(ii) 8Y' disctmJing jigurtls. It is " process just reverse to the adding figures. In this method, the given value is decreased to next lower value of unit of rounding. For exa mple, the value 21,l57.4 is approximat e -d to 21.~57 up 10 un it place. 21.350 up to tenth place, 2.1 ,300 up 10 h undredth place and 21.000 up to thousandth place.

(iii) Approx;mation It) 'h~ whole ,'lumber: This method is also known as rounding of figure, nnd is an usually accepted method, 11,i. method i. the best one as il min i '0''''' the error of approximation. In this method. a value is raised to the next higher value of tbe unit of rounding ifit is more, than half of the unit of rounding and is len over if il is I ess than. half of the unit of rounding.

Q.31 What are, different source. of statistical errors?

5

Ans. Following are the four sources of statistical errors:

(I) &0 rs of origin;

(ii) Errors of inadequacy:

(iii) Errn .. of rnani pulauon; and (iv) Em'~ of inlcrpn;ltuion.

Q. 32 Explain briefly the possible error.

Ans. In rounding or val ues to the nearest of units, tens Or hundreds or thousands, etc .. ,' value IIlSS uian half of (be unit 01 rounding is left over and greater than half of the unit of r,,"ndins is increased 10 the next higher "nil nus, the possible error is" 112" unit of round ing. For example, if the number 23 is rounded 10 the nearest ten, its value is 20 and possible error is:t 5. Hence the value will lie between 20 :t 5. i .e., between I Sand 25.

Q.33 Wh~I"" different sources of primary data? Ans. DaUl obtained from origin. I e. per; ments Or surveys. i.e., the dala co)lected by ; nvestigators or enumerators is known as prim"", data, A Iso the census data. d.13 rcleased in the Reserve Bank of India bulletins. da .... published. by other authorities in original form are Con,jdc",d as primary dura.

Q. 34 Whal are different source; of secondary dala? AI!$, Published thesis, research papers, project «pons. summarised census repon, monthly .b'tr.octs or CSO and different publications of trade and commerce associations, etc., art: the 'Various sources of secondary data.

Q. 3! What ki nd of deficiencies of data are cheeked through editing?

A!IS. Data are edited to remove m~inly four deftciencies 'Which are:

(I) Completeness of data; (iil Consistency of data; (iii) Aocur>cy of data; and (iv) Homogeneity of d.UI.

U rheberrcc htl lch gesc h utztes Iv! ale,; a

6

SECTl0N·B

Fill In the Blanks

Fill In the suitable WMl(s) or phraSB{s) in the blenk$:

I. The statement, "Not all numbers are sutistical; log.mlhm$ for instance, are merely abstt:oc( numbers. Statistical d~ta are concrete numbers, which represents objects", WM'

given by and '

1.. ''Greal numbers are nOl counted correctly 10 a un.il, they are estimated", i, tbe statement of

3. The d~finition, ~'Slati~liC:-$ i:.; the science which deal.'! wilh the collection, eh ... ification and tabulalion of numerical facts as the basis for ex plan arion, description and comparison (If

phenomena", ""as given by _. . "

.4, "Slati'lics is 3 body of melhod, for making wise decis.ions in the face- of uncertainly" is

lhe definition of '!nli.stic< given by _

and .

5. Statistics is both. 0

_____ and an

6. The eredlt of the statement, "A statistician is a pracunoner of the an and ,cjeo~e of

sl..1.tiSlicst! goes to '

7. '111c purpose of statistical methods is 10 .implify gre3I bodies .of numerical data" is the ',l'ltement gi~en by

8. The definilion. "The esse nee of ,tali.ties is n,ot Il1Cf1) coon~ng wtcomparison". was id"""

by '

9. The statemem, "Statistics enable. one 10 enlarge his horizon", goes in the name of

10, The statement. ~PI.nnjng is the order of the day and without ,tnt"I;"" planning is

inconceivable" w". given by .

II. 'TI,e author of the staremem, "The science of stansucs is. mOSI useful servant. but only of great value to those who understand its proper use", was due 10 ~.

12. Sb>(j.\~ ••• an prou ,

13. Usc of statistical methods" most danproos

in tbe hands of .

14. 11l<) '!Dtem.nt, "On average a factory labour has become younger in 1991 as compared. 10 198]"', is

15. Stati'tics deals with only .

16. Statistic<d an.1 ys" helps in (he of results.

17. Statistics is not applicable to _

observation.

13. NOl" hut data an! the subject-

maHer of stetistlcs .

19, Stntlstics ore numerictil of fllClS,

but .U numerical statements are not

'10, By stausncs we mean quantitative data

affllCred 10 11 marked .xt.ot by _

of causes, (»'1" do K.ndalf)

11. SWislics docs not study

11. Sttltisu.cs L, not" science, its a "

23. Statisties are, the straw out of which r like every other economist, have to make ___ , (Manhall)

24. Slati.ticsis the arith meuc of human

25. Planning on the basi. of inadequale and inaccurate statistics is than no plnnning at all. ('Thin/·Fiv.·Y,ar Plan. Plan.ning Commission).

26. Statistics is liable, to be .

27. The d.t. collected from published reports is

known data.

18. Data obtained by conductl ng n survey is called data.

29, Before analysis, the d.ta.hould be

U rheberrec htl lch ge5c h utztes Iv! aleri"

ST.o.TI$TICS IN GENElW.

30. units are better than •• birrary

units,

31. ore used io a mailed enquiry

method.

32_ M~ilcd enquiry method cannot by !!darted if

the respondents ore, "

33. In personal enqui ry method, tbe response is

beuer than method,

34. l'rc1C$.ing is e.ocnl!:d ror preparing " good

or ~

35. Population figures published by the Census

Commissioner oro data.

36. Mi.tak", and swti$lical errors ate .

n. lf a quantity is such that all errors tend to be in the same direction" they nrc "oiled errors.

38. TI,e errors caused by the carelessne .. or Ihe

lnvestlgators are called errors.

39. 'Formula for the estim:uion of biased absolute error is

40. Formula for the estimation of unbiased

absolute error is

41. Iliascd relative error can be CSlimlUcd by Ihe

formula .

42. Unbi=d relal;',. error can be eslimaled by

lhe formula .

43. A survey in wbich informati on is collected from each and every individual of the

populationis known as .

7

44. A"igning number digits to various response whether quantitative or qualitntive, is called

45. A figon: 16..118.7 rounded to !Ill: neateSltenth

pl"""i. .

46. The figure 32,617 rounded to the nt:lrn$l.

hundredth plaa: is ~

47. The figure 32,627 approximated to lit. thousandth pl~ by the method of dtocaroing

figure is ..

48. The ligon: 45.067 approximated to thousandth place by Ihe method of .• dd i ng figure ls

49. The figure, 13.85 rounded 1(> nllC decimal

plnces1 .

50. 'The flgure 13:75 rounded to one decimal

place is '

51. Government cannot do proper planning

without the help of ,

52. Ob$..-v"ions collected tbrough $I!TVC)"' or e<periments are classified lIS

53. To know the area under cultivation of wh",,!, lbe appropriate type of inve'tigation

i.~ _

54. To kn ow the a verage yield of " crop. an 'pprop.i.,o investigation Iype will be

55. The compendium which gathers ;))1 the f""iS and rascinating memories in a assimilable record is known a~ __ ~~

Multiple Choice Questions

SECTION-C

Select the COI7l3Ct altemativ& out of given ones: Q. 2 Who stated that statistics is a branch of

Q. 1 The .taternent, "Statiotic. is both a science and on an", was given by:

(a) RA Fisher

(b) TlPPOl

(c) L-R. Connor(d) A.L. Bowley

applied mathematics which specialises in data?

(a) HOf1ICjl SCl'M (b) R.A. Fisher (e) Ya·lun-chou (d) L.R. Connor

U rheberrcc htl lch gesc h utztes Iv! ate,;"

B

Q, 3 The .... ord 'W.!illic,,' isused lIS' (0.) SingulM

(h) Ptural

(e) Sio!gulw.od plurnl boIh (d) none of th.c above

Q. 4 "SUlli'lics provide. 1001. and techniques (Of research "'(11'1:=", was .!Ialcd by,

(a) John 1. 'Griffin

(b) W.t'l<lng

(e) A.M. Mood

(d) A.L. Boddlngton

Q,S QUI of various definlnons given 'by the following workers, whieh defin'itiQIl is considered 10 be mo,1 Hocl?

(a) R,A. Fisher

(b) A.L. Bowley

(c) M.G. Kendah

(d)Cccil H. Mcyc rs,

Q,fi.. WIle stated thai tbere are three kinds of liel: lies, drunned lies and O1aliOlics.

(A) Mntk Thin

(b) Disroeli

(e) DJUTeI1 Huff

(d) none orlbe above

Q, 7 Who gave !he Slllleme"~ ~A well .wrapped .wtimet i. beller '(han Hiller', 'biglle', il misleads, yel it .""ROt be pinned on you."

(u) Mark, TWi.n

(h) WA Ne;$WIm~er (0) DIIm:lI Huff

(d) o.w Snedeeer

Q, 8 Whic,h of !hero1l"",;.g represents d.Ia', (a) a oi"gle, value

(b) only IWO values in D, 6<1 (e) "group of \Illlues in" set (d) noee of Ibe above

Q. 9 Sntislies deWs wilh:

(a) qualitalive inform.tion (b) qU:UlIilllli •• infoml3lion (e) both (a) and (b)

(d) none of (a) and (b)

Q. 10 Slalislical results an:,

{aJ <%1" per-cent ccoect (b) not.rihsolulely oom:ct (c.) .a!"",Y' incorrect

(d) misleading

Q. 11 If 'a\js,Ibc.acru.al value and "e' is us estimalcd value. the ab so lute enor i&:

(a) a-.

(b) la- .• I

(c) ale

(d) {a - ~)I.

Q. 12 If 'a' is the oc:tunl value and 'e' is its wimated v41ue, !he fomtula for relDlive error is:

(0) aI.

(b) (0 - .)1. (e) I a - e . V. (d) (a - .)la

Q, 13 Dam taken from, lhe publicati.O<1, 'A.gricuhllml Siluarhm in lMkJ' will be considc:n:d 4$:

(a) pri~~

(h) second:u:y data

(0) primary and ...:ondary <!aLII

(d) neilher primary nor """,ndory d.1II

Q, 14 Mailed questionnaire meIhod of <enquiry can be adopIcd ifrespondcnts:

(~) live in cities

(b) bJlve high incom. (c) are edllClllcd

(d) nrc ,known

Q. 15 The stl1t:meftt. "Desi~ning of lin approprialc questionnain:,imelfwins Iudf!he'b:llll.,~, was given by:

(n) A.R- nome (b) W.J. King (e) H. Hug)' (d) H. Secrist

Q, 16 Sl3Ilstical data an: collected for,

(a) collectin,g data wilhoul ..,y pu:pose (1)) a given pWJlOSl'

(e) any purpose

(d) none ohho abo"" Q. 17 Relative enur is al .... ys: (a) positive

U rheberrcc htl lch gesc h (jutes M ateri"

ST,o,nsnes !N GENEAAl

(b) neB'lli""

(0) positive and negative both (d) zero

Q. 18 Stali<tic(J] error refers 10:

(~) Originnl value=- Approx. value (b) AclU~1 value - Estimated value

(e) Actual nlue- Estimaled value

Estirnaled val lie

Q. 19 Method of complete enumeranon is "prl icable foc

(a) Knowing the production

(h) Koowing the quanunn of export and import

(e) Knowing the popu]01;on (d) all the above

Q. ,20 A statislical population may consist of: (a) an infinile number of ,itom.

(b) "finite number of items

(e) either or (n) and (b)

(d) none of (a) and (h)

Q. 21 Wbieh of !he follow;'1S CXMlpl" doe! not constitute an ;nlinile populalion?

(.J Population consisling of odd numbers (b) Peputation of weights of newly born

babies

(c) Population of heights of 15·year...,ld children

(d) Population or head and tails in to ss ing a com .ucces,;vely.

Q.22 Which of the following "an be classified as hypolhc!ie,,1 population?

(a) All Iabourers of a factory

(b) femme population of a country

(e) Population of real numbers between 0 and 100

(d) students of the world

Q. 23 A "udy based on complete enumeration is

known as; -

(,) sample survey

(b) pilot SUl"Vt.'Y

9

(e) census survey

(d) none of the .00".'

Q. 24 If tbe actual value of 0 unit is 415 and i IS eslim,iltcd value is 400. the absolute error is:

(0) -15

(b) 15

(e) 0.0375

(d) - 0.0361 ,.

Q. 2S If the estimated 'volue of an item i. 50 and it, ootu.1 value i, 60, the relative error is:

(a) -20

(b) 0.16

(e) 1.2

(d) 0.20

Q. 26 Who or.igiMUy £II"" !he formula for !he estimalion of errors ']

(a) L.R. Connor

(b) W.I. King

(e) A.L. 'Dowley

(d) A.L. Boddington

Q. 27 Boddinglon gave the formula for !he estimotion of en-on: of the type:

(a) Absolute error biased

(b) Ab><>lolll error unb~

(e) both (a) and (b)

(d) neither (0) nor (b)

Q.211 Bnddington's formul. for ... im~ljon of absolute error (A. E) is:

(a) ,01.1 A.EI.Jn

(b) Av era gc A_E x.Jn (c) Tot.l A.Eln

(d) Av"rnB" A.E x n

(n = No .. of items)

Q. 29 Bcddingtnn's formula for estimanon of relati ve error is;

(0) TOtal A.E.Jn

(b) Avemge A.Eln

(c) Avcmg<: A.Ex,"

~

(d) Average A.Ex..rn

"

U rheberrcc htl lch gesc h (jutes Iv! aleri"

10

Q. 30 Bowley's formula for absolute unbiased error is:

1

('),r,; Averag. (A.E)

2

(b) 3--,r,; (AverAge A.E)

2 . .r,;

(e) -3- (Av .... g. A,E)

3

(d) 2J,; (Ay"",ge A:6)

Q. 31 Staustical results ore: (a) .bsolutely correct (b) not true

(c) true on avelll8e (d) uniYemllly true

Q. 3Z "The statistical law(.) based on trio! and error methods islare:

(a) law ofstati.tical regularity

(b) law of inertia of large numbers (e) boIh Jaws (a) and (b)

(d) none of the laws (a) and (b)

Q. JJ The figure 32.64.616.8 epproshnaied 10 the len III pi""" by Ibe mclhod or di5cardinS figure is;

(a) 32,64,615.8 (b) 32,64.616 (e) 32,64,62{1 (d) 32,64 .61 0

Q.34 The figure 32,64,616,8 appo:ollimafCd to the leplb pi ace by addinS fisure. is:

(oj 32, 64, 6,]5

(b) 32, 64. 616

(e) 32, 64. 620

(d) 32, 64, 610

Q. 35 The figure 26,476 approximated 10 tne hundredth place by discarding figure is:

(a) 26,400

(b) 26,500

(e) 27,000

(d) 25,000

Q. 36 The figure. 47, 616 approxilfllUed 10 bundredlll . place by adding (!gUm is:

PRoGI'IAYoIEO sr ... T1ST1CS

(a) 47,630 (b) 47,620 (e) 47.700 (d) 47,600 -

Q. 37 The figure 43,572.6 approximated 10 the thouWldlh place by discarding figure is:

fa) 43,500

(b) '43,000

(e) 44,000

(d) 44,500

Q. 3S The value 43,572.6 approxlmered to the thousandth place by adding figure is:

(A) 43,500

(b) 43,000

(e) 44,000

(d) 44,600

Q. 39, The figure 4S ,986 approximated to the leo IhollS3Jldtb pl_ by the mclhod of discarding figure is:

(a) 40,000

(b) 46,000

(e) 45.500

(d) 45,000

Q. 40 The figure 45,986 approximated W len lhoosandlh ploo:: lty !he method of adding figure is:

(a) 50,000

(b) 46,000

(e) 40,000

(d) none of the abnve

ANSWERS

SECTlON.B

(I) Willi"", A. Spun and Cbarles .P .. Boaini (I)A •. L. Bowley (3) Lovitt (4) W.A. Walli. and H.V. Roberts (5) SCience, art (6) C.H. Meyers (7) A .s, Waugh (8) A.L. Doddington (9) Whipple (10) Tippet (ll) WJ. King (12) anything (13) tnexperts (14) acceptable (IS) quantitative dal.a (16) interpretation (17) 5i ngle (18) datum (19) .slatements; $talistics (20)mu!ciplidly (2]) individuals (22) scienti.fic mclhod (23) bri<:ks (24) welfare (25) worse (26)

U rheberrec htl lch gC5C h utztes Iv! ateri a

STAnS'!lCS IN GENE!W.

misused (27) secondary (28) primary (29) edited

. (JO) Physical (31) QuC!l,ionnai"" (32) illilerale (33) mailed enquiry (34) schedule; questionn",,,, (35) primary (36) QOI same (37) biased (38) unbiased (39) Average A.E x No. of items (40) Average A.E

x JNo_ of i,."" (41) Biased A.E/E5'imalCd v~ue

UnbillSed A. E

(42) Eslin:uued value (43) census survey (44) coding

(45) 16.320 (46) 32.600 (47) 32 .. 000 (48) 46,000 (49) 13.B (SO) J 1.8 . (51) "wries (52) primary data (53) "eMU' method (54) wnple method (55) ,lali.ticar perspective.

SECT10N.c

(I) b (2) b (3) c (4)c (5) a (6) b

(7) c (B) c (9) b (10) b (II) b (12) c

(13) b (14) c (15)e (16) b (17) c (18) b

(19) d (20) c (21)0 (22)0 (23) c (24) b

(25) d (26) d (27) b (28)b (29) d (30) b

(11) c (32) c (33)d (34) e (35) a (36) c

(37) b) (38) c (39)8 (40) n 11

Suggested Reading

1. Agarwal.B.L.. BaSic 5lalislics. New Agc Intem.tianol (P) Ltd. Publishers. New O<:lhi, 3rd edn .• 1.996.

1. G"p1ll. B.N .• SlaliMic-s, Saltily. Bhawan, Agm, 3rd OOn., 1.978.

3. Harvey, J .M., Sources· of Stalistics. Clive Bingley. 1969.

4. Mc:Ou-lhy, ])J .. IlIlrodur:lil)7I to $rQli$rkal Reasanin,. McOraw"Hi11 Book Company. New York. 1957.

5. Monmncy, M.J., Famfrr;m Fig"''', Penguin Book·,. Baltimore, 1959.

Ii. Reichman. W.l., Use ami Abuse of Statistics, Penguin Books. Baltimore. 1961.

7. Simpson, O. and Kafka, P .. Basic SllIIiSlics.

Oxford &. illH, Calcun., 3rd edn .. 1971,

8. Snderson, T. and Seleve, S" An Inrro<iuClioll to lh. Stalisrkal Ana/ysis of Data. Houshton Mlffl in, ·Boston. ,978.

U rheberrcc htl lch gesc h (jutes /vi ateri"

Chapter 2

Classification, Tabulation and Frequency Distributi0n

SECnON-A

Short Essay Type Questions

Q. I What is meant by classmca~O!I?

A.ns. Cl""ific"uon i. tbe process of =gl_ng thi ng. or item. in. groups or classes accordi ng to their resemblance and affinities and give expression 10 the uniUl of attributes that may subsist amongst the diveniry of jnd.ividunls.

Q. 2 What are llIe modes of cI .... mc,o~on? All$. Oi fferonl modes of classificannn ore,

(i, G.ollmphlcaJ ow.lftc<Uion: c1assirleation is according 10 place. area or !Cgion.

(ill Chf(NIOUJIlIr:a/ cl4$lliJicatiOI!: 11 is accord'.& to the lapse aftime, e,g" monthlY, yeQrly, etc. (iii) Qualila/iv. c/assiJU;aJion: Data are classiMd according to !be .uribures of the ..,bjects or items, c.8 ... sex. qualification, colour, etc, (Iv) Qu.anll,atIv. cln:sfi/icotio'" O:>Ia;ul) d~ficd according 10 tII.e magnitucle of lhe numerical values. e.g .• age. income, he;gh~ weight, cle.

Q.3 What are the objectives of c!<!S$ificalion? .........Broadly. Ill ere ure s i.x objectlves of classlfie:ttion.:

(i) To present the (oct.< in a simple manner.

(ii) To highlli:h( itents which possess or do not posSCII certain IlItribotes or qWllilie&

(iii) To provide. help in making comparison between items.

(iv) To find out mutual relationship between cerll1i n measu= 4IId their effects.

(v) Thpresent tile dab in a manner which is $u;t:lbl~ (or further treatm ea t,

(vi) "Ib prov.ide ""'is for ~bulalion.

Q. 4 Wbat do you understand by quolitativc classlrw:aulIQ?

A.... It is the classification on the basi, of certain attributes or some qualities of items which can not be rneuu:red quantitatively.

Q. 5 Descri be in brief different kind. of c lassiflcation,

A.... DiffcTe>lt types of cl .... ifit:alion =: (i) Classification according to altributer;.

{ii) Simple 'or two-fold or dicltOlomous cI ass ification

(iii) Multiple clll$$ificruion.

(lv) Quantitative cla. ss lficaaen, i.e .. the elasslr!Cation according tovarillte val ues,

U rheberrec htl lch ge5c h utztes M ateri"

Q. 6 What do you, understand by mulli[3<:to,. c1as.ification?

A.... elllssi (xation criterilt based 00. two or more facto,. (attributes) is known as mullifactor cllWification. In this Iype of classification, fin' the data ore classified into two. "" more, cia sses on the bll$i$ of one factar" For each co.mpo.nent el ... ificltion, funher classification' is done on the basis of oocolld facto":Uld so on.

Q, 7 What do you understand by open ClId(s) in gtoup data?

A.... U in grouped classes, me lowe, limi; of the beginning class is not specified andfor the upper limi, of the rugbes! (last) class is not specified, ir is known os grouped data with open end c13S$( es ).

Q. 8 Whnl; nrc: different characteristics of cl as sifje"pon7 Describe eacb clwucterislic in five lin e s.

A .... Different ch3tOCleristies of cLassir.",,!ion ore:

(i) E.r/r4u$tiv~: The c1_ i>hould be such that Ihey cover every item of the set, i.e .. , they should also be complete- and noo-overlapping. For instance, for moriraJ _ the classes should be married, unmarried, widow, widower, divorcee, deserted,

(ii) Slizhility: ClassifICation should be unifOlT'l or sIluIdardised so that the results are COI1lJ'II"IbIc at dirr""'"t occasions or in dlll'erenHrudies. (iii) Flalb.ility: Classification should be amenable according 10 different silUlltioos or requ ireroeOIll of Sl.ady.

(i,vj Hvmog."my: The units of measuremem of all cw..:. should be same. Also like units ooly be accnmmodated in one cl;)S$.

(v) S~itt1bility: Classification be done, according te the objective of the sLUdy only. For insUlllCe, 10 .rudy the financial status of people, it will be wei ess to classify \hem according to their skin colour or their hair colour, etc,

(vi) Arilhnmic accuracy: The sum of number of uni L'l in all classes shOllld be equal 10 the 10101, number of uniu. Also in case of observations, !be sum of observations in "'1 ctasses .houW, be equal 10 thee sum of 011 ebservations,

13

Q. ,. How can onedelml1lne !be numbcrofclosses for a frequency dlstribution?

AIlS. In qlUUltitative classification, dle number of classes depends upon !he cla .. interval. So a formula was suggested by H..A. Sturges 10 decennine the class interval !U!d also the nltmber of classes. The formula is,

i ~ .,-~L:::-.;:.S __ I, + 3.322 log ,. n

where.

= cl";Ss interval

L, = I.arzcst observation S = Smallest observation

and n = total number of observations in the set, A1"o, the dencmirratcr, I +3.322 log,o" is equal 10 the number of classes.

Q, 10 Describe in brief tIic grouping error.

A= U me c](ISSe$ "'" formed in such a way th at the frequencies are .,venly dl>lributed threughout the dus interval. it is justified to assume that !he freqnenci,es, are centered ru the rnid-vnlllC of the class. But in cases where oucll an assU.DIptil)O i. not valid, it lead. to error which is known as grouping error. Grooping, error affects the accurllC)' of rbe results.

Q: 11 darify the difference between exclusl ve and i""lusi"" class interval s,

Am. Following ure Ibe differences between exclusive and i,,,dusiv~ class intervals:

(i) Ih.exclusive class intervels, the upper limit of 4 class, is the low ... limit of tben ex l c!3S$. Also the, upper I imit of 4 class ;5 not included

in. that. class. -

(ii) In inclusive class interval. the upper limit of a <lass instead i. not the lower limit of the ne.1 cI .... The lower limit is gcnemlly greater by unit m eas un:mcnt.

(iii) In inclusive method. both. the limits of" class ""', included.

(ivl To simplify the calculation procedure, inclusive c:]assc;s. are cOnverted into exclusive classes,

U rhcberrec htl lch gcsc h utztes M ateri a

1.

(v) Inclusive, clD$SC$ approao::h is wi!able in case of data given in whole numbers. In I'tSI of the C8SCli exclusive class approach is suitable.

Q. 12 If mid-values of the classes are known. how can tile classes be formed?

ADS. Find the differen"" between two consecutive mid·values. SublnlCt half of the difference from the mid-value and again add it to the mid. value. The values obtained 00 subtracting and adding half of the di ff.",,,,,,, are the lower and upper limits uf the clan of which the mid- value has been used. If m is !1m mid·v.lue of a class and i i. difference between two ronsecutiv. mid· values, the lower nnd upper

class limits are (m-i) and (m+i) respectively. Q. 13 III ustrate excl us ivc and inclusive class Jotervals.

Ans. In eaclusive class intervals uppers limit uf the, class ls not included, e.g .. io !he class 10-'20 those values are included which are to or more and les. than 20. Similarly in the inclusive class il1!crvals, both the limits of • class are' Jncluded.The cl asses nuy be of the type. 5 - 5.99, 10 - 14.99, etc.

Q. 14 Disti ngui.h between real limits and apparent clw limits of a distribution In grouped data.

ADS;. If the distribution is fOr .. discrete variable, the real and app=nl cJ8S$ limits are same. e.g., S-IO. 11.16. 17.22 •... si!ICC there is no recorded wluc between 10 &. J I, J 6 &. 17. etc.

But If the clU$ intervals are exclusive. in thlll case either the upper limit or the lower limit i.' 10 be, ""cluded since the value can be included in one class only. For instnnce. Jet the cl ... ..,. be 5· 10, 10-15. 1~·20. etc. SUPJlOIC the upper ltmits are excluded. Inlhat case, the real (tmit': are 5·9. 10-14. 15-19. ere, Of course, there, is 110 point in between 9 &. 10. 14 &. I~. etc. If the lower Urnits are excluded. the real limits are 6-10, 11.1~. 16-20, etc ..

Q. 15 What do you understand by tabulati.on1 Ab$. It is the process of presenting dill. collected through survey. experiment o. record in rows and columns SO Ihat it can more easily be unclentood and can be used for further statistical ""aly""

Q. IIi What are different parts of a. standard table?

F'RoGIw.IMEo STATIS11CS

A lIS. There are five paris of 8 table,

(i) TItle. (ii) capuons and stubs. (iii) Body; (iv) PnflllOrial. and (v) Source note.

Q. 17 What are the,objectivcs of tabulation ofdala? ADs. The objecti_ of tabulation are:

(i) 10 clarify the objec1 of illvcstiption. (u) To reduce complexity of data

Oii) T" eecnomise r;pace

(iv) 10 depict the ",Iation among data if lt e:<isU. (v) To raciliLaIC analysis of data.

Q. 18 What are the requisites of a standard table? ADs. Rcquisill:< of a 5lalldard table are:

(il II should be sui14ble f~ the purpose.

(ii) Clarity and comple!eness ofllb1c is necessary. (iii) Table should be of adequate size.

(iv) Units 0.1 measurements should be specified. (v) Logical arrangemeat or items,

(vi) TOIais and sub-totals be gi\'CIl.

Q. 111 What..., Ibe main PIIfJl'05C'3 of tabullllion? Ab$. The main purpose. of L.lbulation are:

(i) To pmeJ!1 the haphwud datil in simple and

ccacised manner. (i i) 1O .. ve spece,

(iii) To show the trend of data, if any. (iv) 10 facilit.a~ comparnon of <iota.

(v) 10 detect errors and omissiOl1ll of data, if any.

(vi) 10 fociJjta.~ the pmccss of statistical<malysis. (vii) 10 luIow lite source of data.

Q. 20 Whlll i. the difference between classification and IabulnLioo?

ADs. Oassification i. meant for an:anging the data ;010 ChtullClCristics or groups wbere each. group has the number of item anachcd to iL In case of variable", it i. given in the form of frequenq diotribution.

Tabulation is the .(og.ica.l and .ys:tematic arrang.· ment of dall in rows and columns. In a table, data may be presented in mooi.fied form. as well .•. g .. in per cent, proponlon, total or average values, etc.

U rheberrcc htl lch gesc h (jutes Iv! aleri"

CLASS!FlCAlIO~. TAIJIJ\JIllON AND FREOUENCY DISTRlBUnON

Q. Z 1 Wh at is an original table (c1assi tiCal ion table)?

ADS.. In an origi n .. 1 table. the data are presented in the same form in which they are collected.

Q. U What is a derivative table?

AIls. to • derivative table the data an: not p",.cmted in lts original form but the vihJeS based 00 original observation. are presented. For instance. totals for different c1assiflCation. s, the means. percentage. ratios orprcporrions.etc .• ore pr=nlcdin A derivative table. Q. 13 Distinguish between frequency and cumu-

lative freq .... ncy. .

All$.. FlTquency: .Number of times D variate val ue is repeated is called lts freq ue ncy.

Cumulariv. !lTqutncy:Thi. i. the Dumber of ob.<ervntion. corresponding to IO$S thao (more U"",)

15

or equal to a speci fied value.

Q.:W Differentiate. between a timo. series and a spatial series.

AILs. (il 11mesclies is an ordered data arranged in sequence of time period. 11me periods may be weekly, monthly. yearly. quinquennially. decadal, etc.Tlme series i. also knoWII as hi.nor/eStrUs . (il) Spatial series i. One. in which the daw are arranged aceording to the place or space, The place or space may be loenlitie •• cities. states, countries. etc. -

Q. 25 Explain briefly the stem and leaf display of data.

A...... II is a me1bodofpr=nlationofdala in which each value: is divided into two pam. One pan consists of one or more leading digits as stem and remaining of tbe di gits as leaf,

SECTION·B Fill In the Blanks

Fill In the suitable word(s) or phrase(s) in the blanks:

I. OM.meatiOD is the of raets that

are distinguished by some sig.nifi.an!

Z. For a good cI ass ification. the class should be

and .

3. Classification can be done according 10

4. Quantitative classification leads to .

5, Ycarwisc ltCOrding of data offoad production

will be called elllS,ifieation.

6. The ce nsus daw published for .ityw'se pcpu-

lation in India will be known as _

classification.

7. The data recorded accooIing to sWldIld of educntion like illiterate. primary, secondary. gradunte. II:Chnical. etc .• will be known as ct ... ification,

8. Distribution of '"",ilies acwroing to their size

will be classified as elassifleaiion,

1). The difference between the upper and lower

limit of a class is called .

10. The average of the upper and lower limit. of

a c:lil$$ is knowii as ,

n. There.is a gcneraIassumpl.ion !hat the c I ass

frequency is centered at -the of

the class.

lZ. Departure from the assumption that the frequenc ies are exenly distributed 0"" lite

class inltrva1lead. to error.

13. Formula Ior determining the number of

classes ....... given by ~

14. H.A. Slurges fannul. for determining the

number of classes is •

15. Number of classes depend on .

16. HA. Siurges formula for fl.llding out the class

interval i. .

17. The. number of cl asses and class interval for the distribution of marks from 0 to 100 of SO

students of a cl ... should be and

____ respectively.

U rheberrec htl lch ge5c h utztes Iv! ale,i a

16

____ limits.

19. Mid- val ues of the ~Iasse:s ilIl= Ill... called

20. Freq""""y density of a cl ass is the frequency ____ of class.

21. In lite mid-value of" class inlCrYlll ;. 20 and the difference bet w een two consecu1ive mid-

values in 5, !he class limits are _

and ' _"_'.

'22. All anung,emefI.t of data in rows and columns

is k.nown, <IS ~

23. Tables help in __ ,, , of data.

24. Tabulation 1Mkes, the data easily ~

l5. Tnbulation follows .

u;. facts <oW10l be presented in .the

form of • table.

rT. A ~ purpose table i •• '-'0 known as

or tIIbIe.

28. A senernl, table is a of a large

amoitnl of data.

29. The table w,hid, do net present the data bul

I,," {!>SullS of analysis ""', called _' _

tables,

30. Hendings of the columns of a table IUe known

os .

31. Headings of the rows of a !able 'are called

J2. The portioo of !be IDbIe in whicb data ...,

presented is, desi£Oaood ... of a

table.

JJ. Tbe manner in which the frequencies are distributed :u:cordiQg 10 variate values is

known as .

34. Frequency d;'aibutioos are often constnlCtcd

wi.th Ibe help of .

JS. RelAtive frequency is the ratio of a frequency

10 !he of the di.tnbullon.

36. Pen:eoIage fiequency is Ibe multi-

plied by 100.

:n~ Fn:quencies lidded suceessivdy in an, ordcn:d series giving !he, num~ ofill:ll)S up 10 thai

value are ,,"'led . •

38. A frequency distribution willi upper lim"s~ o.f c10s0ea aDd COrrespo,ndiDg cumulative freqlliOllCi~ is kIlO"''' as type distribution.

39. A fmJueocy diSttibution wid!, klwer limits of el3S:O:S aDd correspo,ndiDa: cWDUlative fre-

quencies is known as Iype disIri-

bution.

40. The graphs of loss than type and """" dian

type di!lributiolUl intersect II ~

41. A grouped series, iowhich eilber !be lower limit of the first group or the upper limit of the lIS! group;' missi,,! or boIh. ls called UI

42. A series anunged. in lICCQ!daPce with eDclI

and every observation i. known as .

43. The dislribution of frequeucies. aa:onl..iDg to

iodividunt van"",· willeS is called _

disai.buti.on.

44. A series of data wilb exel ... lve classes o1ong with !he com:sponding frequencies is called dislllbuuon.

<IS. Given the followjng Ii:1:quency distribution, fill jp the missiQg freq .... i.::ies'

S IS

Cltw F"'queru:y

inttrir'tiI3'

10-20 ,

20-30 10

30-40 12

40-'0

50-60 7

60-70 4 Cumulatior F ... quency

U rheberrcc htl lch gesc h utztes M ateri"

ClASSIF1CATION. TA81JlAlIOH ""0 FREOUENCY OIS1RlBUT1ON

17

Multiple Choice Questions

SecTION-C

Select the COIT8d a//emaliv9 ·out of givBn ones;

Q. 1 Numerioal data presented in descriptive fonn ore coiled

(a) classified presentation (b) tabular pre.~tatiO!'!

(10 )grapI1ioal presentation (d) Ie.tun] present:>tion

Q. 2 Whether classilication is done, lim or tabuIalian?

(a) Classification follows uibulation. (b) Claso:ifica!ion precedes tabulation. (e) Bolb are done .i.multan.eously ..

(d) No criterion.

Q. 3 Fot the rmd-valuesgiven below, ~, 34, 43,53, 61. 70

The fin.1 class of the distribution is: (a) 24.5-!W.5

(b) 2S-34

(e) 20·30

(d) 20.5~29.5

Q.4 In an exclusive type disuiOOtion. the limits ex.luded are:

(a) lower limits

(b) upper limits

(e) either of the lower or upper limit (d) lower limit and upper limi ts both

Q. 5' A $erl.uhowing lbcseLo or all distinct values individually with their frequeacies is known 'os:

(a) grouped freque""y dislribulion fb) simple frequency di.a:ibution

(e) cumulatiw frequency di,trlbuiioo (d) none of (he above

Q. 6 A series $howi ng the sets of all value, in cl asses with their =pon!Iing frequencies is :known as:

(a) grouped rrequency distribution (b) "imp-Ie frequ.""Y di.trlbutio,n

(e) cumulative frequ.""Y dlstribetlon (d) non. of the above

Q. 7 If the number of students in a schocl is ZOO and maxlmum and minimum marks corned a", 90.00 IOrespeclivcly, for the distn"bution nfmarks, the. class interval. (rounded) is:

(a) JO

(b) 9

(0) 12

(d) none of above

(Given 108,;: = 0.3010. log,.3 .. 0.47711

Q. 8 If the Jower and upper limilll of. class are 10 and 40 respectively, the mid-points of the cl"" is:

(.) 25.0

(b) 125

(c) 15.0

(d) 30.0

Q. II In a grouped data. the number of classes preferred are:

(a) min_imum posoible (b) adequ.",

(e) m .. Jmum possible

(d) any ..rtHtr.lriljl chosen number,

Q. 10 0_ interval is measured as:

(a) The sum of the upper and lower limit (b) half Qf the sum of lower and upper limit ( c) half of the difference berween upper ;!!ld

lo we rlimit

(d) the dille",nce between upper and tower limit.

Q. 11 ibc, cl ... ;n"'tval of the continuous grouped data:

10·19 20·29 30-39 40-49 50-59

is:

(a) 9 (bl 10 (c) 145 (d) 45

Q. 12- A grouped frequency distrlbuuon with uncertain fim or lao;t classes is known as:

U rhcberrec htl lch gesc h (jutes M ateri a

18

(Q) exclusi ve class distribution (b) inclusive class disuibution (c) open end diSlribution

(d) discrete frequency dl!lUibutiOll

Q. 13 The distribution,

\.\ll,,~t [,eM than 5 Less than 10 Less than IS Less than 20 Less than 2S less than 30 is of the type: (a) inclusive class type

(b) exclusive class type

(e) di=te type

(d) none of the above

Q. 14 Data can be well displayed or presented by way or:

(a) stem and leaf display (b) cross c1assirlcalioR

(c) two or more dime!1~oruol !able (d) all the above

Q. IS A simple table represents:

F",qullCY S

l2

21

27

31

33

(a) only one factor or variable

(b) atwa}'$ two fa.;1or!l or variable<

(e) two or more number of (actors or variables

(d) a111he above

Q. 16 A complex table represents: (a) only one factor or variable

(b) alwa)'$ two fa.;tors or variables (e) two Or mote facton or variables (d) all the above

Q. 17 The headings of the rows given in the first column of • table are, called,

(Il) stubs

(b) captions

(e) titles

(d) prefnro<y notes.

Q.18 The column hending of. table are known as: (n) sub-titles

(b) stubs

(e,) reference notes (d) captions

Q. I' The series.

Pltu:e

No. of (JCcidJlnlS

per day 10 IS 18 17

7

Delhi Kolk.ta MumbAi Chennai 1rnIo",

is of the IY.I'"' (a) Spatial

(b) Geographical (e) Industrial

(d) Time series

Q. 10 The series.

Year

ProdllCtiOll of [ood

(1.1. ToRMS) 160

168

170 ]72

174

176

1987 1988 1989 1990 1991 1992

is oalqoriJed q:

(.) individual",ries (b) continUOll.! sericl (e) discrete series (d) timescries

Q. Zl The inco.me of five persons is as followt:

P."on Incom~

(&/Month)

,MrA 1,700

MrS 2.300

MrC 1.000

MrD 8,500

MrE 5,400 The .bove series is of the 1ypC: (a) Individual series

(b) Discrete senes

(e) Con!inlU)U$scries (d) TIme scrios

U rheberrcc htl lch gesc h (jutes M aleri"

ClASSIFICATION. TA8UlATIQN ANtI FREQIIEWCY DlSTRlBlfTlON

Q. 22 The series,

Marks 10·30

30-40

40-50 50-60

60-70

10-80

is of Ibe type:

(a) Discrete series (b) Conlinu"us series (e) Jndividuul series '(d) none of the above

Q.·23 A frequency distribution can be: (a) disc",re

(b) conn nDOUS

(e) both (a) and (b) (d) none of (a) and (b)

No. 01 SIUr./.nIS S

14

24

J2 9 2

Q.24. In an individUAl series. eaclI variate value: (a) h .. s sa me frequency

(b) hIlS frequency one

(c) has varied frequency

(d) has frequency two

Q. 15 Which of the following s ta tements is true? (a) An individual series is a particular case of discrete series

(b) An i nd ividual ser les is " panicula. c ase or continuous series

(c) An indjvidual series is a special c ase of discrete, and continuous series

(d) There ts noihmg lilre individual seri .. s

Q. 26 Frequency of " variable is al WIly.: (a) in percentage

(b) a f,action

(e) an integer

(d) none of the above

Q.27 The data given as, 5.7. J 2. 11.79.84.91 will be caned 8$

(a) • continuous series (b) a discrete series

(e) an individual series (d) time series

Q. 18 The following f.equency di,lribution,

19

x: 12. 17. 24, 36, 45. 48.

1 2, 5. 3, 8. 9, 6, isch"'ified as:

(a) continuous distribution. (b) discrete disbibotion.

(el cumulative frequency distribution (d) nOlle of the above

Q. 29 In an ordered series. the data ere (.) in ascending order

(b) in descending order

(e) either (0) o. (b)

(d) nejther (a) 0. (b)

Q.30 The following frequency distribution,

52

Clan •• (}'IO 1).:20 1).10 (}'40 1).50

Fr.quency 3

8

14

20

25

is known 3.$:

(II) continuous frequency distribuuon (b) discrete frequency dlstribatien

(e) cumulative dist.ibution in more Wan type (d) cumulative distribution in less than type

Q.31 The following frequency disll'ibulion,

ClIlS ses 1).15 0-10 I).~

is clas,ified as:

(3) cumulative distribution in less than type (h) cu mul ati ve di.tri burien in more than

F,.qu.,ncy 11

8

3

type

(e) discrete frequency distribution (d) cumulative frequ.ocy distribution

Q. 32 ClassifiCAtion is applicable in case of: (a) qua. htative characters

(b) qualitative charact ers

(e) buth (a) .nd (b)

(d) none of Ibe above

U rhcberrec htl lch gesc h utztes M ateri"

20

ANSWERS

Section-B

(.1) grouping: characteristics (attributes) (2) exhaustive: mutually exclusive (.3) anrlbetes (4) frequency diotribution (5) Chronological (6) gcographical (7) qualitative (8) qoantitlti>'C (9) cl= mtcrval (10) mid-value (!I) mid-value (12) grouping (13) HA StuQl.s (14) I + B22 log," n (15) CI'5, jnterval (16) (L - S) 1(1 + 3.nz log,. 0) wI1ere L = l.rge.t •• lue, S = smallest value and n. = No. of values (17)7 and 15 (.18) mathematical (19) class marks (20) per unit (21) 17.5 and 22.5 (22) llIbulatiou (23) ""oIY.I;& (24) uoo=and:ible (lS) classification (26) qualitative (27) primary or reference (28) n:p<lSito.ry (2.9) 'pl!Cific p~ (30) captions (31) stubs (32) body (33) frequency distributiun (34) tally marks (35) total frequency (36) relative fr"'luen.cy (37) cumulative freqUM<:y (3S) I.,.. I""" (39) more !han (40) median (41) open end series (42) individual series (43) discrete frequency (<W) continuous frequency (45) freq = 14; as, oU-{<I.'<I·27. 41, 48.

SECTI0N-C (I) d (2) b (7) b .(8) •

(3) d (4) c (5) b (6) • (9) b (10) d (II) b (12.)"

(13) h (14) d (IS) a (16) c (17) a (lS) d (19) b (20) d (21). (22) b (2;1) c (24) b (25) 3 (26) c (27) c (28) b (29) c (30) d (31) b (32) c

Suggested Reading

1. Agarwal, a .L... Basic S,mmies, New 1\8" In,emationol (P) Ltd. Publish ers, New Delhi. 3n1 edn.,I99~.

2. Devore, B •.• Probability and SIIltLslifO. for &'ginurinll (lJId the Sciences, BrooksICcle Publishing Co" CnIiromi.o., 1982.

J. GUplll CoB.. An Imrodur:tion 10 Stalisticai M.lhotIs, Vikas P\!bbshing Honse, Delhi, Slh OOn .• 1978.

4. Kenny. I.F. and Keeping. E.S .. Masbnnarics oj Statislf"fl, Part I, O. Von Nos!r.lod Co., NewYork, 1951.

5. Sancilcti, D.C, and ~" V.K.. Statit/ici.

Sulton Chand de Sons, New Delhi •. 7th edn., 1991.

6. Shukla. M.C. <01d Gulshan, 5.S., Srar .. ties, Sultlln Cbnnd &. Co .. New Delhi. 1970.

7. S;n1p5On, G. and JC.nIbo, F" Basic Stotinicl.

Oxford & mH, Cal<:u'ta. 3rd 100.i3O. print, J971.

U rheberrcc htl lch gesc h utztes M ateri a

Chapter 3

'Diagramatic and Graphical Representation

SEcnON-A

Short Essay Type Questions

'Q. I Wh.l are the advantages of diagramaric representation of dala?

Ans. Pollowing are the advantages of diagramati; roproserU81ion of data:

(i) Di~ give. bird's-eye view of complex

diu ...

(iiI They havelong lasling impression,

(iii) Easy to undemand even by a ccmmon man, (iv) !bqsavc time.and labour,

{vI They 'f!lcilltmc comparison.

Q. 2 Give the names of di,agrams which are onedimensional.

Am. 'Bar diagroms and line diagmmo are onedimensional. Different lypes of bar diagram, are:

(il simpl"bar<f~gram (il) mulliple bar diagmm

(iii) .ub-divided or component bar diagram (iv) peroe!IlIIj,"" bar diagram,

Q.;\ Name diagrams which, are two-dimensional. AIlS. Rectangle», circles 3IId pi. dingram. are twodimensional diagrams.

Q. 4 Wltlll ,ore tbe di'lgrams categorised, under threedimensional diagrams?

AIlS. Cu 00<, cy 1 i nders and spheres an: categorised II.~ three-dimensional dillgrruns.

Q. 5 What types of diagram an: known as nondimensional diagrnrns?

ADS. PiclOgmm, are known as non-dimensional diagrams.

Q. 6 WIt.t do you understand by bar diagram and 8 sub-divided bOT diagram?

AI1$ A bar dlagram represents the magnitude or a single factor acco.nliQglo time periods, places. items. eic, But when the magnitude of the factor i. given with its sub- factors, each bar is funher sub-divided into componems in proportion tnthe magnitude of the sub-factors,

Vearwlse Sal ..

S.100 (e",," A"I

::.~

; 19Q2

1991 1990

s 4

fig. 3.1. 50, tUgrwll

U rheberrcc htl lch gesc h utztes M aleri a

22

Such a diagram is known as sub-diviced bot diagram,

CounlrywlS" Tou(lSll" various C~ies

No.. 01 Tour1st

Agoa DOlh, Mad'r8.s Ujjaipur BBngaJor-e _.,...

SriMg.r

o

Fig. 3.2. Sub-<li.lde<l bot dleg"''''

Q. 7 When do you prefer am" ltiple bar d lagram (compound bar d iagrnm j.

A.... lb depict a number of rei eted factors for comparison in vanous years or at :1 number of p lace-s. mulli pi e bar di egrams are p re ferab le.

Q. II Who.1 is" multiple bar diagram?

A.... In a multiple bar diagram, adjoining bars nrc drawn according to the n urn ber of factors and their heights in proportion 10 the velues of Ihe factors in the same order for each period or 1'1 aee, Each bar of "group isshown by differ e nt patterns or colours 10 make them easil y disti nguishable and this pauern is retained in all. the groups. A constant diSWlCC is

.., ""1---------------,

.--

I i

, r I

~ I

I

j

I

I 10

o

Fig. 3.3. Multiple bar diagram

PR.OGFWdMEO sr ATI sncs

maintained bel ween gTO" ps of bars drown for periods or places. Such a d iagram is known as m ultiple Or

compound bar diagram (Fig. 3.3). ~

Q. 9 What do you understand by devlatl on bar diagram?

A.... Devi.ti on bar diagrams are su ltable 10 show the net dcvi.u oIlS d uring various years Or according to different countries or places, etc. In the deviation bar diagrams, posi live, deviation, are shown (0 (he right sid" of the bas" line and "egotive deviauons an: shown 10 the len side of the same bose line, For instance, tile S"P' between i m pons and e .. pens, profit and loss in di Iferen I. years or from different countries can be very we 11 displayed through de viatien bar diagrams (Fig. 3.4).

Net R6.5UrL!i· of a Company

19a6

1001

1990

o

1991

Fig. 3.4. O""I.lion bar di.yr''''

Q. 10 Wnte n short note On duo-direcuonal bar diagram s,

An$. Duo-directi onal bar diagrams are used to exblbit'the two aspects of a single. faetor at a glance gi ve n for d lfferent periods or places. In this type 0 f di cgram ~ one pan of tbe bor remains above the base li ne and the other below the base I ine. The hei ghts of tho bars below and above the I i ne ore in proportion to the valu es of the two aspect scp""'tely wh erea s the bar as a whole represents the f actor, For exam ple, we wanl to show the price of cenaln .. item in differenl years, Thc price cons lsts of tWO parts, (he cost and

U rhcbcrrec h II ich ges" h utztes ~ t atcrla

DIAGRAMAT'C AND GFW'H'CAI. REPRESE:NT"TI(lN

'he prof! '- So pro fI' may be ta ken above the line and cos, be low the line. It i, a son o r sub-d ivided bar diagram (Fig. 3.5).

Pric. 01 110m.

30

20 I-

.;

a: ,0 I- m

~

~

\l. I

~ 10 I-

...

~ 20

SO Fig. 3.5. Duo-dlrecUonai bar dlog"""

Q. 11 \II rile ~ brief note on paired bar d iagram, Ans. When two related factnrs h"vin g di fferent units of measurements are to be displayed for comparison in van OIlS periods or places, paired bar diagrams are suitable. In this di"llram u,ually, lhe periods or pi aces arc ,hown in a strip and hori zontal bars for each factor on: drawn 10 the right and left of the vertical strip or vertical bars ore drawn below and above the horizontal strip, For instance, area rmd pmduetiun of pnddy in differeru years in India can be very we II d isplayed through a paired bar dl cgram (Fig. 3.6).

Q. 12 Whnl Ott sliding bar charts? Explain in brief. A ns, SI id i ng chart is a bi lalora! ch an in wh ich two componenrs of a factor are represented by IWO parts of {he bar. One p art i, 0" {he left and the ether is on the right of the bas. line. The scale may be the absolute "umbers or in percentages. Sud, a chart is suitable i nsituati on, such as a n umbers arrested in" criminal cole or patients operated for different d iseases, What percen t"8e or number of suspects have been cleared and who' "'" still under tria]. How many patients ore cured und how "",oy of them sti II

23

Area and. fJ",roWctbn 01 Paddy

t sea-a!

19711-71

tl)6O.El

111511-51

eo 60 40 20 0 20 40 60 eo

Fig. 3.8. PB.irfid ,bar d[sgram

su ffer, The two com pone" IS can be beuer displ ayed by , sliding bar d ingram. Tho base li ne may be ",p"',e"!ing the type of opentlon" kind of court cases. etc, 'I1Ie, pen:entagc or numbers in the two components. cured and not cured, cas", cleared and not cleared may be changing, from lime 10 time. For each type of operatlon crime (faclOr). a separate ,I id ing bar wi 11 be drawn.

PalfEI nts Oparatad

N'ot cured Cured

95 70

20 0 20

70 95

Fig. 9.7. Slldng bar chM Q. 13 Wh.1 is a broken bar di agram?

An.s.. Often an in vesti gator comes aCtO_SS cases where SOme figures are very' large as compnrcd to

U rheberrec htl ich gesc h (jIZI"S M aterla

OIbers. In Ihls siluarioo, if the seale is chosen for proper portray of .m.n value. by bars, lbe b .... for large values will .'I"'nd 10 a unpalatable .i,.e. Agnin, if the ..:.1" ill chosen for proper di.pla.yof large values by ban, lbe bars for $lllllll. val oe will become noo-ex.illcnl. Hence. to remove Ihi. discrepancy. broken han; "re COOltrucled.

Fi "I, ~small scale i. taken and bars on: erected III all period. or placee up 10 the higb.e.<ll small values andlor a round off · .... luc. Then willi a gap anolller base lin. and a new scale for I.rge values is chosen. Ban Are constructed for remaining val ... on !he new ba!!e line. Soc" • bar diagr.tm is knowo as bmun bar diagram. These diagrams be inlerpreted very carefully 10 avoid any wrong conclusions.

170 160

n

Un

100

se

A,ts Com Law

FIg. U. BtoI<8IIba' (ia,g<am

Q. 14 Who' is the b .. i, of comparison in bar charts?

AlJS.The basi' of comparison in bar chltU is Iio"",, or 0.00 dire.:tional.

Q. 15 Whalls the difference heiwcc:n ber d~,JfilIlIs (cbans) IUId colUmn chllrls (diagrams)?

A... In the bar diagram., Ih e bars are arranged horiwnlally on a vemcal basi' line, whereas in the column chon&, b:u: ore arranged venically On a hori~1al base lin e,

Q. U. Whal Ire Ihe differences between • bar diagram and I recIaI!gIIl..- diagram?

PROORAMMED STA1ISTlCS

ADS. The dirr.",n",,'S are:

(il In bar diagmm all b ars are of equal width. (ii) The widlb of ban in b:u: diagmm Is chosen arbilrarily for qualitative characters lik e places, stales. coun1ria, etc .. or accordin,g 10 the equal class interval s,

(iii) In a reclangutar dl.lgram, there is always. vnriahl e andus magrulUde. Th e width of the rectangles"", eccordmg to variate valu es and hcigm, of the rectaDglcs an: IICcordi!l~ to \heir respective magmrud es,

(Iv) In bar diagram. comparisons nrc based. on the heights of bars only whereas in rectangular diogram., comparisons are based on the orca of the rectangles.

Q. 17 E~plaln ~ line diagram In two lines.

Arw. A line diogr.tm i. a one-dlmensinnal diavam in which the beigh! of the lin e repr ese nts Ih e freqUCIlCY corresponding 10 lbe value of the item or a !'actor.

35 30

10

I

0 s .. ~ it .i! s

.. ~

::E ~ '" FIg. 3,g. Una diagram Q. 18 Expl.i.n in brief a. pie-chan,

ADS. A pie-chon is • .in::uw diagram whkh is usually used for depicting tbe components of. single factor. The circle is divided lnla se grnen ts which are in proportion to lbe size of the components. They are saown by different pauerns or colou .. 10 make them OIlmctive (Fig. 3.10).

U rheberrcc htlich gesc h utztes M ate,i"

OIAGAAMAnc ~D GRAPHICAl. REPRES£NTA,T!ON

India'. Impom. from various oourca. (per com) 1985.J!8

FIg. 3.10. Pie cheri Q. 19 Explain a hislognm.

Amo. A hislogram is a bar d i.gram which ls suitable for frequency dislributlon, with cominunus classes. The width of ~II bars is equal to class interval and h"igj,ts of the bars are in proportion 10 the frequendes of the respective classes. In this diagram ban touch each other but, One bar never overlaps the other (Fig. 3.1 I).

O!sullUtlon of Wooos

250 - 900

s

~ 0200

~

.;

2'; 100

In:g8g8ciH!!;i e ~ g ~ ~, a ~ <CO Cb ~

W.geoiMonlh (R a. )

fig. 3"n. HllOgoam

25

Q. 20 Describe a frequency polygon.

A!IS. When, the mid-points of the tops of the adja· cent bars of a histogram rue joined in order, then the graph of lines so obtained is called a frequency polygon,

F",quoncy DI."lbu~on 01 Wag ••

,350 300

~ ~ ~'~

Wa,gooIMorith (A .. 'I

Fig. 3.12. FfI'qu...,y polygon

Q. 21 Discuss a frequency curve in bri,ef.

An$. A frequency curve is a graphical represeatanon of frequencies corresponding to their verlate values by • smooth curve, A smoothened frequency polygon represents a frequency curve.

Dls1rbJtlon 01 III a""

Fig. 3.13. Frequ_y curve

Q. 22 What is • false line in reference to a graph? An" When the variation in the magnitudes of a variable is smatt as compared 10 !he variate, "IU0:.5,

U rhcberrec htl lch gesc h (jutes M ateri a

the yenieal ,scale ebusen as zero at the origin fail, [0 depict the fluctuation prominently as desired by the in"""tigator. lience a line parallel to absciS$3 (Xuis) is drawn a little above it and the point joining the ordinate (Y-ub) is [:u;en as origin which u~uolly represenU tbe minimum value (an approximate value by dis.:ardlng figure) from amongst the magnitudes to be liken on [he, vertic • .! seete, Then proper ..,.,1. is chosen I!IelWIring distances from this foIse base line, Such a process makes the fluctuations conspicuous. The grapll $0 obtained i$ called gee wltk gruph. False line on the graph is usuany shown by ~ SAWtooth line.

Q. 13 Whot is the range curve (chart)?

AM- To depict !he spread of highest and lowest values of a hllle ,series, the band fanned by the une Rraphs of highest and 10we51 values is known us ra",~ curve. Range curve Can also be represented through bar <egmenu drawn a' cac:b periodwhicb arc of magnitudes of the differences between higllest and lowest vallOC$. These bws ,ran from the lowe't price and go up 10 the high""t ""Iue.

0. 14 How do you d,aw a bistQirBlIl when the widths of all classes ,are not equal?

AtIS. When widths of all classes of a frequeney distribution are not equal, beighl$ of the ban are taken in, prcporuon to the frequency density (frequency perunlt interval).

Q_ 15 What, is ~ ratiD ch.an. or a seml-lognrithmic graph?

AOL It i. a line graph obtained by plouillg lbe poinlS (>:-. log y) in such a way thai the ... -vaIue, are taken oIong X-axis on n naturlll scale .nd values log )' oro lake n oIong Y-uis. Hence, rntio ch.an. is a graph of the points (>:-. log y) ploued on an ordinary graph paper: II is also called semi;logarithmic graph in the $Cnse that log-values are used only for y (Fig. 3.14).

0. l6. What do yau undcntand by a graph?

Am. A graph is n display of points and line s, In n ,,,,ph of paired value. t.r. y), so called the eoordinlles of a point. are ploued on a graphp:>pcr by ,uitably choosing Ibe scales along X-.~.is (.m.ci_.s.) and Y-ws (ordina[c). 'The planed points are joined

440 430 ~ 420

i (10

~

82

83

84 Yur

85

88

Fto. 3.14. RallO 0lWl

by straight Ii nes in thelr sequeece of occurrenceThe figllR! so obtaioed is coIl.cd a &f8PII. 1lte &rapb depicts the !rend, flucruaUon., VlII'iability. etc., very prominently (Fig. 3.1.~).

Two II{ mO", graphs mlUin on the same graph paper bI~ing a common scale .IOII! X -nxis and Y-axis facilitate the comparison of data tremendously.

Pop"ldon aIId A_bUy 01 PulHs

50

PopuIalkln (C'.'.')

FIII_ 3.150 Graph

Q. l'7 Describe an ogive clU"'IC in brief,

A..... It I. graph plotted for tbe vori.te values and their cOJl:Csponding cumulative frequencies of a freqllClley distribution. Its shape i'jus!' like clong.ted

U rheberrcc htl lch gcsc h (jutes M ate,i a

S. An ogive curve is prepared eith .... for more tluin type or les. than type distribution.

M

40

i 30

!I

: 20

~ ! 10

:z:

10

30 <10

20

FIg. 3. tl. 0gI'va curve

Q. 28 WMt arc !he uses of oglve curve?

AIlS. Ogive curve i .• useful in finding out quarules, deciles, percentiles, etc.

Q. 29 AI what point the ogives for more than tyP" and I"". than type dlsuthution intersect?

A.... The ogives for more thon type lll!d le .. than type d istributions intersect at the med~.

Q. 30 Explain briefly a Lorenz curve,

A.... Lorenz Curve is a ,pccialtype of cumulative frequency curve which i. used to por:ll1ly the oow 10 ,ndi calc whether a facter is equally dislributed in re Ialion 10 the other raClor fcr eertain segmeru of !he population. It was origilUllly developed by M.O. Lorenz and is named aft er him.

Q. Jl How COQ one draw a Lorenz curve?

A.... Following ,steps are i",,,,lved while drawing a Lorenz curve:

I. The vuriate values giving infwmation lIhout the segment of population are ignored.

2. Cumulative tolnl' for the mll,gniludes or frequencie, of the two other lilctors lire found nul'separately.

3. Cumulative totals ,are ._pressed .. percentage of their respective grand 101:11,.

27

4, Paired cumulative pemmtageo areploucd on • graph paper choosing'ame scale along 0<.1 frum 0 to 100.

S, P101ted pointS me joined by a smooth free hand curve, Thi. curve alway. stan. from the origin and tenninat<.la! the end point (100. 100) (Fig. 3.17).

IOOr-------------------~

eo

60

o

20

eo

eo

lOt!

40

Perc,nl, ot Plreonl

Fig. 3.17. Ltnr\leurw

Q. J2 Wh., i. the importance, of the Lorenz cune1 AIlS. In case of equal distribution Qfboth the f!ICtOJ'S in the segment of pepuletien, the graph will be • 'traight .1 ine. Bul. the Lorenz curve forth... fmm straight line show. the inequality of dislribution of two foetor s,

Q. 33 How are Ih. data portrayed by pi etograms? .Ans. In pietegrarns, !he data are di splayed by th e pictures of the items to which the 00111 pensln, A .ingle pietou'" represents a fi_ed number. For example, th.e pOpulation ;, shown by man, milkproduction by milk cans, neelS of aeroplan. by the piClu"" of ae roplanes, etc,

Population of Indi. in 1993 i. 89.6 croee, This can ·00 shown by a pictogram having tit. piclure. of nine men. each man repn:.<entins. ,10 erorc. Pictegrnms are lb. least wisfaetory type of dingrnms. They are inaccurate 100. Even "'en they are preferred by novice and dilenante people.. OisplJ!y of data through piclojlnUt1. was inili.atcd by Dr. Otto Neun.th in 1923 (Fig. 3.18 J.

U rheberrcc htl lch gesc h utztes M ateri"

28

PcpulaUon 01 I ndiB. in C'8n~U.!l YeranI

Fig. 3.18. Plctog"",,.

Q. 34 WIt.1 are col umn charts with circular bose line?

A ns. It is a sort of bar diagram cneirculer base. Sue h column churn ore very attractive and appealing 1,0 display monthly da", of" year Or hourly data of a day, ete. In this diagram, the periods aretaken on a circle just like the dial of clock and. eqllidisLI\nl concentric circles represent certain frequency. Far i nstanee, let the inner circle represent a frequency

No. 01 Ac:eJdfi nb,P8 r Hour

Nighl A

equal 10 10, the second c in: le would rep resent 20 and so on.

Columns (ban) ore drawn On each period of the circular dial where... the length of the column. measured through co ncenmc eire los. is equal to lite frequency of that period. Such a chart is known as column chan on circular base (Fig. 3,19).

Q. 35 Describe briefly a step bar or column chan, AIlS. It i. n techni que forportraying componen IS divisiOn> of on agg,..gn", Or components percentages. In step b!ir chan. item names are taken .long vertical base line and bars !ire taken horitonl..a.Hy 011 ;)Q arithmetic scale, In n step column chan. item name, are taken along n horizontal base lin 0 and bars nne taken vertically On an arithmetic sent". In thi type of chan s, the next bar or column starts where prevlous bar or column ends. SI<:p bar or column charts nrc used as n sebsunne of pi.,.chnns, . al thou gh, these charts nne not as appealing as the, pie-c harts.

EmplCyrn.nlln Publil: S""'o,

CI 25 00 75 100 12S 1:50 '75:200

(LaI<h.)

CenHal (lOYL SI.leGOY!. Owt;sIGovl. local Bc(H8S ,

CO"-!lrut:1lOn Manulactunng SOO1I'd1y , Ml'lmg,

~------------~

Fig. 3.20. Sl~ bar chart

Q. 36 What are overtopping bar or column charts? Am. Tbese are digression designed multiple bar or column charts, In this Iype of charts, one column (bar) penetrates into the neXI column (bar) by half of us width. Bach column or bar i, shown by d i ffereot panerns or colo" rs, Of course. such charts save SJ>B'" as the spread of group of column, Or bar. is ,ize.bly reduced (Fig, 3.21).

U rheberrcc htl ieh gesc h utztes M atcrla

DW)~TIC AND GRAPfIICAL REPRESENTATION

Allinclia ProcIuctlCn (M.ToM"l

o Alco

• Wheal

• Puis ..

80

40

o

19l!O-ll1 I 9l!~6 1990-91

FIll. 3.2.1. Overlapping bar chari Q. 31 Wrile sixl;n es a bout tri1inenr chan.

All$. The lriliooar chans enables 10 display three variables simultaneously in. lite form of compoaen ts or elemen IS of a qu;mtity as a whole. It i. a 100 pel cent OhM" i.e., !he sum of the lit"", components is always equal 10 lOOper cent. The trilinear char! is drawn in lit. form of a equilater.ll triangle by .. ,Ubcnting each side in pen:entage divisions ranging from 0 to 100. The lines are drawn parallel to tho sides of the rriangle in 1he manner of co·ordinalCS.

Q.38 How do yOI! deftne. cltut,diagram orsrnph? ADS. It is technique, for portrayin8, numerical rAe'" through do,",-lines. =s. volumes. 0Ibtr sc"rn<:tr:lcal fOrmll and symbols .

Q. 39 Wh.t rue 1ho advantages of clwu. dillgnlms oed ampb,?

ADS. Charts. diagrams and graph. facilitate the prcsenwioQ of quantitative values in a simple and easily understnnclllb!e form. They fadliUlu, comp'lrison 3nd analysis or data. Graphs in patticul4r !how f1uctuation s, trends and relmionsllip. between lite variables. In comparison to fables and tl:!Ilual forms, charts IIfId graphs are beuer and cas ;Iy understood by a eommoner,

Q. 40 What factors an: responsible for the choice of a chart. di.gram ond gr"Ph?

AI!$. The main (actors responsible for the choice

of a chart or gruph are: ~

(I) Objecllve of !he display of data. (2) "fYpc of dall1.

(3) Sill: of chan or graph.

(4) Audience for wbom a chan or' graph is prepared .

(5) Funds availnbl.c,

(6) Predilection of tbe sc.icnlilt

SECTION-B Fill In the Blanks

FDI in tile suitable won:1(s) or phrase(s) in Ihe blanks;

1_ Diagrams "'" 4n01her form of _

2. It is comparatively easier to understand ____ lItan nllm.cricnlligll res.

3.. TIle osi ves of less lItan type and mote than

Iype distributions .In_I-III _

4. Histogram and historigwn are _

s. In a histogram bars e""h oilier.

6. A histograrn i~ for geographical classificaucn,

7. Frequency polygon con be drawn with. 1hc

help af _

8. diagram con be sketched (or

gwgraphical serie s.

II. Sub.divided bar disaram depicts the distri-

bution of of a (""tor.

lao Paired bar diagram. are ~uitable for lite data

of two related foclOffl having _

units or measurements.

11. bar di.zram is suitable fo.

.howing the dlffcrcnces between bud •• t

U rhcberrec htl lch gesc h utztes M ateri"

30

provisions and actual elpendilurc of PWD in !he lost ten years.

12.. Histograms can be drawn only for ____ diSlribulions.

13. Pie-chart i$ always ~

14. SqU3l"<'S are dimemional

diagrams.

IS. Cylinders are dimensional

diag)"Oms.

16. Pietograms an: dimeRSional

diagranu.

17. A straight line in a graph indicates the

18. A zjg·zag grapb shows the ofa

series.

]9. The heights of bars with unequal cl ass

i .. terval. are proportional to _

10. In reclangular drogrlllll5, cOITlporison is based

On of the rectangles.

11. Line diagram is suitable when there are __ . __ variate values in rhc freq ..... cy distributions.

n Multiple b er diagram has " of

bars for each YeM or place.

23. WIlen more than one factcr i.s 10 be displayed

for cemparison during vari"",, years, _

PRoGRAMMED STAllsncs

bar diagram is suitable,

24. l\vo related facto .. having different unil.!l of measurements can' be displayed suitably by ____ diagram.

25. A graph of variOlll values and cOrte$pOndin,g

frequencies is known as _

Z6. A smoothened frequency polygon is koown

"'----

rT. The srapb of the points (x. log y) is known as

2B. Ratio chans are also known"" _

ch.arU.

19. Lorenz curve WIl'I initially given by .

30. Lorenz curve indicates ill. _

of distribution s of lwo factors of Q population,

31. When Lorenz curve turns OUllO be • ,!might line, it its concluded thai the two factors

have dislributioQ,

3l. More the distance of Lorenz curve from the line of equal dislribution. more is the ____ in rhc distributions oftwo (aClO" _

33: PiClograJIIS ""' satisfactory.

34. Pic!Ograms were originated by _

35. Column charts on circular base arc jnsllike

3Ii. Them i. no zero base line in a semi-loge-

ri1llm~ graph since _

Multiple Choice Question.

SEcnON-C

Select the· COfTIlct alternative out of given ones:

Q. I Chan. and graphs are the presentation of numerical facl.!l by means of:

<a) points and lines

(b) are. and other gcometticol forms (e) symbols

(d) oil the above

Q. 1 Graphs and charts facilitate: <a> companson of vahres

(b) 10 know the trend (c) to know relationship (d) 311 lhe above

Q. 3 The purpose oerved by diagram. and chan is: (a) simple presentation of data

(b) In avoid Ulbul .. tion

(el to avoid textual fonn

(d) all the abo ve

Q." Choice of. particular chart depeads on:

drheberrcchtllch geschulzlcs ~Ia\eria

DIAGRAWlnc NlO GRAPHICAl REPRESE~TAllON

(0) the purpose of study (b) the nature of di", (e) the type of a .Ydien"" (d) all the above

Q. 5 Rectilinear co-ordinate chan j, also referred as;

(a) Cartesian co-ordinate graph (b) rectangular graph

(e) both (a) and (b)

(d) none of (a) and (b)

Q. 6 Trilinear chan is used to portray sirnultaneously:

(a) two variables (b) three variables (e) four variables

(d) nny number of variables

Q.7 The shape of 0 trilinear charts is ih .. of a: (a) cone

(b) cube

(e) equilateral triangle (d) pyramid

Q. 8 Which of the followings is a one-dimensional diagram?

(D) nor dlegram (b) Pie-chon

(e) Cylinder

(d) A gmph

Q.9 Which of the followings is not 3 twodimensional diagram?

(a) Square diagram

(b) Multiple bar diagrom (e) Rectangular d iDg ra m (d) Pie-chan

Q. 10 Wbich of the following statement ;.5 not correct?

(a) The bars in a "istag",m touch each other (b) The bar in • column chan touch each

oilier

(c) There are bar diagrams which "'" !:nown "-! broken bar diagram,

(d) Mulliple bar diagram. also exist

Q. II Non·dimeru;;onal diagram' are also known as:

(n) cubes

31

(b) spheres

(c) piciogrnms (d) .11 the above

Q, 12 \\lhich of the sietcmenc is correct?

(a) Histograms nnd historigrams are 'imilor in look

(b) Cube and square dil'lgrams arc similar in look

(e) Pie·chan and ogives (d) none of the above

Q. 1.3 0 give curve occur for.

(a) more then t~pc distribution (b) Ie ss than I~I'" distribetion (e) both (a) and (b)

(d) none of (a) and (b)

Q .. 14 In an ogive curve, the points are ploued for: (a) The values nmJ frequencies

(b) The values ond cumulative frequencie' (e) Frequencies and cumulative frequencies (<I) None of the above

Q. 15 A semi-logaruhmic graph or ratio chan is obtained by plotting the poin ... :

(aJ Ix.y)

(b) 1101'. x, log y) (e) I~, log y)

(d) lx, log )'/~I

Q. 16 Ogives for more than type nod less than type distributions intersect at

(a) mean

(b) median

(c) mode

(d) "rig in

Q. 17 When Ihe ,al""& are large in magnitude in a chronological ser] es and variation amongst values is small, " graph is better drawn by choosing:

(a) a false base line (b) wid. scale

(0) narrow scale

(d) none of the above

Q. 18 In a bar diagram, the base line is: (al horizontal

(b) vertical

(e) raise base line

U rheberrec htl ich qesc h utztes M aleri a

32

(d) ony of the above

Q. 19 In a column chan. Ihe bose, Ii ne is: (~) horiaontal

(b) veJ1ical

(c) al an angle of 4S· (d) fnlse base I i no

Q. 10 In ~ column chan. bon nrc, (a) horizontal

(bj vertical

(e) slanting

(d) none of tile above

Q. 21. In a bar diagram, the bars are: (II) horizontal

(bl vertical

(c) slanting

(d) none of the above

Q.2Z In case of frequency distribution with classes of unequal widtl!., the beighlll of b= of a histogram are proportional to:

(n) class frequency

(b) class inter>als

(e) frequencies in percenlllgo (d) frequency den,itie,

Q. 23 Yenrwi.., production of rice, wheal and maize for the I as! ten yc.m can be displayed by: (a) simple column chart

(b) subd ivided column chart (c) broken, bar diagr.un

(d) multiple column chart

Q. 24 Profit and loss of. firm, during various yean can be displayed through:

(a) simple bar "i",gram

(b) duo-direcuonel bar diagram (c) deviation bar cluut

(d) mu ltiple bar di.grom

Q. 2S WIlen for some coon tries. the magnitu<lcs on: small MId for other. the magnitudes ure very large, to portray the dall!, it Is prefe=d to construct:

(a) dev iation bat diagram

(b) duo-directional bar diagram (e) broken bar dlagram

<d) any of the above

Q. 26 With the help of hi<lograrn we can prep=:

(a) FreqU#ncy pol)'gon (b) frequency' curve

(e) frequency distribution (d) aU lhe .bove

Q. 27 Historigram ~ sui!llble for: (a) time series data

(b) chronological distribution (c) none of (n) or (b)

(d) both (n) (IJ'Id (b)

Q. 28 When we have the number of court Cases of different categoriea and information about number of cases sen led. the information ClUJ be beue. portrayed throu$h:

(a) slidi.ng bar diagram

(b) histogram

(c) paired bat diagram (d) column chart

Q. 29 lb show the maximal and minimal values in a time series, thesuuable chon is:

(a) deviation bat diagram

(b) range curve,

(c) histoeigram

(d) aU the above

Q. 30 With the help of ogive curve. one can determine:

(a) median

(b) deciles

(0) percenules (d) aU the above

Q. 31 l.oreot curve i. an indic'.tor for the distribution of two factors;

(a) being eq ual

(b) being unequal

(e) both (a) and (b)

(d) .neither (a) nor (b)

Q. 32 Greater the distance ofLorenz curve from the line of equru distributioo:

(a) more is the inequality

(b) less is the inequality

(c) h as to tellnothing aboul inequality (d) none of the above

Q. 33 Pietogram. are: (a) very occurate (b) least accurate

U rheberrcc htl lch gesc h (jutes M ateri a

DIAGRAIAA TIC AN!) GIW'HlCAl REPRESENTATION

(e) mOSlJy used

(d) scientifically correct

Q.34 Pic1ogram. on: generally used by"

(~) =togrnphcrs

(h) dilellante Q. 41

(e) scientist

(d) ~II the ~bove

Q.35 PiClogram' are suitable for the data in:

(a) counts

(b) interval.

(0) fTllClion

(d) none of lIie above Q.43

Q.36 Pictograms ore shown by,

(n) dot.

(b) lines

(c) circles

(d) pictures

Q.37 In II column chan on circular b ase , bars are:

(a) vertkal Q.44

(b) horizonllli

(c) slanting

(d) curved Q. 38 If there i. an increase in a series at constant rate, Ih<! graph will be a:

(n) convex curve

(h) parabola

(c) concave curve

(d) • straight line from left bouorn 10 right lop

Q.39 If there is a decrease in n s erles at constant rate, the 8,"ph will he .:

(0) hyperbola

(It) a ,Slmightline from left lOp to right bottom (c) a ecnvex curve,

(d) acne of the above

Q.40 A semi-lngarithmic gr.tph of a series increasing by a cnnstant amount will be e:

(0) stmight line at Mgle of 45°

(h) a <(lOYe.> upward carve

(e) "concave upward Curve

(d) • convex dDw,nward curve

Q. 41 The suitable chart 10 empbasiic the di.{fcrencc between two time series. of which one is al a higher leYd. is:

33

(a) range c1tan

(b) de>iation, bar chart (e) p.ifl:d bar di.grnms (d) band chan

When there is • pronounced skewness" the desirabte scale to plot the frequency distribution is:

(a):uillimclic scolc

(b) multiple scale

(e) logarithmic scale

(d) any of the above

When there are , large number or values .in an indivjuu,1 serie s, preference for portraying the d<ll. goes 10:

(0) bar diagram

(b) column chan

(e) (ine chart

(d) scatter diagram

An .hl,rnative cban to pie-chan Is: (a) step bar diagram

(b) !'OWIngul ar cbatt

(c) .• phere

(d) none of the above

Q. 45 The graph ofllte successive point. ofa distribution joined by straight lin es : in 'll!!~tieal term.inology is known as:

(a) frequency distribution

(h) frequency curve

(c) trend

(d) cumulative di.,tributioo curve

Q.46 A deviat'().nal or bilalcrnl chan willi 100 pcr cent component columna is also known us; (3) duo-directional chari

(b) fio.lling, column chan

(c) sub-di v ided column. chan (d) range chart

Q. 47 Pie-chart represerus tbecnmponents of a fbctor by:

(0) percentages (b) .ngles

(e) sectors

(d) eircl es

Q. 48 The immigration and OIl!m±grntion or people in a number of countries and ,;also the ncr

U rheberrec htl lch ge5c h utztes Iv! ateri"

rmgration con be, beuer di'played by: <a) duo-directional column chart

(b) smss-de\'ialion column chart

(c) net deviation column chart

(d) range chart

Q. 49 Common ferm of rectilinear co-ordinate grapb!s;

(a) band chart (b) surface chart (c) Slralo m chart (d) all the above

Q. 50 An arilh metlc chart o>an:

(a) have only one amount scale (b) have multiple OII1,ount scalc (e) be, without any amount seal" (d) none of the, above

Q. 51 II i. ~ 10 find commutlltive fn:quenclt'Ji in order to draw Q/an:

(a) hi$logram

(b) frequency polygon (e) Og;"" curve

(d) column cItnrt

Q. 51 If we plDl the points of a less than type or more, than type Jrequ.,ncy di5lribution. the shape of grnph is:

(a) 011;"" CUIVc

(b) sc auerdiagmm

(e) zig-zag curve

(d) parabola

Q.53 Histogram is suitable, for the data presented as:

(a) continuous grouped frequency distri-

bution

(b) discrete grouped frequency di'tribution (e) individual series

(d) all U", above

Q.S4

A histogram can be drawn for the distribution with unequal class interval. by considering: (a) cl ... , frequency

(b) height of bars proportional to class interval.

(e) height of ban< prcporticeal to f."'I""""Y density

(d) an the above

Q .. 55 In a histogram with equal class intervats; heights of b ar are propcrtlonal to:

(a) mid-values of the classes

(b) frequ.,ncies of respecti ve classes (~) either (a) or (b)

(d) neither (a) nor (b)

.Q. 56 The data relating to the number of registered allopathic and homeopathic doctors in six dilIennt states can be most opproprilll.ly represented by diagram:

ta) line graph

(b) hiUOgram

(c) pie-diagram

(d) double bar diagmm

Q. 57 Histogmm can be used only wbcn:

(a) class interval. an: equal or unequal (b) class interval, are nil equal

(e) class intervals are unequal

(d) frequencies in 01..,. interval are equal.

Q. 58 The most appropriau. diagram to represent tbe data rebting to the monthty <npe!ldilUre on different item. by a famity is:

(a) histogram

(b) pie.di"gmm

(e) frequency polygon (d) line gntph

Q. 59 The gross income. taxes IUId net income of B manufa.:lurerduring diffe",ntyears can better be represented by:

(il deviation bar di.gram (b) broken bar diagram (c) paired bar di>ogmm

(d) duo-directional bar diagram

Proportion of males and females in India in di fTerent occupation, in the year 2000 can most properly be represented by;

(a) sliding bur diagram

(b) deviation bar diagram

(e) $Ub-di~ided b ar dill,,"",

(d) multiple bar diagram

Q.60

U rhcberrec htl lch gesc h utztes M ateri a

CW3lWM.llC NIl! G!W>tlICALREPRESEN.T"T1ON

ANSWERS

Sectlon-S

(I) r.obulntion (2) diagrams (3) medlnn (4) not same (5) toech (6) not suitable (7) histogram (It) Bar (9) Comp<ment'l (10) different (II) Deviation (12) continuous frequency (13) circular (14) IWO (15) three (16)' non (17) tn:nd (18) nuctu31ions (]9) frequency densities (20) area (2 t) too llUUly (22) group (23) multiple (24) paired bar (25) frequency polygon (26) frequency curve (27) raiio chlllt (28) """i.logarithm,le (29) M.D. Lorenz (30) inequality (31) equal (32) vari:ibilily (33) least (:14) Dr DUo Neurath (35) bar diagram s (36) log 0

= ~ 1;0.

SECTION-C

(l)d (2) d (3) d (4) d (5) c (6)b

(7) C (8) a (9) b (10) b (11) c (.12) a

(13) C (14) b (15) c (16) b (17) a (18) b

(J9) 0. (20) b (2l) a (22) d (23)d (24)c

(25)c (26) d (27) d (2S) a (29)h (30)d

(3llb (32)" (31)b (34) b (35) • {36)d

(37) c (38) d (39) b (40)b (41)d (42) c

(43) c (44) a (4S)b (46)b (47)c (48) b 35

(49) d (50) b (5 J) c (52)" (53) a (54) c (55) b (56) d (57) b (58) b (59) d (60) a

Suggested Reading

1. Agarwa~ B.L, Bade SUl&tics. N""" Age Internenonal (P) ltd. Publish ers, New Delhi, 3.d edn., 1996_

2. Colvin. F.S. and SI.nlon. B.s .. Hwulbook of Graphics R"pf<sml,,/ion. John Wiley &. Sons. New York. 2nd MO., 1979.

3. Devore.. J. L.. Probability and Statistics far Engineering and the Sd"Mts, BrookslCole Publishing Coo, C.Jiforni~, 1982.

4. Gupta. C:B,. An /mroductitm 10 Stalls/ieol Methods, Vikas Publishing HOU.'le, Delhi, 8(h edn., 1978.

S. Kenny, J.E ""d Keep'na. as" Mathematics of Sta1ist~s, Pan I, D. Von NostnUId Co., N"""Yofk, 1951.

Ii.. S~ncheti. D.C. and Kapoor, V.K." Statistic.

Su1!lII1 O!lI!ld &. Sons, New Delhi, 7th edn .• 1991.

7, S'h"kl a, M.e. and Oul.han, S.S., SIOtisrics, S, Chand &; Ca., New Delhi, 1970,

8. Simpson. G. and Kalka, E, Basic Statistics, Oxford &; j,BH, Ca!culIlI, 3rd indian print,

1971. .

U rheberrec htl lch gesc h utztes Iv! ateri a

Chapter 4

Measures of Central Tendency

SEcnON-A

Shon Essay Type Questions

Q. 1 What dQ yoP undentond by measure of Hannonic roC31L, (iv) MediiUl; lv) Mode; (vi) central tendency? Quartil"" (vii) Oetil ea ; (vii) Decile.; (ix) Pen=tiJes, ott.

Ans. IL is a single value within the range of data which represents a aroup of individual valu", in a simple and concise maneer $0 thlll the mind ca 0 get a quick understanding of the gen era l size of the individuals in the group. Since the value lies within the range of <lalll,il i. known as • measure of central tendency (value).

Q.:: Quote the statements about eeatral values given by John I. Griffin, R.A. FISher 31LdAL. Bowley.

An!. The Statements of the prominent stalisticians about the central volues are:

John L Grilli", An a.verage may be thOUght of ..

a me as u re of central value, .

R.A. FIsher: The inherent inability of human to grasp in its entirely a large body of numerical data compels us 10 seck relatively few constants that will adeq ualely describe the data.

A.L. Bowley: Averages arc s!atistical constants whiclt enable u •. to comprehend in. single effort the ii gnificance of tho wbcle

Q. 3 Name different measures of central values. "A"t)!J. Dirk",.t measures of ocnlml values are:

(i) Arithm~tic mean; (ij) Geometr.ic mean: (i.i;)

Q. " What are tit." properties of a good mew;ure of ccn(ra! teDdcocy?

A.... The desired propeni.cs of a good measure of central tendency are:

(I) It should be bIIsed on ill] obiaYntions of n set

of values.

(ii) It should be rigorously defined. (iii) It should be easily computable.

(iv) It should be leas!. affected by extreme values. (v) It should fluctuate le .. t from sample to

sample drown from the same population,

Q. 5 Whll! are the major classifications of averages?

ADS. Major classifications of a~'el'ligcs are:

(i) Mathem.tical D,ve"'lle&-llfilh.metic mean, geometric mean, harmonic mean.

(;i) PositlOlUll ~ian. mode, quaniles, quintiles. octilcs."decilcs, percentiles.

(iii) Commercial nverages-moving average, progresSive average, composite average,

What are the functions of averag .. ?

Ans. Various functions of averages are",

(i) It presents a simple and conci<cd picture of

U rhcberrec htl lch gesc h (jutes M ate,i a

large and complicated set of data.

(ii) It makes possible and easier 10 compare two or more group" of data.

(iii) II provides a study of differenl groups of data.

(i~) It r.cilitotes the tmerpeetauon of data.

(v) It help" in taking decisions, For CJ<amplc, one can tell whether the .income of a fnmily i.$ above or below the average income.

Q. 7 Whal are Ihe limiunicns of on average in genem.?

Am. The gen ... allimitations of an average are: (1) A mean represents a group as • whole, not an individual.

(iiI Orle.,n average is a v.rue which docs not •• ist in Ihe set of data.

(iii) Sometimes an overage gives a value which i. not f cas ible, e.g., ove::.ge size of a family i. 3.62.

(iv) An .vertlge is incapable to throw any light on

the conscnnion of the series.

Q. 8 Deline arithmetic mean (average).

Ans.. Arilhmelic mean (A.M..) of. set of data may be deli ned as the sum of tile values divided by the number of values in the set, lIS formula is, A.M. '" Ix;/N for i ~ 1,2,3 ... " N.

Q." What are the merits of arithmetic m ean 1 Ans. Meri Is of arhhmeticm ea n are:

(I) II is easy to calculate. (ii) It is rigidly defined.

(iii) It is based on all observaticns,

(iv) Observations are not to be arranged in order. (v) II provide s s ound basi. for comparison of

two SCt1¢$_

(.n It can be calculated if panial calculerions arc

ovai lablo, ~

(vi;) ltis less susceptible 10 sampling Ilucmatlons, (v;ii) It ls most suitable for further algebraie treaunent.

Q. 10 Whal are the demerits of arithrnetle mean? AIlS. Demerits of arilhmelic mean are:

(il It is 100 much affected by extreme values. (ii) Mostly il does not cOITCSJXlnd to .ny value of theset of observatieas.

(iii) It cannot b •• atculared for frequency dislributi on with open end classes,

(iv) It does net convey any information about the spread or trend of data.

(v) II i, not uuilllb!e. measure of central value in

ease of highly skewed distribution,

Q. 11 Prove tIIal thesum of deviation. from mean is zero.

Am. Suppose xl,xl'.' ".%.11 ore n. obsctVations in a:

sec TIle sum ofdevlotiOM from rn ea nx I, S{XI-X).

I-I

. .

~ L.tj - L,.t'j ~ 0

j .. 1 j.1

Q. 12 Give !he formula for weuJating [tIC"" of a set of values from coded data,

Am. The formula for ealculati!\g mean from coded data i_51

•

L!ixi

x;;;;;; a+_'-_' __ ~c

tr;

1_1

where, " is the constant value which ls subtracted from each observation and c is' OJ cnnstant which divides each value obtained afier.ubtracling D.

Also f. is Ihe frequency of the value x, and

x; '" {Xi-a)!e.

Q. 13 What is the crfect of c ban go of origin and scale on arithmetic mean?

Ans. Arithmelic mean is increased or decreased by the eunstantvalue added or subtracted from each

observation respect; vely, Also it is .!_ Ii me of Ole

c

original mean if each observation is divided by c and c times the original mean if each observadon is multiplied by c.

U rheberrec htl lch ge5c h utztes Iv! aleri a

38

Q. 14 What is the difference between ,umpl. mean and population mean?

Ans. Population mean is bused on the velues of each and every item of the population, wberees sample mean i. based on the values of items ,elected in the S3mpl. from the population. If "'" h av .., a random sample.theathe sample mean is 110 esti,mate of population mean.

Q, 15 There ore (0 + I) observations in a sample, If X, is the mean of the first n numbers and:t:z is the mean of the 1 as , n numbers" prove that

Xl E Xj ... 1. (x ••• -XI) n

Am, Proof: By Form"]',

XI-1.(XI +xI+"'+ x.}

"

~ X1 +.!.(.r,,+! -Xt)

n

Q. 16 E, press weighted mean in brief

An.. If x"x2' ... , x, are !he values having weigltu WI" W1!" 'I W,1t: respectively, the weighted mean,

;;:;;;; w1_..t"1 + wl.x.2 +.~.+w,t.xk

\VI + W';z, + , .. + WI

Q. 17 Whal would be, the weigh led mean or " natural numbers if weights n", the corresponding numbers?

AIlS. Using Ibe fotmula" tbe weighted meDJ1 i=- I.l +22+." +n."

1+2+ ... +n

I' +2'+", +,,' 1+2+ ... +n

PROOAAMMED STATISTICS

" (20 + IX" + 1)/6 n (n+ 1)/2

2.+1

-3

Q. 18 \Vh;)1 is a mOVing average?

ADs. In this type of.veragc, we get 0 ""';es of overages OUI of a series of data. rn. thi. we lake a fixed number of beginning items and find its overage. To gct the next average, delere the fi",t item of the previous group and add next (II1e item of the series 10 it and find its average. Continue th;s process lill ,an the values are exhausted. This gives ... series of avereges, For example, suppose there Are five ;1.111.'1, a, b, c, d. ~ in the series. 'Ibkil\g three !Ierns in a group, Ihe moving average> will be,

a+b+c b+c+d c+d+~ ~'---3--'---3--

Q, 19 Where do youumise moving ave"'ges? A.... Moving QVcrnges are utilisedmo.lly in. the analysi. of time series data to know the trend.

Q, 20 What do you understand by progressive !I\'CIUgc,?

ADS. In progressive average. we find our the average of tbe firs' two periods value and Ihen go on addl ng the value for the third, foonh. fif1h period, etc., and calculate the average< of three four, five, etc .. period' successively. Such averages are called progrtssivt O"''''g'$. For example, if a, b. c, d and ~ arc' the values of five successive periods. the progressive averages arc::

o+b a+b+c o+b+c+d o+b+c+d+~

-2-'-~- 4 5

Q. 21 !'oJ w hal purposes tbe progressive averages are uti lised?

A11$. Progress; ve averages are' used 10 know the average profits in the starting years of commercial iO$tih1lions.

Q. 2Z What is meant by compcsite average?

A11$. The average of !he """",£es of diff"",m series ls known as composite average, The formula to,

U rhcberrec htl lch gesc h utztes M ateri a

MEAS.URES o. CEHTIW. TEND<NCY

sum of the ",'CI'IIg<:S of individual series compcstte average ];I; -===:::.:,:;

number or series

Q.23 Whal iscombined or pooled m ea n?

"ns. The mean of the data Ihal would be obtained b) combining the vatues ofall individual ,~ph~$_

Q. 24 Give the formula for finding out the combined (pooled) mean (I f three samples when their mean, and sampl e s izes are given.

Ans. Suppo,," i',I' i.l• i.1 are three sample moons base on nl.n2, and n.l observations respectively, Formula for the pooled or combined mean is.

~ n~i'. +n2.i:z + 1I.~jl

Xl2..' _. -_.-

111 +n"2 +nj

Pooled mean is the mean obtained on combining all sam ples,ounsidered as a single samp Ie.

0. 2S Elucidate trimmed mean.

AIlS, S" PP'''''' we wan 1 'a JO 1"" «:01 trimmed mean, then the mean of the observations b)' •• eluding JO per cent smallest and 10 per cern largest observation. is known as .0 per cent trimmed mean, NOI necessarily lO per cent trimmed mean ha, 10 be found out, it can be any other chosen pereemage which is considered ,suitable.

Q. 26 In" freq u.ncy distribution, the mid value of tlre fi~ class is (0 and each cia'! has equal class interval s, '11,e cumulative frequencies of the classes nre 3.18,31,40 and·45. Give the origi",,1 frequency distribution ned find it. '11""n,

Arus. TIle first cia" will be (10 - 2.5) 10 (10 + 2.5). i.e.. 7.5 - .125. The second cias., will be 125 - 1.1.5 und SO on. Their corresponding frequencies 'will be 3. (18 - .'I). (37 - IS). (40 - 37) 'nJ (45 - 40), i.e., 1.15. 19., 3 and 5., Hence the frequency distributiou IS:

1.5---12.5 12.5-17.5 17.5-22.5 22.5-21.5 21.5-l2.5

3 15 19 '3 5

39

'In the 000",", distribution, the mid-values oflhe classes are 10, IS, 20, 25, 30 ~I"'C(iv01~. Hence. ~'" mean " 0 f the d istribu tie n is,

JO~ 3+ IS~ 15+20 x 19 ... 25 d+30d

I'~

860

="4S-

" 19.11

Q. 27 Define geometric mea n.

Aas. Geometric mean is the nib root of llIe product of " val ues of a set of obsc rvati on" By formula,

G ~ (~'L ,A', •..... k".J"".

Q. 28 What are 'he merits of geometric ,""",,1

Merits of geometric mean are:

It is least affected by extreme value,.

h is based on ,,11 observations of the set.

It is suitable for further algebraic treatment.

\Vhal are the dcment-5 of the geometric mean? Demerits of the geometric mean are:

I", calculation is semewhat complicated.

It cannot be calculated if any of Lho value in the set is 'zero,

(hi) If anyone Or more values arc negative, either geometric mean. will not be calculable Or an absurd ,",I"" will be obtained.

Ans.

til (iiI (iii)

Q.29

A ns, (il (iiI

Q. 3D For what type of d.", genmctric mea" is the most appropriate meaaure of location?

An.. It is appropriarc r", 'VC1"a8ing the ratios of change. for average of p,oportion~, etc,

Q. ,31 Where geometric menn is considered (0 be 111" best ... rage theoretically?

A:ns. Geometric mean is considered.most suitable il,\Wage for index numbers.

Q. 32 If G, is the geomemc mean of n 'I' observations Xi' X:2~- -.~ xiIIl' and 02 is the geometric .nccn of "2 observations Y., Y'2,1 --.! )I.flll and lel G be the geometric rnenn of all (n, +n,l ebservanons.prove that,

logG~_I_- -. {.,logO, +", .logO,) nt+n:r

U rhcberrec htl lch gesc h utztes M ateri"

40

Am. By furroul., geometric mean of (n, ... II,) omcrvations is.

logG,,_I_log(x, x x·y·

n. +~ -' t·····.n,! ~

(y,·h .... y.,)}

_ I"

We know, log. a.M. ,,- £."Iog x.,

n ,

or Llogx, "n log a.M. ,

I

10gG" _ .. ---- {rr, IOJ:,G, +n.l log G,}

"1 +n:2

Q.33 Explain briefly the inver sc property of geomerrlc mean,

ADs. If iG is lhe geometric mean of .(ltX2t· •• , xJllt !h" geometric meon orthe inverse ofthc observations 1/ X,.1f.1"2 , ••••. 1/ x •. is IriG- Thi' is known as in.verse property of geometric mean.

Q. 34 The arithmetic mean of 'WI) numbe .... i. 10 and their gCOltlCtric mean is 8. Find the numbe, ..

A.... SUI'(>O"" the 'wo numbe .... are x, and x,. Then by formula.

A.M.: XliX' "to G.M.",F,-:;; "8

(x, + x,) '" 20 and ... , Xl m 64

}'lc ••••. }'., )

PROGIIAMMED STATISTICS

(x, -x,f -(x, +%.,)'-4", .. , "400-4 ~ 64 "144 (",--"2)" ... 12

Solving and

.. , + .. ,-20 -",-x, '" 12,

'" ,,16, x" =4

we get

HeDCe, the numbe rs are 16 and 4.

Q. 35 Define harmonic mean and give the formula for h:umonic mean.

A.... Harmonic meanis the inverse of the ar.i,n.· metic mean of the reciprocals of the observations of a set.

The formula for harmonic mean i.,

"...!:J.fL_ for I ~ J,2, ... ,k

E,M",

where,.'; is Ill. frequency of x, for j = J, 2, .. _, k. Q. J6 Wha, are the merits of hannnnic mean? A.... Menls of harmonic mean are:

(i) 11 is based on all observations of a set.

(iil [I is a good mean for a bighly vnri.b!e series" (iii) 1'1 gives more weigblag" 10 the small values

and less weigblage '0 the large values.

(iv) It is better than weighted mean since in this, values are automaticolly weigbted.

Q. 37 Whal ore the dementsof hOrtnOhic m .. n7 ADs. Demerit> of harmonic mean are:

(;) lis calculation I, complicated.

(ii) [f any .0)"" is zero. il canllOl be calculated, (m) lIS volu e is gencrnlly nOl '. member of 'he

series.

Q. J8 Con we use coding while calculatiag geometric mean or harmonic mean?

".... We eannot use ending in gecmetric mean Of harmoni<: mean as no easy expression is available

U rhcberrec htl lch gesc h utztes M ateri a

MEASURESOf' CENTRAL TENDENCY

which relates the mean or coded dam lo the mean of the original data,

Q. 39 How will you calculate mean in case of a continuous f,equenc'Y distribution?

AIlS. To calculate the mean of a cnnti nuous frcqucncy distribution, first find the mid-value. of the classes and use them as variate values. Then U-~ the formula for the weighted mea" with frequencies .. weigh".

Q.40 Given two vatues '~J and ~" prove Ih.1

A, ~!.<: O.M.2: H.M.

ADS. We know,

A.M.= x, +x1 ,a.M.: AA,RM.=b,:r,

2 ..1", +x'l.

Again, (F. -ii;)' 2:0

or. (x, H,).[X;X; 2:2(AA)'

a.M.,2:H,M

A. M. 2: G. M. 2:. H. M.,

Q. 41 "'pres' geometric mean in terms of arijhmeuc mean and harmonic mean.

Ans. The relation between geometric mean • aruhmerle mean and harmonic mean is.

G.M. '" JA.M,": RM,

Q. 42 Express harmonic mean in terms of geometric mean and arithmetic mean.

41

AIlS. The expression for harmonic mean relating lo geomeiric mean and arithmetic menn is.

H.M. = (G.M.)' /A.M.

Q. 43 For what type of va lues harmonic mean is suitable'?

AIlS. Harmonic mean is suitable when th,e values are pertaining 10 therate of change per unit time such lIS speed, number of item. produced per day, cnntracts completed per year, etc, 1.11 generat, harmonic mean js suitable for lim e, 'peed, rates. prices. etc,

Q. 44 Define quadratic mean.

A.... The square root of the arithmetic mean of the square of numbers of a sct is known as quadrati.c mean.

Q. 45 When do you prefer a quadratic mean than a mean.

AIlS. When one has some ne gauve n umbers in • set, quadratic mean is preferable than any other mean, The reason being thot cOI15idering the sign of negaUve numbers, the mean value is considernbly reduced. If the sign, of negative numbers at. ignored, the mean thus obtained depend. on an al.gebraic IH.. lIU( tbe quadratic mea" is devoid of the ""ahem.tical.buse, and has the same scale and units as the original values.

Q.46 Comment on the value of qUlldralic mean '" compared 10 ari1hmelic mean.

Ans. Qencmlly the quod,,!!ic mean is greater than the arithmetic mean. The reason being that 'he di ffercnce between a number and its square increases exponentlally as the number g,e!S larger" Hence, the squareroot of the average of squared numbers results into a higher value, than arithmetic mean of original numbers.

Q. 47 Find (he arithmetic mean (A.M.,,) and quadrauc mean (Q.M,) for the following set of ,.1ues and comment on the mean values.

X:

2

3

5

8

12

A.M. = 2+3+5+8+12

.. 5

~6.0

U rheberrcc htl lch gesc h (jutes Iv! ateri"

Su m 0 f squares of numbers,

EX' - 4+9+25+64+ 144·~ l46 246

A.M. of squared numbers ~"5 049.2

Q.M.~ J49i ~ '1.()1.43

Q.M. is greater than A.M. as is expected on the basis of our theoretical knowledge.

Q. 48 Where do you make usc of quadratic m ean in general.

Ans.. Quadratic meal} is absolyt~ly ul_ili~ed I_n findiog the standard deviatlon of a se1 of dam. As a mauer of fact. the standard deviation i. nothing but the q uadrauc mean of the deviations from data mean. Q.49 What do you understand hy median?

AIlS. It ls a most preferable measure of location for asymmetric distributions. Medi.n is the value of the variable thai divides the onJec"d set of valu05 into two equal naive s, 50 per cent values are to the left of the median and 50 per cent are to the right of the median.

Q. SO What arc the point,. in favour of median? An.. Points in favour of median are"

{il Melli an is not influenced by extreme va lues because it is a positional average.

(iil Median Can be calculated in case of distnbution wilb open end intervals

(iii) Median can be located even if the data ore incomplete.

(iv) Median can be located even for qualitative facto.1S such a.I ablllty, honesty, etc.

Q •. 51 What arc the demerits of median? An.. 1'Q1Iowing are the demerits of median:

(i) A •. light change in Ihc series may bring drostic eba.nge in median value.

(ii) In case of even number of observations or continuous series, median is an estimated value-s other lila" any value in the series, (iii) It is nOl soil.ble Ior further mathematical treatment except its usc in mean dcvintion,

Q. 52 How will you. calculate median in case of ungrouped data?

An.. To calculate median of ungrouped data. followi nJ: steps be followed:

(i) Arrange the data in order,

(ii) Filr an Individual or dlserete series of N items, the value corresponding to (N + 1)12'" i temls the median value when N i.s odd ..

(iii l When N is even, median is the average of

N'" and ( No. 2)"' ; lem. s.

2· 2

Q. S3 Give the interpol"tion formula for calculating median 0 f grouped series.

AIlS. For grouped data, (he Iormula for median with USUAl rotation is,

M~ '" 4, + N/2- c ~ / f

where. I" '" lower limil of the median class

c ~ cumulative frequency itlSt .b< .. e the median cl as s

f ~ Frequ"!1CY of the median ct ... , / = class interval

Q,54 Whal is !he impact of extreme value, of a set on median?

AI1lI. Medi.n is not affected by the extreme values of a set,

Q. 55 D iscuss mode in brief.

AIlS. The variate value x having maximum frequency ;0 • distnbation is blown .as its mode. Q. 56 What was said about the mode by Yo Lun Chou?

AM, Yo l.un Chou SWornc"t about mode is. "The mode i. that val ue of a series which appears most f requently than OIl)' other."

Q. 57 Con there be more than one mode of a distribution ?

AIlS. Some distributions have equal peak. and hence mode i5 not necessarily unique. There wi.1I be as many modes of distribution as the number of peaks in it

Q.58 How would you calculale mode of. groupell frequency distribution?

Ans. For frequency distributinn with classesIn

U rhcberrec htl lch gesc h utztes M ale,i"

MEASURES OF C.ENIFlAl. TeNDENCY

usccnding order, the formula for calculating the mode is.

~ 1.,;,+ /0-1-,. d 2/0 - I~, -I;.,

where. the class having the maximum frequency is known as model cI ass.

I .. = Lower lim it of the modal class I '" class interval

6,= Difference of modal frequency from its preceding class frequency

4, = difference of modal frequency from its following class frequency.

In the second formula,

4, and J are same as abo v c. 10 == modal class frequency

L, '" Frequency of the preceding class I~, = Frequency of the following cL'ISS.

Q.59 How would you find out the mode of " discrete frequency distribution'/

Ans. It can be determined merely by inspection. The value having maximum. frequency i., the mode of the distribution.

Q. 60 How can One locate mode graphically in case of grouped data'!

An.. Prepare a hi,logo"1I. Consider a bar having ma .• imum 'height and the bars to the left and right adjoining 10 it. loin tho top left comer of the highest bar to the top len comer of the right side bar and lOP right corn", of the higheSI bar to lb. rop right comer of the bar to its left The abscissa value of the point of intersection of the joining lines is the mode.

Q.61 Discuss qunniles and their importance. Ans. Three variate values of the viltinblet say A't which divide the se ries (frequency) into four equal pam ore called quartile. for the corresponding distribution of x. There are three quartiles namely Q" Q, and Q,. 0, is called the lower quartile and Q" the upper quartile, 25 per cent valu es are less

43

than 0, and 25 per cent values are Inrger than Q., and the rest 50 pet cent values lie between QJ and Q:." Q'2 divide d,~ series into two equal parts and hence, it is the same, as median.

A. regards their importance, quartile. ate widely used in economics and business. They are also helpfb! in dct"01,ining the shape of a distribution,

Q. 62 Give tho formula for calculating the quartiles. of a continuous d is-lribul i 0[1.

Ans. Formu I" fllr calcul ",iog ," quartile (i = I, 2., 3) lor a conunuous distrihution having K"clos.e, is.

iN --c

In+-4--~ I 1

where, Q, = "" quartile which is to be calculated 10 = lower I i mit of 11,. t", quarti Ie class

N ~ Total of all frequencies

c = cumulative frequency of tho class just above the quartile do .. ,

1 = frequency of the quartile class I = da. ss i merval.

Q. 63 Discuss decil es in brief.

An.. TI,. variate value, which divide the series (frequency) into ten equal paris arc caned deciles, Hence, there are in all nine deciles denoted by D,. D2 •.... o; The uerns before D, and after D. are JO per cent. Also, the number of items that lie between ""y two deciles is also 10 per cent, D,. the fi fth dedi di 'ides the seri es into halves and hence, it is the same as median.

Q.64 Give the formula for finding the decll es of a continuous distribution.

Ans. TIle formula for calculating deciles of • continuous. distribution is...

iN ---c

D =ld1Q___xl

, " f

for i = I. 2, .... 9.

A 11 notation can be dec oded by rcplac i ng the word quartile by decile !IS in the formulaforquartiles" Q. liS Describe percentiles in brief

U rhcberrec htl lch gesc h utztes M ateri"

,AIlS. The value" of the variable x which divide il$ di:s!ribu!ion into JOOcqual parts W clIlled percentiles, There ate in .11 99 percentile> and are denoted by P,,. P,. , .. " P,,,, respectively, Area between My two pCn:I!~'ilC"$! before p. and Ilftcr Pw is O_Ol~ i.e., it

I

repre<enl'IOO part ofth. population, 50m percentile

is same as median.

Q. 66 Give the formula for celculatlng percentites of a continuous distribuuon.

An.;. Formula for ca lculating ,. percentile is,

il>l -c

P, ~ I. + l.OO__ ~ I. f

for i~ 1.2 •.... 99.

All notation can be decoded as for quartiles by repl""i ng the word quartile by percentile ..

Q. 67 How !O determine quartile s, deciles and percentiles in case of. discrete series?

An.;. Qualtliles; The value. of the variable x cortes-

(1)1 +.1)" (1)1+1)'' d 3(1)1+1)

pending to 4 "-2- .n--4- th

item. of an ordered discrete series are the values of Q" Ql and QJ rcspe<:tively. The position of the required item can easily be adjudged with the help of cumulative Jrequeoc] es,

DuU e s: Similar to quartiles, the values of the vnri-

_ /(/>/+1) _ .

able x corresponding to ---10-- th Item for,,, I. 2,

..,,9 of on ordered discrete series is D,. The position

i(N+1)

of -I-O-th item can be located with the help of

cumulative frequencies.

Pm;emiles:IUSllikc deciles, thevalue of the variable

. ,;(1)1+1),.,

x corresponding 10 """"iOO th nem for , ~ I. 2" ... ,

__ . 1(/>/+1) , .

99 of an ordered senes .. p,·'100 Ih item In Ihe

series can easily be placed with the help of cumulative frequencies.

SEcnON·B

Fill In the Blanks

FUI if! the suitable word(s) or p.hrase(sr in !he blanks

I. "The statement, ~An 3ver:age may be thought of as a measun> of central ~nlue," was given

by _

2. The statement, "The ;nherent inability of human to grasp in its .nurely a large body of numerical da"'. compel, us to seekrel.tively few constants that will adequately describe

the, data," was given by _

3. "Average. are otat;stical constants which enable us 10 comprehend in a single' effon the signit1.nnce of the whole", was ,lated by

4. Measures of central !endeney are also known

._. 0-' _

5. Arithmetic mean is very much affected by

6. Any mcnn is b ased on •• lues of

a set.

7. An average the data,

8. In an open end distribution can-

nUl be determined,

9. If j is subrracted from each observation of a set, the" the mean of the observation is

reduced by _

10. If each value of a set is divided by 10, the

mean of obscrvntiaO$ is of th e

mean. of the original observations,

U rheberrcc htl lch gesc h (jutes M ateri"

11. The mean of the values II, 12, 13, 14 and I~

is ~.

12. The mean of 8 numbers is 15" Afterwards a new number 24 is added. The mean of the

nine numbc::o;. ig _

13. TI,. sum of the deviation. from mean is

14. The average income of a person on working for tbe first ftve days of tbe week is R •. 3~ per d.y and if he works for the first si. day, of the week. hi. average income per day is Rs.40. Then" his income for the sixth dlty is

IS. A fnctory employs 5 persons in the ftrst week of the month and pays Rs, 25 per day per penO". In the second week, he employs 8- persons and pays Rs, 30 per day per perso n. In the third week be employ. 7 persons ond pay. Rs, is per day per person. The average payment per day per person by the employer

is _

16. Weighted mean i, more _

unweighted mean.

17. In c as e of 15 per cent trimmed mean. only _ per cent observations ore utilised.

18. The arithmeuc mean of n natural numbers

for I to n is _

19. Sum of square, of the deviations from mean

is always _

20. The geomeeic mean of four numbers '2. 4, 8

a nd 64 is _

21. If Xt, x, ..•.. xn .nd YL'J'~ ..... y, are the variate v.lues of two variables X and Y and their geometric means ore G, and G, respecn-

""Iy, men the geometric mean of (x./y,) i.

ll.. Geometric' mean cannot be calculated if :my

",Iue nf the .et i. _

23. Geometric mean is suitable when the values

are given ... or _

24. If arithmetic =0 of certain values is 9 and their geometri.c mean is 6, then the harmonic

mean is _

25. IJ the harmonic mean of the two numbers a

and b is 5 if a: 5, then b i. _

l6. Harmonic men is suitable ns an average fOf

finding out the _

27. If the two Qbserv~tions .x, and x ore, such

that x, =-x,. their hlll"lllOnic mean is '

28. The, relation between A.M .. C.M. and RM. j,----

29. The variate value in a senes which divides

the series tnto h.aJve, is called _

3f). Tbe sum of the ""'olute deviations tnkm1 from

median is _

31, The median of the series 3. 18,7,20. II. 12.

9. 17. 22 ls _

32. Medi"" i.s f<lf each and every

than

distnbution.

3J. Two ogivcs, lb. one less man type and the other more than type, intersect nt __ . __ "

34. Medi:m is a more suited average for grouped

data with classes.

35. The variate vul tie having """imum frequency

in a frequency distrit..It_ion is called ,

36. The mode of the distribution of values 5, 7,

9,9.8,5,6.8,7,7.5,7.9,2,7 is .

37. and arc least

affected by the e.m:mc values as a measure of .e.trallende.cy.

38. Out of all tbe measures of central tendency

_ is the only measure ·which is

nOI unique .

39. is not appropriate for further

algebraic treatment as • measure of central value.

40. The distribution which has only one mode is

called _

41. The. d.isuibution having two modes is called

42. If the frequencies in dL'I1ribulio. have ups OM downs, the mnde is determined by ·the

method of _

43. In an individual or di"""'te series .( N: I r

uem is known a" first, _. _

U rheberrcc htl lch gesc h utztes M ateri"

44. Second quartile ts same as _

45. Third quartile and percentile arc

• s..1.me_,

46. Second quartile and decile are

equal,

47. The, decile which precedes 72th percenule is

4lI. The, perce" tage of v,ol ues which lie between

p)!) and P 56 i. _

49. 25th percentile is same as quartile.

SO. Si .. th docile is same as per-

centile.

51. Percentage of values lying between first and

s-ht1h decile i5 _

S,:!.. Percentage nfvalues whicn are greater than

·66th pcn:cntile is ., _

53. III a frequency distribution fourth decile and ____ ' percentile arc same,

54~, The relation between second quartile Q" sixth decile D. and 8!J'l' percentile is

55.. Med;an is same as decile.

56. Medi"" i. same '" percent; le,

57. Percentiles (quartiles, decil .. ) are always in

____ order.

58. The VlIriatc value! which divide a series into

five equal parts are called _

S9. If in 0 serie s, 20 per cent values lI!egreille{

than 75. then ' '" 75,

60. If in 0 discrete series 30 per cent ,yalues arc

less than 35, then decile '" 3,5.

61. If in • discrete series 7~ 'per cent values are

greater 'han 20, 'hen quartile ,",

20 or percentile = 20,

6:!.. 111e general name, fOf quartile. quin'iles.

octiles, decile. and pereemnes is _

63. 'The sunemen I •• The mode is that value of n series which "wc= ino" frnq""n'ly than

any other". wasgiven by ' _

PIIOGRAI.IMED STATISTICS

1>4. The ,.I,'II;Or! between fill! quartile and 'Z!ilh

percentile is ~

65. 'The ,relation between 4th docile and 4Slh

percentile .i" ~

66. Percenlage, of yalues ,ha, Iieberwecn 'entit

and eighty fifth percentile is '

67, The number of type' of statistical facts are

~ Of _

68. For n symmetrical distribution, median in

l"rlm of ~uartiies is _

69. Geometric mean is '_ of the

nrlthrnetlc mean of logamhrnicvafues of a variable,

70. Harmonic mean is , of the arith-

mene mean of the reciprocal of the observanons of a dnl.a. sel. '

71. Qun(!rntic mean i, preferred most, if the data

set contains some, number.

12. If a variable takes some non-negative v~lU<'.\ x I. x1'· . '. x.. then the inequality Ihat holds

bet .... een H.M., O.M. ""dA.M, is .

73. Quad"n," mean: has scale and

"nil' 111 thnt of origin .1. observation.

74, The. applicru::ion of quAdratic mean is evidently

found in _

7S. The mean of 20 obse rva tion.! is 7, 'If each coservatlon is multiplied by 3 and then 5 is added to it, then the mean of the new do ta set

is ; _

76. Tbe mean of a set "f 10 obser;31ion.l is 4, Another set of 20 observations is added. to it which mnkes the mean of lite combined set equal to 6. The mean of second set Js

77. If 2y' - 6,< " 6 end the mode of y .. 66, then

the mode of x ts _

78. An average is thai Yalue of a distribution

which repreSent.< _

79. No average, can be regarded AS _

ror .11 limes and an data,

80. Mode Is defi ned.

U rheberrcc htl lch gesc h utztes M ateri"

Multiple Choice Questions

SECTION-C

Selact' the eottec; alternative oul of given ones.

Q. I Mean is a measure, of:

(a) locaticn (centrsl valoe) (b) dispersion

(e) correlation

(d) none of the above

Q. 2 Which of the following is. me cs ure of central value?

(0) Median

(h) Standard devi atio n (e) trle:m deviation (d) Quartile deviation

Q. 3 Which of the folJowi_ng rcpr<:.IenlS medi.1n?

(a) First quartile --

(1)) Fiftieth percernlle

(0) Sixth deci le

(d) None of the obove

Q.4 If a.constam value 50 is subtracted from each observation of" set, the mean of the SCI is: (a) increased by 50

(1)) decreased by 50

(c) is not affected

(d) zero

Q. 5 If a constant 5 is added to each observation of u set, the mean i 5;;,

(a) increased by 5

(b) decreased by :;

(e) :; times lbe original mean (d) not affected

Q. 6 If each observation of a set is multi piled by 10, the mea" of the .. ne.w set of observations: (a) remains 0, e .ame

(b) is ten times the original mean (c) is one-tenth of the origi nat mean (d) is increased by 10

Q. 7 . If each value' of. series is multiplied by 10. the median of the codod value, is:

«(1) not affected

(0) 10 limes the original median value

(e) one-tenth of the originnl median value

(d) increased by 10

Q. S If eachvalue of a series is multiplied by 10, the mode or the coded values is:

(a) not affected

(b,) one- tenth of the <>rigi nal modal. value (e) lO-time, of the original modal value (d) J()().tim'cs of Ihe original modal value

Q. 9 I r each observation of n ",t is di y ided by 2, then the mean of new \1'31 ues:

(a) is two limes the original mean. (b) is decreased by 2.

(e) is half of the migin.l mean (d) remain, the sam.

Q. 10 Which of th. following relations Among the location parameters doe s . not hold?

(a) Ql = median

(b) P ",= median

(c) D, ~ median

(d) D6 = median

Q. II I f the grouped data h .. open end classes, one

can not ca leu late: -

(a) median

(b) mode

(c) mean

(d) quartile,

Q. 12 Geometric mean of lWO observations can be calculated only if:

(.) both the chservnuons ere positive (b) on e of the tWO observations is zero (e) one of them ;, negative

(cl) both of them arc zero

Q. 1.3 Goomclric mean Is beuer lhan other means when,

(a) the data are positive as well as negative (b) the data are in ratios or pcn:cmOjlC5

(c) the data. are binwy

(d) the dat. are 00 into",.l scale

Q. 14 Geometric mean is a good measure of central value if the data arc:

U rheberrec htl lch ge5c h utztes M ateri"

48

(a) categorical

(b) on ordi nal scale

(e) in ranos or proportions (d) none of the above

Q. 15 Harmortic mean i. better than other moan. if the data are for:

(a) speed or roles

(b) height' or Jength.,

(e) binary values like 0 _lid I (d) ratios or proportions

Q. 16 The correct relationship between A.M .• G.M. andH.M.is:

(a) A.M. = a.M. = H.M. (b) a.M.;" A.M;"< H.M. (c) H.M.;"O.M.;"A.M. (d) A.M.;"<G.M.;"<H.M.

Q. 17 E~U'erne value have no effect on: (al .vcrage

(Ii) median

(e) geometric mean (d) harmon!" mean

I 4

Q. 18 Geometric mean of two numbe .. iii and 2S

is:

1 (a) iO

I (b) 100

(e) 10 (d) 100

Q. 19 Expenditure durin, r,f$I r,,,,, months ora yeas is R .. 96 per month and during last seven momhs is Rs. 120 per month. The lIVWIge e.penditure per month during wbole yeor is: (i) Ro. 108 per month

(b) Rs .. 110 per month

(e) Rs. JOO per month

(d) Rs.216 per month

Q.ltI A~ 5lrength of eleven members = J 1.0.

A....:mge strength of \be first six members = 10.S. Average strength of the last ri. membe rs = 11.5.

lbc average suength of the .i'lh member is: (aJ 10.5

(Ii) ItS

(e) 11.0

(d) 10.0

Q. 21 The average of the 7 number 7.9, 12,.r., 5. 4, II is 9. The mi ss ing number .r. is:

(a) 13

(Ii) 14

(e) IS

(dJ 8

Q. 21 The mean propordon of 0.16 and 0.01 is: (a) 0.4

(b) 0.17

(e) 0.085

(d) 0.04

Q. 13 A lraj n covered the first 5 km of i IS journey at aC!peed of 30 kmllt and no.' 15 km at a speed of 4S kmllt. The average speed of the train was:

fa) 35 kmllt (b) 40 kmJh (e) 32 kmlh fd) 42 kmJh

, Q. Z4 The second of lhe two samples has 50 item wilh mean 15. If !he whole group hill! ISO item. with· mean 16, the IlIC8fI of the fint sample is:

(a) 18 .. 0

(b) 15.5

(e) 16.5

(d) none of the above

Q. 15 For a group of 100 candid.tes. the mean was found to be 40. Later on it was discovered that. wloe45 was misread as 54. The correct mean is:

(n) 40.50

(b) 39.85

(e) 39.80

(d) 39.9t

Q.:26 A di,lribulion consi sts of lbree group. having: 40,50 nod 60 ilem. with mean. 20. 26 and 15 respectively. The mean of the dislribuLion is: (a> 20

(b) 18

U rhcberrec htl lch gesc h utztes M ateri a

(e) 22 (d) 2S

Q.?7 The average age of SO student. in abus is 20 years. When the age of conductor is included, the """l1I8e age isincreased by One year, The age of the conductor is:

(a) 51

(b) SS

(c) 71

(d) SO

Q. 28 The average of five numbers is 40 ~od the average of another four numbers is SO. The average of all numbers taken togeth.r is:

(a) 44.44

(b) 45.00

(e) 45.55

(d) 90.00

Q, 29 The a",rage temperature of two cities on first six days of a week is same, 11>e temperature dropped in one city all of a sudden on the ,""",cnch day of the week. 'The, average weekly temperatcee of the two cities differed by OSC 'The di lference between Ihe si. days average dai Iy temperature oftwo cities and the seventh day temperature of the mil er city is:

(a) 2,5

(b) 3,.0

(0) 35

(d) 2.0

Q. 30 What percent.age of value, i. gre.tel than 3.rd quartile?

(a) 75 per cent (b) 50 per cen I, (c) 2Sper cent (d) 0 per cent

Q.31 Wlm' percentage of values is I",. than 3rd decile'!

(0) ~O 'per cent (h) 70 ret cent (e) 40 PC' cent

(d) none of the above

Q. 32 Wh.t percentage of v.1 ues tie. between 51h end 25th percentiles.

(a) 15 per eem

(b) 30 per cent

49

(e) 75 per cent

(d) none of the above

Q, 33 There were 25 teachers in a school whose, mean ege W:t$ 30 ye3I'$. A teacher retired at Ihe age of 60 year.; and a new teacher was appolnted in his plru:e.The mean age of teacher.! in (be school was reduced by one year. The age of the new teacher was:

(a) 25 years

(b) 30 years

(e) 35 years

(d) 40 years

Q, 34 If the A .M. of 0 set of two observations is 9 and its OM. is 6. Then the H.M. of the set of observations is:

(0) 4

(b) 3./6

(e) 3

(d) 1.5

Q, 35 The A.M. of two numbers is 6.5 and their G.M. is 6. The two numbers are:

(a) 9.6

(h) 9,.5

(0) 7.6

(d) 4,9

Q. 3Ii If M., Q, D and P stand for median. quartile, decile and percentile respecfively, (hen which of the following relation between them is (rue?

(a) M.", Q1 "'·D~" P", (b) MJ= Q, = D, = P" (e) M." Q," D." P,. (d) MJ=Q~=D,=P",

Q.37 Which of the following relation is true between 3rd decile and 30th. percentile?

(a) D,=P.,.

(b) DJ '" 'P:!i<)

(e) D,=P""

(d) D1",P""

Q" 38 Wh i eh of the dod les are less than first

quartile'? j

(3) D, nnd DJ

U rheberrcc htl lch gcsc h (jutes M ateri"

50

(b) D, and D} (e) D, and D,

(d) None of the above

Q. 39 In. raclory there are 60 1"" eent Iabourers, !lO percent scribes and 10 per cent e~ecutivcs. On "~Cnlllc, the salnry of a labourer is Rli. 16(lO p.m, of a scribe Rs. 3000 p.rn and that of on exeeuuve R s, 8000 p.m. The average salary of. employee, in the, factory is:

(") Rs. 4200 p.m

(b) Rs. 1166 p.m

(e) lb. 2660 p.m

(d) None of the three

Q. 40 The mean of seven Qbserv3ljons is S. A new ohservati on 16 is added, The mean of eight observnti OIlS i~:

(a) 12

(h) 9

(e) 8

(d) 24

Q. 41 If the sum of N cbservaticas is 630 and their mean is 42, then the ,".title 0 f N i,~

(a) 21

(b) 30

(0) 15

(d) 20

If Jhe two observations are 10 and -10. then their brumonie meanis;

(a) 10

(b) 0

(e) 5

(d) '"

Q.,42

Q. 43 If the nbservauons are 5 and -5. 'heir geometric mean. is:

(a) 5 •

(h) -5

(e) (I

(d) none of (he abo".

Q.44 If the (1010 observations ore 2(J and -20, lheir arithmetic mean Js:

(a) 10

(b) 2(1

(e) u

(d) none of (he above

PROGRAMIIEQ STATI$11C1l

Q. 45 Can. quartile. " decile and a percentile be 'he median?

(8) Only qUllltilc bu, netd ec ile aod pen:enlikl, (b) Quorti Ie "ad decilebur nOl, porcentile (.)rkcll., and' percentile bu, not qu:u1lte (d) Quartile, decile and percentile, all the

three,

Q. 46 When all the observations are same, tben the relation between A,M" O.M. and H.M, ;.: (n) A.M, '" C.M. ~ H.M.

(b) A.M." O.M. ,<: H.M.

(e) A.M." G.M. " H.M.

(d) A.M." C.M. c H.M.

Q. 47 'The pcrcCllt:lgc, of valu.. used in case of 10 per cent trimmed mean is:

(a) 40 per cent

(b) 60 per cent

(e) 80 per cent

(d) 20 per cent

Q .. 48 The ... rage of '2n nat,u131 numbers from I to 2n is:

(a) (n + 1)12

(b) (2n +,1)/2 (c) n (n + 1)12 (d) n (21! + 1)12

Q. 49 A man goes from his noase to his office, al the "peed of 20 kmlh and returns (rom hi. office to home a! the speed of 30 kmlh. His mean speed is:

(a) 24 kmIh

(b) 10J6km1h (c) 25 ktn/h

(d) none of the above

Q. SO Mode is tl .. tvalue in a frequency d i.1ribulion

which (X'SS"SS." -

(0) minimum frequencY' (b) m;t.imum frequency (c) frequency one

(d) none of the abo ve ,

Q. 5J Tbevalue of the variable cOITeSponding to the highest point of a freque ncy: dis!ribution curve represents:

(a) mean

U rhcberrec htl lch gesc h (jutes M ateri a

MEASURES OF CF.NTRAI. TENDENCY

(b) median (e) mode

(d) none of the above

Q. 52 A frequency diSlributioo having two modes is soid to he:

(a) unimodal

(b) bimodal

(e) trimodal

(d) wi tho. t. mode

Q.5.3 If modal value is not dear in. distribution, it can be .. scertamed by the method 0 f:

(a) grouping

(b) guessing

(e) summarising (d) trial and error

Q. 54 Shoe size of most of the people in India is No.8. Which measure of central value does it represent'?

(a) mean

(b) second quartile (0) eighth decile (d) mode

Q. 55 In a discrete series Ie wing (2K .. 1) observaucns, median is:

(a) K'" observation

(b) (K" +1)'" observation (c) (I( .. 2)12"' observation (d) (2K +1 )12'" observation

Q. 56 The median of the variate val ues I I. 7, 6. 9. 12, J5. 19 is:

(a) 9

(b) 12

(c) 1.5

(d) 11

Q. 5"7 The median of lite variate values 48. 35, 36. 40,42,54. 58,60 i"

(a) 40

(b) 41

(e) 44

(d) 45

Q. 58 To find the median (mode), it is necessary to arra nge the data in:

(n) """ending order

(b) descending order

51

(c) ascending or descending order (d) any of ihe above

Q. 59 For a grouped data, theInrmula for median ls based on:

(a) interpolation method (b) extrapolation method (c) trial and error method (d) iterative method

.Q. 60 Which of the mea,ure of central tendency is not affected by extreme values?

(a) mode

(b) median

(c) sixth decile (d) all the above

Q. 61 The middle value of an ordered series is called:

(a) 2nd quartile (b) 5 tn decile (e) 50 percentile (d) all the above

Q. 62 Formula for din:el1y calculating the men" X of an ;ndividtl4l series is:

(a) X ~ IX N

(b) X'" I[X N

(c) X ~ A .. Ed.< where d.<""' X-A

n

(d) X ~ A+ 'Ed.< where dx « X-A N-

Q.63 Formula for "aleu13!ing the mean of an individual series by short-cut method is:

X~ 'EX

(n) N

(b) X - tjXlN

- 'Ed.<

(0) X-A + 'N

(d) X-A + tid.< N

U rheberrec htl lch ge5c h (jutes M ateri"

52

Q.64 r"nnula forcakul,u;ng the mean of adiscrete series by direct method ;J!

(a) X ~ l:..X,r.I

(b) X ~ l:.;/;X;/'L;/,

(e) X ~ .11.+ l:.1t. dxtfl:.fi (<I) "one of the above

Q. 65 Formula for calculating lhe mean or. di>cre!e series by sbort-cur method is,

(a) X~IXIr.!

(b) X ~ l:.!,/l:.!

(e) X ~ II + l:.dxl!.l (d.) X _ A+ l:.ldx

tl

Q. Ii6 1!.s.600 pet day are paid on. r=h fnrm to it! 50 daily paid labourers. A worker get! live unpaid holidays in n mooth .. The aver age income. of a d:J.Uy paldlaboueer is,

(a) Rs. 250 p.rn

(b) Rs_ 300 p.m

(c) Rs. 350 p.m

(d) Rs. 400 p.m

Q.67 The variete values which divide .... ri es (frequency distribotion) into Iive equal pam are called,

(a) qui01l1 es

(b) quartile,

(e) cctilcs

(d) pereentll es

Q. 68 The. varime values which divide a series (frequency distribution) inlO four equal p.:uts are called,

(al quimlles

(h) quartile"

(e) deciles

(d.) percentiles

Q." The variate values which. divide a series (f,equency distribution) intn leQ equal parts are coiled,

(a) quaniles

(b) decil es

PROGRAMMED STA TlSTlCS·

(e) octites

(d) pcn:enliles

Q.70 The variale values which divide a se ri es (freqUCllC)' distribution) into 100 equal pam. are known as:

(a) octile.

(b) quart; I",

(e) percentile,

(d) decikss

Q. 7t The ".,WIe· values which divide a se ries (frequcncy distribution) io(o eight "<1""1 pam

ace known as: -

(a) quarti I es

(b) quintil es

(e) decile>

(d) octiles

Q. 71 The Dumber of partition values in case. of qcanil es is:

(a) 4

(b) 3

(e) 2

(d) I

Q. '73 The number of pll!1i!'on values in case of qUiDlil es cannot exceed:

(a) 4

(1\) 3

(e) 2

(d) I

Q. 74 For decil.s. tbe total number of partition value. are:

(a) 5 (h) 8 (e) 9 (d) 10

Q. 75 For percentiles, the IO!al number of partition values are,

(.) )0

(bl S9

(e) )00

(d) 99

Q. 76 The second decile divides the series in lire ralio:

(0) 1:1

(b) 1,2

U rheberrcc htlich gesc h (jutes Mater; a

MEASURES OF CENTRAL TENDENCY

(e) I : 4 (d) 2: 5

Q. '77 Eightieth percentile d ivides a f re que ncy distribution in the ratio:

(al 4: 1

(b) 4: 5

(e) 3: I

(d) 3: 2

Q. 78 The' f,rst quartile divides a frequency distribution in the ratio:

(II) 4: I

(b) I: 4

(0) 3: I

(d) I: 3

Q.,79 Fourth octile divides a frequency distribution in the ratio:

(~) 1: I

(b) I: 2

(e) I : 3

(d) I: 4

Q. 80 The (inn qullrlile i. also ,k:nOWQ as: (.) median

(b) lower quartile (c) mode

(d) third decile

Q. 81 The third quartile is .l!o called: (a) lower quartile

(b) median

(c) mode

(d) upper- quartile

Q. 82 In case o{weightlld mean, the ac 'CUJ1II..j' 01' utility of' the, mean:

(a) decreases

(b) increases

(c,) i. unnffected

(d) none of the, above,

Q. 83 In certain situations weighted mean and un weighted mean can:

(a) be equal

(b) never be equal (e) both (al and (b)

(d) neither (a) Md no, (b)

Q, 84 Weigh u:d mean is more accurate if we use: (a) estimated weighLs

53

(b) arbitrary weights (e) ,cal weights

(d) any of the abo ve

Q. 8.5 IF .3, .s, ,S, ,7, 3Dd 1.5 are the mspectlve weights of the value' 10, 15, 20. 25' and 30. then the weighted mCIIII is:

(a) 20,0

(b) 23,42

(e) 16.58

(d) none of tI", above

Q. 86 If for a discrete series, the assumed moan A", 50. 1:/<if '" 45 f"r<if '" x -A. 1:/'" 12. the" the mean of !he series is:

(a) 46,25

(b) 7.92

(e) 49.17

(d) 53.75

Q. 87 The mean "f the following discrete series (frequency distribution),

x: 7.12.16,22,25

t : 4,5. a, 3. :1.

is:

(a) 16.40 (b) 15.09 (0) 20.80

(d) none of the three

Q. 88 If for an individual series. assumed mean, A ~ 25. L '"_ ~ - 21 fo. <if ~ X - A and N ~ 7 then tbe mean of the series is:

(a) 20

(b) 21

(c,] 22

(dl 6,57

Q.89 Given the follt)"wing less than lype frequency dlstribanon of income per month,

Incom« (Rs'.) Ie ss than No. o/I'.rsons

1500 [00

1250 80

1000 70

7.50 55

500 32

250 12

the median class of income is:

(a) 750·1000

U rhcberrec htl lch gcsc h utztes M aleri"

(b) 1000-1250 (e) 250-500 (d) 500.750

Q. 90 For the d istribu lion gi "eo in quesuon 89. the modal class is:

(~) 250.500

(b) 500-750

(e) 15Q...1000

(d) none or II", above

Q. 91 For the distributlon given in qoesnon 89. ihe upper qoertile clsss is:

(n) 51)(1.150

(b) 151l-Jooo

(e) 1000-1250

(d) 1250-1500

Q. 92 For the distribution given in question 89. the sixth decile class is:

(.)500-750

(b) 750-1000

(c) 1000-1250

(d) 1250- 1.500

Q.93 r"Or the-distribution given in question 89. 30th percentile class is:

(a) 250-500

(b) SOO- 75{1

(e) 750-1.000

(d) none of the above

Q. 94 Given the fDliowinJl frequency distribution of income of employees,

lncome lU./mOlllh

0-2$0 250-500 500·750 15Q..J,OOO 1.000-1.250 1.250-1.500

No, 0/ emploJ~~s 12

20

23

15

]0

2{)

The median tneome of empl,oyees is: (al 625.00

(b) 760.00

(e) 695.6-5

(d) nunc of the above

Q.95 FOr the frequency distribution given i. quo .. ion 94. the mode of tbe distribution, is: (a) 23

PRoGiIAMI-IEO STA nsncs

(b) 6(;6,66 (e) 513.33 (d) 56U8

Q. 96 For the grouped data given in question 94. the thi rd quartile is:

(a) 1.200

(b) 1.125

(e) 1.183.33

(d) none of the above

Q. 97 Sixfh decile of the connaaous freq lIency d •• tribution given in question 94 is:

(a) 833.33

(b) 1,000.00

(e) 70.00

(d) none of tho above

Q. 98 2O'I'perecntiio of Ill<: grouped data given in question 94 is:

(a) 250

(b) 350

(e) 500

(d) 550

Q.99 The more than type frequency distribulion of nge of tho patients . on a pru1icul~ dayin a hospital is M given beIDW.

Age (yeors) No, oj pa/;ems

,. 10 152

,. 20 128

,. 30 1.13

,. 40 77

,. 50 ,.60 >70

.nd up [0 80

the SOIh percentile class is:

(a) 30-40

(b) 40·50

(e) 50~60

(d) non. of lile above

Q. 100 For the more than lype distribution ~jven in question 99. the modal class is:

(a) 30-40

(b) 40.50

(e) 50-60

(d) none of Ill<: above

36 22 5

U rhcberrec htl lch gcsc h (jutes M ate,i a

ME.'oSURES OF CENlWo.!. TENOENCY

Q. 101 For the more 1~'Q type distribulion given in qu es t ion 99. the 1i"'1 quartile class is,

(a) 3(l.40

(Ii) .4(l.50

(e) 50·60

(d) none of the above

Q. 102 In the nseending seri.es of qucsdon 99. fourth

decile class is: .

(0) 3(l..40

(b) 4(l.50

(e) 5(l.60

(d) none crr lbe "rove

Q. 1.03 The mode oflhe more than type di,tribulion given in question 99 i.,

(n) 31.78

(b) .41.56

(c) 42.56

(d) none ofthe above

Q. 104 TIl .• median (second' quartile) of ~ more than type distribution given in quesiion 99

(a) 37.33 (b) 40.00 (e) 40.14 (d) 50.00

Q. 105 The seventh decile of the cumulatiVe distri· bution of question 99 is:

(a) 42 .. :>.4

(b) 4S:78

(e) 57.66

(d) 41.66

Q.I06 Thirtieth percentile of the more Ihan type distribution of quesnon 99 is:

(3) 31.83

(b) 53.50

(e) 30.S8

(d) Done of the IIbove

Q. 107 The upper quartile of the distribution given in question 99 is:

(n) SO.OO

(b) 49.51

(c) 45.00

(d) .41.00

Q. 108 If we ptot the more Ihnn type """ less lh."

55

type frequency distributions ohlle same set of d.1Il, their graph. ;nte~. at the poin! which i. known 11$'

(a) mediun

(b) mode

(e) mean

(d) none of the above.

Q. 109 Mean of a set of values is bucd On: (a) all values

(b) SO per <=1 values (e) lim and last. value

(d) malimum and minimum valu.

Q. 110 Which. mean i. most affected by extreme values?

(I) Geometric mean (b) liannonic!n<3!l (e) Arilhmctic mean (d) Tritnn'led_

Q. 111 The measure of centralYai uc which C&IInot be cokulaled wilb open elld classes in case of grouped data is:

(a) median

(b) arithmetic meat! Co) mode

(d) thin:! quartile

Q. 112 Harmonic mean gives more weightqe 10: (a) .mall values

(b) large valueo;

(e) posi.tive YallI<$

(d) negaIive vaIucs

Q. UJ Harm""ic mean 1;_ le$$ weighta£c 10: (0) positive vaI\1C5

(b) ncgative values

(c) .mall values

(d) II~ values

Q. 114 for ["""er algebraic lleaIJIIeIII geometric meI!I is:

(a) suiral;lle

(b) nol,uilablc

(e) sontelimes suitable (d) none of the above

Q. 115 Fo, further nl&eliraic treatment harmonic mean is:

(.) $IIit>.ble

U rheberrcc htl lch gesc h (jutes M ateri"

56

(b) nor suitable

(e) 50melimu suitable (d) none of the above

Q. 116 For a highly variable series. the rnosl",ltable mean is:

(aJ arithmetic mean {bJ geometric mean (c) barmonic mean (d) none of !be above

Q. 117 Thepercentege of ""lues in a set of values which are less than (rno", than) thc median value is:

(a) 100 per ee nt (b) 7S per ceol (c) 50 per cenl (d) 2S per cent

Q. 118 The percen!agC of value:s of a set which is beyond the third qWll1iI<o is:

(n) 100 p<r CM!

(bl 75 per cent

c) ~O per cent

(d) zs per cent

Q. 119 The percentage of data or. set wbich are to !he left of $CVCnth decile u.:

(a) 30 per cent

(b) SO per cent

(c) 70 per cent

Cd) 90 per cent

Q. 120 The percentage of dot:! of a .sel whi.clI is to the right of ninetieth percentile i.,

(a) 0 per cent

(b) 10 per cent

(e) 90 percent

(d) 80 per ce nt

Q. 121 In. class test, 40 students out of SO !"Wed with mean marks 6.0 ODd !he overaU avenge of class marXswas 5.S. The avera,. marks ofstudenla who fuiled ""=:

(a) 2.S

(b) 3.0

(e:) 4.8

(d) '35

Q. lZZ Seven persons gambled .iUi ng on a table, Four persons lost on an average Rs. S~

PROGAAMMEO SrATlSllCS

whereas the. other three gained on an average Ri. 711. Is the information WOM believing? (li) Yes

b) No

(e) Not certain

(d) None of Ibe above

Q. 113 The average marks of secuen A ore 6S and that of section B ore 70. The averas" of both the sections combined is 67. The ralio of number of .<!udenl$ of section A 10 B is·, (a) 1 : 3

(b) 2; 3

(e) 3: 1

(d) 3: 2

Q. 124 Tbe partition value which divide a series into IWO equal. pans is known us:

(4) second qWll1ile

(b) third quint;lc

(e) fourth octile

(d) suth dcciJes

Q. US In a dislribution,!he value around wltich the. items tend to be most heavily com:entruted is called:

(0) mean

(b) median

(0) third quartile (d) mode

Q. 126 Geometric mean can be used to lind out (a) population growth

(b) growlh rate of GNP

(0) beth the above

(d) none of (al and (b)

Q.l%7 Formula for the. combined geometric mean iGM!?,' or 1'NO series of sizes 0. and '" ODd their C.M:. OM, and OM. respectively is:

(al OM,," ",logGM, +"" log OM, ",+""

(b) OM" ~nl x OM, +n, ><OM, nl'+nl

(e) OM" ~ OM, lognj + OM, log"" n, +""

U rhcberrec htl lch gesc h utztes M ateri"

M EASURE:S OF CENTRAL TE:NtlENCV

("IIOgGM1 +)

(d) GMu D Antilog n21ogGM2

"I +":!

Q. 128 Sum of the deviations abo. I mean is: (a) zero

(b) minimum (e) maximum (d) one

Q. 129 Sum of the absolute deviations about median is: .

<a) UfO

(b) maximum

(e) minimum

(d) one Q. 1 JO Sum of square of lb. devietinns abcut mean is:

(a) maximum (b) mirurnurn (e) zero

(d) none of the above

Q. lJl Histogram is useful 10 determine graphicaUy the value of:

(a) mean

(b) median

(e) mod.

(d) all the above

Q. 132 Graphically partition values can be determined with the help oL

(aJ frequency pol y gon

(b) bar diagram

(e) line diagram

(d) ogive curve

Q. 133 The suitable measure of central tendency for qualitative data is:

(a) mode

(b) arithmetic mean (e) geometric mean (d) inedian

Q. 134 Two series baying the same mean, medi an and mode may:

(a) have same values

(b) not have same values (0) both (a) sod (b)

(d) none of <a) and (h)

Q. 135 Wcighted mean gives a higher value than unweighted mean if:

(a) all the items have equal weights

«» larger itents have higher weights and sm all items have lower weighlS.

(e) larger items have II/weT weights and smaller items have higher weights

(d) none of the olio ..

Q. 136 In a frequency distribution with open ends, one cannot find out:

(a) mean

(b) median

(e) mode

(d) all the above

Q. 137 The overage of " obseIvlllionsx I' "'i! •... , x. is M. If xli. "",laced by x , !hen tho J>eW average is:

(a) M-xi +;r'

(b) M-xl+x'

n

(c) (n-OM -X, +x'

"

(d) nM -Xl +.1"' n

Q. 138 In CMe of 15 per cent trimmed mean, the percentage of observations utilised is:

_ (oj IS per cent _

(b) 30 per cent (0) SS per cent

(d) Done of the above

Q. 139 If Ior valu es of X,AM. ~ 25, H.M. ~ 9,th."

theGM. is: '

(a) 17

«» 15

(el 5.83

(d) 16

Q.. 140 The second qll-llrtile of the following .. 1 of dala,O, I, -J. -2, 6, 4, S, 8.12. 10. lJ is: (a) 4

(II) S

(e) 6

(d) 8

Urhebermchtlich 9 schutztes ~la1crial

Q. 141 The measure of 'cerilral teildei1ey which - teniains iiillillet\;d by d,trem;o ~il(irii is:

(8) me:in' (bj mode

(e) haniioTiio mean

• (d) georrietric itI.eati

Q. 141 .i"he mean of !he square. or fll'sl eleven narural numbers is:

(~) 4(\

(b) 23

(e) 48

(d) 42

Q. i43 The poiol of irill!r=tiOQ of I"", ttimb~ti\le frequency curves provides:

(a) !IIellII' -

. (b) miJde

(e) ",.dion

(d) lim quartile

Q. 144 Tho percentage. of ibobis Ib a ~bi:y dimbullon lying lielw=! uppd alid i"""':" quartiles is:

(a) 80 per cenl (Ii) 40 per ceni (e) $0 per <::enl (d) 25 per cent

Q, l.~. A penon has been deputed to find the averige income of ra<W!y eJ!1p1oy~. Th pm.ide a eoiTecl picture ofa~c iiitorlie; he .Ii.oold find out:

(a) geometric meah

(b) welgllted IiiiIari

(el prqj""'l~ itlOaJi (dj"MlthmeiiCilieGil.

Q. i.«; Th know tbe IteM or a ij~ lIietles dill, prefetlible type 6f li"';hIic is:

(a) progctulye Ii'mlile

(h) movihli a~

(c)weiglited hIeah

'," ",(ti) ij\ilict"niilb~.

Q. 14' if We !mow iii", inea;; v.loes liHiiilliblitt of units <if vaiiOllS gtoujiS; Iiiett <iIl1l ffii!ali \IilUo! fOr all 'he K""'iis cal! ~ oIiIalileill!; !indliili 0111:

(0) geometric. mean

(b) Weljllte<l mean (e) pooJtd iri4iI

(d) .proztes.iYe i'rIe4n

Q. 148 iii a frequency di.slflli!irlon ofa IlIrge number of vaiues, iIie m<!de i5:

(al lari:e51 dbsetvatJoh

(b) ",,"Iiesr Yliiue

(eJ IibseMlion ""ili hWhi1um fn;quency (d) maximum frequeilcy lit in "h..,vltion

Ii 1 .. 9 The relation beiw1;en qlladtatic ioeaiI (Q.M.) and arilhmetJc mean. (A,M.) Is:

(aj Q.M. '" A.M.

(b) Q.M. > A.M.

Ie) Q.M. < A.M.

(d) Q.M. *- A.M.

Q. 150 A ,01 of .oJ"o. contains A few neg.ti"" ,,~Iu~., The average of set can better be measured by:

tal arithmetic mean (b) geornemc meIil (el JitogteS5]ve iiitaii (d) quadratic IiIeiiH

ANSWERS

Sicfldit>B

til jijj;ill Griffiii (2) ItA. Fisher (3) AL Bowley (4) typicBi \lallie, stinUiiIirJ InUSIlit (5) cxttemc values (l) YRriate (7) Silililfiarises (8) mean. (9) :;

(10) lli Ili (I i) 13 (Ii) 16 (13) zetn (14) Rs. 6S (IS) IU. 21 {Iii} 8eiiiiilt! (1'7) 10 (is) (il' + 1)12 (19) itiidiihilfti (20j B (2 i) NiliiUj (01(0,> In) zero (23) tauOj jlt<ijIOriIIiils (24) ~ (25) ~ cui a.mge

ljjeeiI (2'1J liiiIiIllY tiS) a,M,,, .JA.M.kH.M. (29) medi;u; (30) mlniiiiuiil dh i~ t:i.2) Ilnlque (jj) iiM!ciian (34) open,;lid ~~)ffibilo: (31;) 1 (37) medlan; liitlde (lSJ iiltidil (39) ili~ (40) Illillii<$l (41) biillildii (42) tI6\ipliii (4j) qiiltllie (44) ~1alI (4S) ,gil, (4~j nl\ii {4t, b; (4~) 26 per cent r4!1i j~t tlt !lilt (!SO) 6i:l'" (' I) ~ 1'& cenl(Sl) 34 jler cent (53) 401ft (54) t1i < ~< 1>10 (SS) fiM (56) ftftji!.lh (!S1)a5tiiidlbi UillillillUiIes (S9) PIO (60) third (61) first; twenty"fihh (6i) fm<:lil es

U rheberrec htl lch ge5c h utztes M ateri a

MEASURES QF G~T~

(63) Ya Lun CliO\! (~) a. "& (~) fJ, <!' fu (66) 75 per cent (67) nll~caIi ~y~"", «(j8) (Ql + Q,)/2 (69) anillog (70) ~tiprot.1 (71) oeg;!live in) H.t.1. ~ Ci.ttI. S A.M. (73) same (74) lIIandard devjation (75) U (76) 7 (77) ~ 1 (78) en!ire d.!:l rig) ~I (80) nol rigidly.

SEcnON-C

(I) a (2) a (3) b (4) b (5) ~ (6) b

(7) b (8)e (~) cOO) d (I L).~ m)~·

(13) b (14) c (1~). W!) d (17) b . O~) a (19)b (W)" (~mc mIll m)b @)~_ (;ZS) d (U). (;21) e as) J ~H ,3(1) c (31) a (32) d (33) c. (34) a (35) II (l6) d (37) c (38) b(39) c (AP) b (41) c (42) d (43) d (44) c (45) d (46) • (47)~. (48) b (49). (SO) b (5-1) ~ (5~) b (53) ~ (54) d (55) b (56) d (57) d (58) d (59) ~ (60) d (61) d (62,) a (63) c (64) b (65) d (66) b (67) a (68) b (69) b (70) c (71) d (72) b (73) a (74) c (15) d (76) c (77) a (78) d (79) a (80) b III I) d (87) b (1l3) 8 (1I4)·c (85) b (86) d (87) b (88) c (89) II (90) b (91)c (?2) II (93h (94) c ~5) d (96) b (97) a (98) b (99) b (00) b O,Qlj d (10],). (103) d (l~) c O()ll)!!- 0(6) a (1Q'ry 11 (IQ8) II (109) a (.liD) c 0 HJ b (I m, (113) <I (114) b nU) b (116) c (117) c (1l8) d .(119) c (120) b (121) d (122) b (123) d (12.4). (125) d (126) c (127) d (128) a (129) c (13<1) b 031) c (132) ~ (133) d (134) c (135) b (I lWrl (131) 4 ((38) ,d

(Q9H (j4Q) b (l41) I> (l4fj ~ (143) c C!44) c 04S) b (146) b (147) c (148) c (149) b (15(1) d

Ei!.l.9Qe!lted Reading

1. Avl"wal, R.J-.., ~~!>~ri.s/lCf,· New I\~l' lolmlJUiona! (Il) LuI, Publ~, l'J!'!¥ .~IJj, 3rd !'<Ia.. 1!l96.

~ rre!l<l. J£,MDlle", EI'/IU!nlqry SlqlU!WS, J'r!ollli!;l'-H!II1 pf lPdi@, New R!:Lhi, 1981.

3, Qo9a, A.IlI., G~pIa, ~.K.. ~. Dasggp!!i, B" . ~ltIals rd SI!JIi.sr~s, yo!. t, tJJe Wurid

~,Calcul!a, 1977. -

4. ~ s.c. 0IId K.Qpoor, '1.1(., FundQltJelllals qf .Malhema#CilI StQlistia, $ultaJlO1!Jo!!d /JI. S0Q8, !"ew I">dltj. ~!It qlQ .• 1983,

$. ~ P.G. a,m:l JC$6!'l), jU., /lq.sic S!qli.sl/rl _ for ill'S;" .... anr! q"nomia;. Ipbp Wjley. NI:!W ygrt, 1982.

~. ldl'l!:<,J •• Wassefman. W. and Whitmore. G.A .• ;'ppiied SlaJulI<$. Allyn 8Qd Bacon. London. 19112,

1. 0st40, ,Ij. •• $lt!lini<:s In S es earth. Ox(ord"

1II!f. Cal!:\l~. L?tiP. . ..

J, SeI~. Ci.~. ~ V:~III, S .. J!:, ~'mtfll(!{Y $tQl!,tI{cs. ~m Pol~ J!iIbI~ipJ;' t(!'W

Y<:>rk. 1,98;1.· . '. .,

f. W~.fI.M: a.JJ4 U\>.I .• !g/vll#ItI!iry S/iUis· . jklli M.I/JP4., JMl ,I!.im:b.n pniI WIIIS!?".

~YorI::, 1969. . .••.. _.' ..

,t, Wi~, i.J!.., !J,gilJ.llln~ ·S{p/#rIcs, _ '!Yl~Ih.op

1'u1>li~~J\'.-I,!116, ,.

;, .

U rhcberrec htl lch gesc h utztes M ateri"

Chapter 5

Measures of Dispersion, -Skewness and Kurtosis

SECTlON·A

Short E •• ay Type Question.

Q. 1 What do you u~taod by dispersion of • set of values?

Au. Expcrierlce. tello Iit_t in ml!lY Aituatioftl. the spread of values ;, different bullheir cenlnll values are .. me.. All the 1'IIOI'e, • central value provides no information abOOtthe ~ng of valUes in a OIl! of <Iota. Hence, =lain me&SU<Ca an: ""01 ved wbich ",O""ton lite ~'" of values in numerieal~.

are blown as meaS\IJeS of dispersion. -

Q •. Z9 Quote the statements about the term dispenicm sjvcn by (0 Reigleman, (;1) W.1. King, (iii) Spiegel. (iv) B .. C. Brookes and w.F.L. Dick, (v) A.L. Bowley.

AIlS. The $Utcments given by various workers on: as follows:

(il RacJeIaan: Dispersion is the e:<tent 10 which Ihe magnilUde or qualilies of the items differ. thaI is the degree of diversilY.

(ii) W.I. Kin,: The term dispersion is used to indicate the facts litar within a given group, the ile"'" differ from another in size or in Dthctwords th"", is • lack of uniformity in their aize.

(iii; Spleph The degree 10 which numerical data tend to S)n"Id about an avenge value is called the variation or dispersion of data.

(iv) ac, Bmokto &lld W.F.L. DidI:, Dispersion or spread is the degree or the scatter or variaIion of the variable about a central value.

(v) A.L 1IowIe7: Dispersioo is lite measure of

the variation of the items.

Q. 3 Name dilfen:nt measures of dispersion. Am. Following am the dif[ernn, measures of dispersion,

(I) range, (U) Interquartile range and quartile deviation, (iii) Meau deviation. (iv) .Median absolute deviation, (v) Variance (vi) SlaOdard deviation, and (v·jjl CoeffICient of variation.

Q. 4 In what manner. John I. Grimn and, C.T. aork and L:L. SchkBde, defined meas-ures of dispersion,

AIlS. The definitions given by the statisticians named in the qllCSlion am lIS given helow:

,J oItu L Grif!I .. , A measure of variation or di'pcr;ion descrlbes !he degree of scatter shnwn by the oo se r-

U rheberrcc htl lch gesc h utztes M ateri"

MEA.SU.REll OF O1SPERSION, SI<E¥mEss AND KtHnO$IS

vations IUId is u,ually II1ClI$1II'ed as lUI average devia· lion about someeentral valu~()!" by any orderSlJUistic. Clark and Schbde: M.ell5ure. of d;"permoD aJe m.e3SU>e5 of scatter about lUI """"'8".

Q. 5 What are ""Iui$ites of a good measure of di.5p<nion?

All$. Main requ isltes of lUI idul measure of dispersion can be given a. follows:

(i) It mould be based On all the observations, (ii) Its unit should be same as the unit of

measurement of items.

(iii) II mould be rigidly defined,

(iv) It should follow gencnl rules ofmathemalics. (v) It:should not be subjected 10 complicated and

iediollll cakulauoes,

Q. 6 What are the uses of dispersion.? All$. Main uses of dispersion are:

(i) II tell. about !lie reliability of • measure of central value.

(il) II make. possible to compare tWO series of data in respcc! of !beir vuilbi1ily.

(iii) Measure of dispcl'"Sion provides the basis for the eontml of vari.abUity.

(iv) U has Q wide application in almost all lields

of smti.tic..

Q. 7 Define range.

Ans. The difference between the large.1 and smallest values of aiel of data is called its nlDge. Ronge is mown as lowest value-l~ value.

Q. 8 Give. the merits of range,

Ans. Merits of range are:

(i.J II is the easiest. measure o.f d ispcrsicn.

(ii) It can alway. be fouod out viSW!lly l e., it involves no caleulatiens.

(ii i'J II ls one of the largely used measure of dispersion.

Q. 9 Give dements of range, AIlS- Demerits of range are:

(il It depends on two extreme values of a series.

Thus, il give. no information about rhe

obsc""ations lying between .mllll""t and largest values.

(iil II is hi&hly susceplible to sampling floctuations.

(iii) II i. nOl .• uitable for funher malhcmatical treatment.

(i.v) Addition or deletion of • si ogle value may

change !be e.1tire complex of "Lnge.

Q. 10 Define coefficient of range,

Alii. It IS a pure number: given as llI.e ratio of difference betwe.:a lbe lOUJe$t and smallesl values 10 !be sum of !be largest and smallest values of a set of data, Numerically, coefficient of range is (l..- S)I if.. + $). lesser the coefficient of range, bl;uer it is,

Q. 11 E~press jntuq.~le (I.Q.) range.

AIlS. It is equal 10 tit. liiffmace between 1M_Upper and lo_. quanile$. Symbolically, it ;.,. equal 10 (Q,- Q,l. This = of dispeISion tells about tbe ranie of the mid<lle ~O per ceOI values of • set of data. In Ibis me .. u~ lower 25 per cent and upper 25 per cent val~es are. e~cluded. It ls pot • good measure of dlspetsion as it tell. nothing about 1M dispersion of valuei around .verage_lt hardly fulfils any of the ""IuWtes of a good measu~. of dispi: r$1011.

Q. 12 Defioe percentile .... ge.

A.... The di fferencc between 90th pen:entile V' to> and 10th percentile (P,,) is caIled percentile nmge. It i. denoted IS PO() - P 10" lis vallie" Is same Us D.- D I where D I and D. denote 9th and lsI decile •. This measure is IMn: useful in Cdocalion.

Q. 13 Define quanile deYiaUon(Q .. D.) WId ghoe lis Important features.

A.... II is half of the interquartile range. i.e; (Q, - Q,)12. It is lUI ebsolute measure of dispersion. Hence, to compare two series, a relative meuure known as coefficient !If quartile devialion;" given wh.icb is 5ymbol~UY expressed as

(Q, - Q,)/(Q, +Q,).

Q. 14 For a symmetrical distiibotion,how Can' One determine 1M upper and lower quartll es wilb the help of quarti.le deviation?

U rheberrec htl lch ge5c h utztes M ateri"

62

AnI. For a .ymmetrical.dillribl!li(H!,!be 1!PP"f and lowel "!WIiles CAA be detenniaed by .Ihc fonnula (Ql +Q. D.) and (Ql- Q.D.) IQlpec6vcly.

Q, I S W1I~1 ano IlIe good points of quartile deviation?

Ans.. Followi ng are tile good poiall of quam1e dIoviotion:

(i) It is easy 10 calculaJe aodulldetstatld.

(ii) It can be cnIculaIed ia. C&Se .,of open ef!d ft-equeocy disaibutiollr .. WdI.

(iil) I! is nO! affected by 2S per Cenl upper arid 2S

per eent I.,.,...,. c.<1mnC. Yiilue..

Q.16 Throw lighl OQ'~,~i..lion (M.D.). A.... Range aod quartillO ,o!F<ia!io<!t _ pcIIilionaI .....,_ of dispersioo. 'wba-ca, .~. deviation ia a measure of dispmioa which iii bPCd O!I all'valuc:s . of • set of data. 11 if iktined .. !be .vena" of !he absolute deviaiiDGI lua..rlQll! ., '_"liIuaIly, tlic mcu, median or.1I'IOde. It is oauaIly ~ by 1;. To elmfy''''berber !hoi .__",UJed in me8I1 deviation is mcaa, median Or IIIDde, • sUms i.

~ 10& such ... 6.,.5 ... «11 ... ,::

iThe fonnull.los..QkWalja, ","".lev. ;fn obsenatJc:io. X,i, ~ ,r, X.;., ' ..

:Ii = !.LIX,:-~ .

n'i •

for. i ~ J,i ~.,: «:

AlSo A may be aay ~ _ out· or _. modi"" and mode.

For a f~uet!Cy c&IribuIiOll ill .which the variate . ~Iue x, occun!, limes (;" J; 2 .... , k).1he fPfl!lllJa

for mean dc~iaJion is. .. .

'. ~

Here. il is worth ClllJlhuising !hat mean deviation ;. for minimum ~bout the mcdiaa. That iii wily IIICIdiIII is ~ly. used."" aa ~c.vaI!le abouI. ~.Ihe mean devillion is ealcula&ed.

Q. 17 Discuss coeffrcieDI or ._deviation.

,PRoGRAMMED STATlsncs

AIIIL .~ deviation bas the same unit of raeasurement ..: thai of the variable x. If fwo series have diffen:nt units of measulUllen!, the seri es cannot be Compared. Hence foe comparing any IWO series, an unilless 11IeIISURO. is given known as co.jJicienr of /IUIJJI amQtion. II is the ralio of mean deviation 10 the averall" 'A: used in calculating it. l!s formula il! -,

Mean deviation Cocf[.ofM.D,= . 'A .•. J< 100

II i. mul~plicd by 100 10 express coeffici en ! of mean deviation in percentage.

Q. II Write Ibe meriu of mean. de. lation,

AM. Following are tbe merits of mean d~ iall.on: (. Ii uti!iso.s all the obscNatiORS of the set

2. II is simple (Q calculale and undentand .

3. II is least afftcrcd by eslreme values.

Q. 19 Mention the demerits '~f mean deviation. AM. Following are Ibe demerits of mean deviarion:

I. The fiRm"'! weakneA of mean deviation. i. , thalin iu calculAlion IICpliVC diffen:tH:C.i are ,c(laoidered positive without MY sound

_ing.

2. II is nOl atneruIblc 10 funber algebraic uearmenL

3, II cannot be calculated is case of open end(')

frequency di.lribulion.

Q.20. Define IIIld dilCUSS variance ..

Am. The ·.verage of the square. of the d~lalion. taker! rrom' meari i. ul1ed variance. The populalion VIIIiaacc ia ,.nerally dcMIed. by (II and its estimate ( ..... pte ".nance) by r, For N populalio.n valu"" XI' X1, ... , XII having.lhepopulalion m ea n ]1, Ibe populalion varillllC<:.

where

;-I,,2, ... ,N p",r.IXdN

An. estimate of .". based on II .sample. vIII ues x" -'1' ... , x •• the ..... ple variance,

U rheberrec htl lch ge5c h utztes M ateri"

MEASUR~ OF DISPERSION, SI<EW!ESS AHO KUImlSlS

for and

s ,,,_l_~('<i_.i)' I!-I~

j" [,:L ... , n ;"I:.,xdn.

For a frequency distribution of umpJc, values xl' .<" .... x. having fn:quencie" fi, 11' .... t. respectively. The sample variance,

.' --I-Eli(XI -.t)' n-I· i

1 { z (l:J;'<')'}'

,,-- Lli'<. --'--

n-l, n

where

for i" 1,2, .... k

Q. 2 t Discuss merits and demerits of variance. AIlS. II posse sses aU tbe requisites of a good f)le3SU(I! of dispersion ~cept that its unit is square of the unil of measurement of variate values. Hence. many times it becom es difficult to ""tuclly adjudge the magn itude of vari3lion. Atso variarK:e is .en.ili vo ui extreme values. Vuriance ts the backbone of st:nisllc,.

Q, 22 Mean deviation is minimum while Ulking the deviations (rom median. Then wily do we take sum of square of deviation. from mean while calculating the variance?

Ana. Deviations from mean are used for vuriWlce because SIlQ\ of .squ;u-e.o of deviation. from mean 10 minimum,

Q, 23 How docs vari.oec i s a ffooted by the change of origin and scale?

Ana. Vuriance is indePendent of change of origin, but is affected by the change of scale. If e,lICh ob5e,..,.ation of a dwa ser is multiplied by r, then the variance of new data set, is c> limes the variance of ori2inal data.

Q. U What is the difference betwee!l absolute and relarlve dispersion?

AIlS. If tile unit of • mc.uure, of dispersion is in . SIlI!IC terms as thaI of the observations of • series, it

63

is called absoluu me<lSl<'" of dlsptf;fiolt" e.g., beigbt in em, weight in k;: .. income in rupees, etc. In this C..Q&e two series" having different units or dispersion, cannOl, be compared, Hepce, roc comparison of IWO series having diffetem units of measure"",n!, one requires an unit less measure of dispersion. Such measures are termed as retatlv« measu",~ or cotffident of di.penion. For tIIis, • measure of dispersion 10 usually divided by mean used in its calculation 'And multi.plied by 100. Thi. provide. the m eas ure expressed in percentage which i. fil for comparison or any two or more series.

Q. 25 Define and describe in brief stllldard dev,iati.on.

AIlS. The positive square root of the variance is called standard deviation. TIle· idea of standard devja~oo was firs! given by Ka,rl Pearsoo in 1893.

Symbolically,

and

It fulfil. .11 tile, requisites of a good measure of dispersioo except th;u. it i. sensitive. !O C1treme val ..... That is why it j. known as standard deviation,

Q. 26 Explain ,median "bsolute deviation (MAD). A.... Slandanl deviation is .fftctcd by extreme values. Hence, the median absol ute dev lailon i, an altcro:itive measure efdispersion, Median absolute deviation is defined as the median of !he absolute deviation taken from median. It is seldom used as it is not eas i[y amenable to further ,dgebr,aic treatment, Moreover, it is not involved in distribution function.

The forrnuln for median absolute devialion ii,

M_AD - Modi"" lX, - X ... "I

Q. Z1 Comment on the variances of median and sample mean based on a sample of sill! n from a normal population having variance 0-2.

ADS. Vuriance of sample median for $alDples of size n from a normal population i. 11 .,' /210 "'h ich is greater titan !he vorilll>Ce of sample mcaII. i which is equal to ,,' / n,

U rheberrcc htl lch gesc h (jutes Iv! ate';"

Q. 18 How .ariability is indieatcd by a measure (If dispersion?

A_ Lesser thevalue of a measure of dispersion, more is the degree of nearness of observations. It is also an ,i ndicator of homogeneity of values of a series.

Q.l9 Wh.t are the different names given to stllndnr:d deviation?

AI1$. Standard deviation i. a[so known as m."" error. mean .r;qu!Jre. ef'ror or TOol mean ,Iqll.an: dtvil1lion from nuDn.

Q. 31) What is the efrect of adding or subtraeJing (I constant 'c' from each observation of a set on variance or sIDndord deviation?

Ans.The variance or sIandard deviation remains !he same as they an! illdepoodent of change of origin. Q. J I Standard deviation is eq .ivaleat to wbot Iype of mean'!

AIlS. As a matter of fact, standard deviation is nothing but !he q-uadraric mean of the deviations from data mean.

Q. Jl Why ore m.ea" deviation (M.D.) and standard deviation (S.D') nat calculated using the same average?

ADS. Mea" deviations and standard devi ation are not calculated using !he same average because M.D. is minimum when tbe deviations are taken from median whereas mean square deviation is least when deviatioos are measared from mean.

Q.3,3 Give the fornIula for stllJldard deviation of a population of fitl;1 II n.tum! numbers,

Ans. The, formull! for standard deviation (a) of 1i",1 n natural numbers i.

Q. 34 How is variance affected if each cbservatien of. ""I i. divided b,Y a cnnstant 'd',

AIlS. The variance of the new set of values will be

I .

d' times: the original value of variance.

Q.. 3S How is van.""e affected if each observation of a setIs mulliplied by 'd'.

All$. Thevariancc of the transformed set of 'values will be tP times the variance of the orighmlset of values.

Q. 36 What is coefficient of variation and its: importance?

ADS. Coefficient of variation (C. V.l is the rati~ of the standard deviation and the mean. Usulilly ;!-is .... pressed in percentage .. The, formula for or>efficieot of varia.tion is,

C.V.~~xlOO.

mean

It is a reIM;"" measure and is most suiuble to com pare any two series. As we know, the si .... of measure of dispersion also depends on the s ze of measurement. Hence, iI, is very 8pIlroprialc measure of dispersion to compare two serie$ which differ I;ug.ely in respect of their means. All the more, a series or a set of val ues having lesser coefficient of variation.., compared 10 the Dther i, 111<1'" consi&tent Q. 37 How can the variance of two groups or series of data can be combined (pooled)?

A.... Suppose or, CI~ :XI, X, u.rd s; N, ..., the variances, m •• ns and ,izes of two groups of values ",spectively. Also let Xu be th.eir combined mean, The combined variance, of two groups is given by the formula.

Nt (a: +dn+N, (ai+4)

c.

NI+N,

where" XI -XI' »dl and X, _- XI' '='d,. The for· mula for combined variance, can be .,tended to any numbel" of groups.

Q. J8 What is the advantage of combined va,-llUlCC? ADS. Many times we know the means and variances of indi vidual series or groups of data of known size s, Then for some stotisticalanalysis, lhcir pooled variance is required, By the fonnul. for pooled

U rheberrcc htl lch gesc h utztes Male,;"

Mc.o.sullES 01' DISPERSION. SKE'M<ESS AND KuRTOSIS

variance, il can e as ily be obt:lined withoul original dala. Also a 101 of ume nnd labour is saved.

Q. 39 Define and disc uss moments in brief.

Ans. ,IIi moment may be defined as the avemge of the ,m ,,'ponenl of the deviations of the varillte values of a series aboul mean. III moment about an arbitrary mean 'a' is given as.

for

I', ~ E(x - a)' r" I, 2, .... r

where. E stands for expectadon, Moth.matical expectati On bas been discussed in chapter 7. Read it from there. Fur a series having N values X,. x, ..... X II' the ,.. moment about 8JI arbitrary origin '0' is given as.

for

II~ B ~ ~)X; -a)'

;e J. 2, .... N.

If 0" 0, 'the momenl is known as ,.. raw (simple) moment lind IS given as

I,", II~ ~NL..X,

Again, if " " 11. the ". mom, en L i. known as central (ebsoluie) momCIII and its fonu,ula is.

h i~ int,~ling, to re ... ealthat an alg~brL\ic relation between II;. 11; and 1', can easily be established. II helps in determining one when others urn known. Q. 40 Wb,at are the uses and importance of moments.

AIlS. Each f,equency diotribulion is specified by its moments especially the first and second moments. They also help in determtning the shape of • distribulion since skewness and kUl10.is Ire usually measured with the help of moments. Mean and variance arc nothing, but fim nnd second moments respccti><ely. In practice, moments of order higher 1han fourth are rarely required.

Q. 41 What ls kn.own as probable eITOr?

Ans. Tw,Hhirds of the standard dev;.uon is known

65

as probable error, It is alsoequrulo quartile deviation. Symbolically,

2 P.E.~ '3"

=Q.D.

Q. 42 Wh.t does empirical relation. exist between quartile deviation. mean deviation and standard dcviaLion?

AIlS. Relauon between Q.D .• M.D. and S.D. i s,

5 2

Q.D. ~ 6M.D.~ '3S.D.

or 6 Q.D, " 5 M.D. = 4 S.D.

Q. 43 Give an empirical relation between range, standard deviation, quartile deviation and mean deviation,

AIts. The empirical relation between range, standard deviation. quanile deviation and mean devialion i, OS follows:

15

R·~ 6S. D.~ 9 Q.D. c"2 M.D.

Q. 44 Gi><e re latio ... of the fint fOlU' central moments with the raw moments.

A.... 'The, relatio ns of the first four moments ].II' ~'Il:! and II, with the raw moments ].Ii .11;.].1; and J.L~ are:

III =0

].I, =].1, -].Ii'

].I, = 11; -31L;)li +2}1i'

].I. = Ii. -4).1;).Ii +611·i IIi' -311;'

Q. 45 Write lite relations of the fmt four moments about the arbitrary mean '11' and raw moments.

ADs. Four moments about the arbitrary mean '0' i.e .• in terms of raW moments can be expressed as:

111'=111-0=].1-0 II~ = 11; -21l11; +,,1

II~ "11, -3a 112 +3a'II; -0'

",i :;;;; III ~4D 1-';. +6a.l'p~, ~4a] poi +a1

U rhcberrec htlich gesc h (jutes M ate,i a

66

Q. 46 What will be the clJ~t of coding (c)I'IJ)ge of orillin WId . ..:ale) of dala on moments?

A.... Let a COIWant vaiije c is subll"acu:d from each variate val ue and then divided by h, L~., the

x-«

coded value J ~ -h-' Then. tile raw mamenlS can

be detconincd by the foLlowi~g relations,

~i,"(*~M,)d

~\_(~~fAl)"h2

~l"(~~M')d' p~ ~(*~M.'Jd'

wbe",. there exists .k varilllc val .... Xl' X., ... ,X, having frequencies I" Il •...• I, respectively and for I = l , 2 _'., k. Also r..t; = N.

T'he. retalions .how tIIallhett: is 110 cffcet of ~311g. of origin. on the moments but the scaI. factor .ppcan in mutriplication if e;ICb vahle· .aft ersu btr.lcuog t .is divided by h. Once. we know the raw IDOmetI.lS. we can fmel the central moments using the relalion. beiween two types of moments.

Q • .,WhOI are the speci3ladvl!I!flIges of standard 4eviatiOfi ?

.us. Th. spcci.lllClvantage;o of$lllPdard deviation are,

(i) Standard deviation carries greal imponano:e in .ampllng melhods. .

(ii) II is lea.s!SC!ISitive to sampUng f1uclu;ations. (iii) With the help of .• t3lldard deviation. il is possible to ascertain the area under the normal curve,

(iv) It hIlS gre4!uliliry in leWoe; of bypolbcses which ocher measures D{ di$per.ion hardly do.

Q. 48 What is meant by skewness?

Ans. lack of symmetrY of tails (about mean) of a

frequency distribution cUfVl'i~ ImP"'.D ~ S~I'W~. Symmcu-y of lui!, mean' thlll ~ fn:gueQ~y .Df ~. points at equJII distaoces on b!l!!! &l~',of I!i!o ~!L1nl of the Curve OD X-axis i .• ~. Also, the .'11'1'3. ~I!'lOf the curve at equidi.,ant inletYa" on. bIl!Il sides of the centre is also equal. i)q>,artl)re fWIIl J)'mmettY 1._ to skewDCSO. h. i. adjudged by !lle eJOIlpIiQn of !he righl and left uti). of the curve.

Q. 49 Witat is po$itiva .1JId DCJati~eslq:WIICSI!? A.... If the left tail of the freq\lUll<)! curve j s. 1JlO~ elongated than righl util, il is bt"""" as ne,Jl,iye skcw~ and a .revetSC si.uatiou ~ 1.0 positive dewDCSS.

Q •. 511 Wllal purpose is .erved by m~urinJ skew.oess?

~ ~lIJ'eofSkewp~ iJ)dical,es)o wl!31~~JIl and in wluUdi~tion Ib.e diAril>\ltiPn of a v4ri~1t di{f ers from synll1ll'iJ'y of ~ fre:qIl!'DI'Y c=. ''PIll curve may haVe positive or p~,;I1i .. e .t<ewneu. J,Ioth posinvc and oegaliv,e skewness caa never occur simultaneously.

Q. 51 What cbanges occur in the positiWl ofmC&ll. median and mode in case of positive and oega!ive skewness?

~ In a negative skew Cl,lry~, ,tlt!: C)lqJI !!lid median are pulled 10 the left )".~ in ~ posj!ive skew curve. the mean andmcdian IU'C pulled .0 the right Also il should be Icepl in mind tIun median m w:lYs lies in between mun and mode.

Q. "52 How can we know abolll skel!'llCllS?

A.... Sk~'WIIC$' can be known by two me.d\od$ given below,

(il Graphically

(ii) By IIUlthematical me3S\l1CS.

Q. 5;3- Dfscribe graplLical method of !lel~!iog skewness,

AI>s. Draw a (roque",,>, cun" by plottinjl the poiPl$ fDr v_ari~ value.s WId .com:spon4ing freljlleD!;il:s. Judge by n;Jked "YI'S wIIdI!I:t .the, Ilills .or.~ .!'.I!WC arc symmetrical Or ROI, If not ,yl))II}e(rk:lll.1bc CJltYe is skew. Po&itivc .kewnessdcpcn<ls,-whedtcr Ibc rilhl Iail i,more clongaJed than left \iiI. Por~~ye .sk.ewheu~ a reverse $ilUalio.1) t:ll$lS. .

Gr1!phlcm ,",,!hod is jus. enough 10 btow ~t

U rheberrec htl lch ge5c h utztes M ateri"

skewness. aut it doc, nOI give the idea about th" extent of skewness. Funber slight skew ness can not be detected by naked eye"

Q. 54 Give different formulae for rneasuelng skewness: a3,

A.... Different formulae for meowring skewness nn:!.;

(i] II owley's formula for measuring skewness in terms of quartiles is:

. Q,+QI-2Qz

a_,.:::: . Q;\~QI

In many books measure of skewness is denoted by J or S,.

(ii) ,]{eliey go,,," tbe formula in term. of percentiles 8IId deciles,

Ke.lley·s absolute measures of skewness are,

S, ~ p .. + PIO -2P",

~ D., + 0. -2D,

These formulae are not pr.nically used. lnstead, lt is measured as coemole"! nf skewness whleh is given as.

S _ p .. +1';0 -2P",

• POll-flo

D., + DI-2D,

. ..

D., - DI

Kelley', formulae are seldom used, (ill) Karl l'ea_tsOn's measure of ,kewn ess,

ITICIIn-mode ct_,,--

S.D.

(lv) Karl Pearson', formula for a wide class of fn·,wooey distributiol1$ in terms of moments is,

, 11 -1',

I ~ I'l

11, giYC, only the measure of skewness but not the directinn of skewness. So anotber measure.11 is defined as.

YI-5.-~-a, I't

Q.55 How do you interprel the value of measure of skewness?

A.... If the measure of skewness is zero. it means that the frequency curve is symmetrical. Hence. for a sy rnmerrical curve p, "a, =r. ,,0 wIlieh i rnpljes lhat QJ + Qj "2(h; p .... + 1';0'" 2P",; D.,'" 0, - 2D,; mean _ median" mode. Atso, f.I" '" 0,

Again. if the va lue of S,. Y, or a, is positive, it reads 10 positive skewness, In Ihis shuatlon, !he frequency curve has elongated rightl.il. A negative value of S" r, or a, leads 1.0 negative skewn ess and tile frequency curve hIlS " long wI on the lef! as compared 10 Ihe right tail, Greater the nIlIgniludc of St~ '1'.1, or ttJ~ mnre is the sl~C$.$'.

Q. S6 For" moderately skew dlstribution. whot ",!.tion between mean. median and made e~i$ts? A.... For a moderately skew distribulion. lite reliltion between mean, median and mode ls,

Mean - Mode" 3 (Mea" - Median) Q. 51 Wltal do you undersund by Kunosi.?

A.... Kunosi, mea ns bulginess, Kunosis ",lolo'!O the peaked ness of a frequency curve as compered to • normally peaked curve" If " frequency distribution curve is more peaked or flat than a normally peaked curve the" il is called I K~rtic ':u",e, This property of bulginess is called /(urto,i3.

If II freqllC""y curve i, more peaked than normel then it is c.lled a l.probnic curve MId if 11.;1 less peaked than normal. lt is called plat}*·unic curv e. In terms ofKunosi s, a normally peaked curve i. known illS fIlesobrlic curve.

Q. S8 Kunosis is adjudged around which measure of cCJ!lnIllendtne.y?

Am:. Kunosl. i" adjudged around mode of the frequency distribution.

Q. 59 How can one bow .boul Kurtosis?

An.. Kurtosi, can be perceived simply by looking a rrequency distribution curve. But such a perception becomes difficult ,flhe curve i •• Iightly kunk

Th overcome thi., dimC\1lty of subjeClive judgement. it i. mathem"tlcoUy measured lIS the ratio of fOUJIh moment to tbe squ",. of the second moment.

U rheberrec htl lch ge5c h utztes M ateri a

68

Symbolically,

~,= 11;

).I,

Q.60 How to ascertain Kuno.is with the help of Il,?

AIlS. If the value of II, i. more than 3, the curves is lcprokuruc and if less than 3. the curve is platykun ie, Rlr a mesokurtic .U!\'C • .a, = 3.

Q. 61 What are .... ndard rneasu res of skewness and kurtosis given by Karl PCJlf$Qo?

Ans. Karl Pearson gave, two convenient quantities as gamma measures defined as

'l'1~../P. =, 11,;, 111

" 11, -311l

'1', =1',-3 ~ --,-

112

If 1, is negative, the frequency curve is ".:gati vely skew and if positive, the curve possossespo.itive skew ness, Whereas fl, gives only 'he magnitude but not direction.

Again, if V, :> 0, the eurve is leptokunic, ifl, <: 0 the curve i. plalykurtic. When Y, " O. the curve is mesokurtlc or we can say, there is no Kurtosis.

Q. 62 What is the purpose of mOIlS"ri., skewness and Kur\osi&?

and

Ans. 1'wo measures namely, skewness and Kunesis determine !be shape of the frequency CU!\'C which obviously ",fleelS on the 'YP" of distribution.

Q. 63 What is the role of averages, measures of dispersion. ,\:ewne$S and KUrIO$i.?

AI)$. Avenges, measures of dispenion. skewness and Kurtosis are CQ,mplementary '0 each other in underSlanding frequency di<lri betinn,

Q. 64 'Why melln devialion and slIIodard deviation are not calculated using the same, avcrnge?

Ans.. The same: average i~ not U$W in calculnting IIB:.1n dcvi,ation and slnndard deviation because mean de,viation i. minimum ahout median and standard deviation is minimum about mean as the sum of "'I uares of '.he dev ial ions from me." is mi n imum.

p!{OGRMtIlEO STATlSTICS

Q. 6S Wbat is Sb"J1P'1l'l'. correction?

ADS. W.p. Sheppard pointed out lit., in case or continuous frequency distributions at the time of calculating moments.u is presumed that frequenci es are centred at the mid-points of the class intervals. Sitch a presumption i"trodu ces some, error in the coleulatio. of moments. Hence, he suUe51ed some corrections i.n various moments, These: carrections are known as Sheppard's eorreenon, They are as follows:

Corrected 11, = 11, (no correcuonneededj

/'

11 •. ' .. 11, - T2 (/ is til. class inl.rwJ)

Currected

I-lJ = III (no correction needed)

[' 7.

1I·~I1'-2"'+240f

Corrections for higher order moment. than fourth con be seen inn ,,,",,book ; f needed.

Q. 66 Who, are iiI"$! and second "<>efficient or

Corrected

A.... The ctICff",icQlS of skewness <1, ~ ../P. ~ y, are lite firsl coefficient of skew ness, A!lotlter measure of .kcwnes., which is mainly useful forme ... uringslight skewness Is known as moment coefficient of skewness and i, given by tbe formul ••

. . ... ../P.{Il, +3)

Mom.nt cacIT. of skewness '" 2{ )

5Jl,-6Il, -9

where Jl, aod .a, ore already given.

Moment coeffieient of s\:ewness is known as sec-ond coejJlc/t!fI' 0/ ,kewness,

Q. 67 Can second coefficient of skewness be u$Od 10 determine the mode?

Ans. Yes. mode can be determined by the second coefficient of .kewness using the relation,

= - second coerf. of skewness x S.D,

U rheberrec htl lch ge5c h utztes M ateri a

MEASURES' OF .D ISPERSION, Si<Ew;£ss AlIO KuRTOSIS

69

SECnON-B Fill In the Blanks

Fill in the su;tablfl wonl(s) or phraSfl{s) in thfl blanks:

1. The 'tatement, "The term dispersion is used to indic.te' the facts that within. group, the items different from one Mother in sire or in other word s, 'tIIen: is lack of unifonnily in

their size" was given by _

2. "Dispersion is the extent to which the magnilude, Or qualilies of the ,ilems differ thai is the degree of divemty"i, the statement about

dispersion given by _

3. The defin ition, "Dispersion is the measure of the varintion of the i tern," was giveo by

4. "Dispersion or spread i. the degree of the scauer or variation of the variable about a central value" i. the definition of dispersion

given by _

So defined dispersion as, "The

degree to which numerical data tend to spread aboul lUI .verage value is called the variation Or dispersion of the data."

6. Dispersion WIIS explained lIS, "measures of dispersion arc me:tS1JJeS of 1<:001er .bout an

average" by _

7. , definition of dispersion is ':A

measure of variarion or dispersion describes the degree of scatter shown by the

• observations and is usuwly measured us an average deviation """ut some central value or' by an order .. atistic,"

8. Measure of dispenion which utilises only

t,wo observarions is _

9. is !be me as ure of dispersion

which utilises only extreme values,

10. Coefficient or range is the, _

measure of dispcnlon.

JI. Inter·qlWlil. range is equal 10 _

12. Percentile range is given lIS _

IJ. Percentile range in terms of dedi", is given

.. _---

14. Mean deviation is based on _

15. Mean deviation is minimum about •

16. mean deviation suffer. with the lacuna that. it

considers all. differences _

17. Coefficienl or mean devlation about a central

value ~c· is given as _

lao The formula for calculating mean deviation about median for a discrete frequency distribution is

1,9. Slllnd<1nd deviation is the _

variance.

20. Other name, of SIan dud d ev i arion are

or

21. B"'I measure or dispersion is _

22. Measure or dispersion suitable for comparing

any two series. is .

23. The relation between variance and standard

deviation is _

24.

____ tho value or coefficient of range, better il is,

____ devialion con be obtained io case of open end i ntervals,

26. Quartile, deviation is not affected by ____ per elml el:tmnc valu ....

2:1. Quartile deviation is a measure

or dispersion .

'211. The relation. 2Q, ~~. + Q, holds in case of

25.

____ distribution,

29. Mean deviation ls calculated by considering

_ deviations.

30. is highly susceptible 10 sampling

fluctuatioru;.

Jl. Measures of dispersion tell aboot _

of central value.

32.. Interqunrtile range is of quartile

deviation,

U rheberrec htl lch ge5c h utztes M ate,; a

70

33. Medinn absolute deviation ts the _~ __ of the absolute deviations taken from median.

34. moment represents variance.

35. The relation between probable error and

standard devi31ion ;, _

36. The ratio of quartile deviation to mean

deviation is _

37. The r3l;O of 'I uartile deviation to standard

deviation is _' _

38. The mm:ncots about mean are called ____ moments.

39. Momenl, about origin zero are called ____ moments,

40. Moments of odd nrder ahaut mean for a

syrnrnetrlc dlstribetinn are _

41. Shepptll"ll·, OOITf:(;tiao adjum Lbo error due 10 consideration of frequencies located ar ____ of the class interval"

41. For a ,ymm<l!Oc distribution coefficient of

~ewn~is _

43. Q, $. and 'y, arc coefficienll

ofskewneu

44. Moment coefficient of skew ness is call ed ____ coefficient of skewness,

45. Measure of central tendency which can be determined with the help of second coefficient

of.k~is _

46. In C3$I) of normal distribu.tioo. mean. median

~nd mode are _

47. Less is the coefficient of varialiOll. _

cons"jnent is die series.

48. The relation between III and r, is

49. Filr. leplokut1ic CUM:, the relation between

J.l. and,... is _

50. F'or a mesokurtic curve. II, equal to

51. Filr a plotykUlti. curve is Y, is _

52. Fo. a plntykurtic curve. the, relation between second OIId foonh mornem i.

PROGRAMIoIED STATlSncs

33, Jl, gives the measure of skewness but

54. Kurtosi, means , of the frequency

curve.

55. Skewn"". m.e4II$ of the frequency

distribution curve.

56. Sum. of squares of the deviations from mean

is _

57. The relation that bolds between percentiles

in case of symmetric dislributionis .

58. For. ,ymmc:uic distribution. lSI, 51b and 9th deciJes are connected lIS _, _~~ _

59. The relation of thin! cetltr.II momeot with

raw moments i, _

60. There is effect of change of

origin on !he valu es of moments.

61. Change of scale the value of

moments.

62. , can. be calculated about an

arbitrary origin ..

63. With the help of a measure of disper.;iOll,

namely. , area. under the normnl

curve "an be determined.

64. Karl Pearson's fonnul~ for measure of

skewn ... is _

65. IJthe coeffiCient of KuIIo." is equal to 3, tho

frequency .,",,,,e i. _

66. If the coefflclem of KUrtosIs i. 8l"ater thlD

3, the distribution ;s _

61. If Ibe mean, mode .nd standard deviati.o.n of a dillribuliOll llli:4&, :l8 BOd 10 respectively.

the distriblllion i$ ~ew.

68. If for an asymmetric distribution. the mean, median BOd S.D. are 2S, ] S and 10, fl'$jleCt;·

vcIy the distributio.n is skew.

69. If II, ,,0 and ~ '" 3. it is known as frequency _~ __ curve.

'71t. Menn i. not equal to in case of

skew distribution..

11. For a. symmetric distributiOll. upper and 10 ......

qumilcs are equidimn! from _

U rheberrcc htl lch gesc h utztes Iv! ate,i"

M.EAS\!Rcs OF OlSPERSJOH. SKEW!<ESS N<O KuRTO$lS

72·. If skewness lsnegailve, the mean is _

mode.

73. For moderately skew distribution, (he rei arion between mean. rued ian and mode is

74. Me:rn deviauon ca"never be, ~_.

75. For ".l<:uloti.n8, variance, deviations are taken

from~ _

76. Coefficient of variation is usually expressed

in _

77. Graphical method of measuring dispersion is

through _

78. It is necessary 10 !he data before

"Loren. curve is drawn,

79. C""fficient of mean deviation IU\d coeffident

of varialion are _

80. If quartile deviation of a series is 30 and median is 45, the coefficient of mean deviation is _~ __

81. If mean deviauon is ] 6 and mean 3D, Ihe

coefflciem of variatinn is per

cent

82. If quartile deviation of certain i!emS is 10 WId mean is SO, the codfo;ien( of variation is ____ per cent.

83" I f team A has mean score. 7 and variance 15, team B has mean score 6 and variance 9, ____ is mcm: consistent.

84. Stan~rd deviation is • _

me .. ure of dispersion.

85. Range is a order measure of

dispersion,

86, In;)J>y distribution mean deviation. is _

,sta~rd deviation.

87. the, distance of Lorenz Curve

from Ihe li ne of equal di st ribution, the ",,'1'" is the variabilit.y in tbe.series of values.

88. Tbe .verageofsqo=ddeviations from mean

i, coiled _

89. A measure, of dispersion is a measure ("If

reliabililY of an _

90. If ~ distribution has. mean ,,7.5, mode '" 10

71

and skewness " " - 0.5. the variance is

'L If the maximum value ill 3 serJC"S is: 60 and coefficient of range 0.5. the minimum value of tbe ~rie$ i$ __ ~~_

9~ Formula for coefficient of mean deviation is

93. The standard deviation of ')1,1' five

observations 5, 5, 5, S. 5 is _

94. Mean deviation is always q!lBrtile

deviation.

9S. The surn of squares of deViations taken from mean 40 for 9 sample observations is 18S.

The coefficient of variation is _

96. The sum of we deviations from modi an 24 for 12 observation. is 72. the coefficient of

meun deviation is _ .. _

'J7. Range of" set of value. is 16 ~nd its mini· mum value is 2 t, the maximum value is

98. If mean and standard deviation of 8 observalions in a sample are 9 and 4 and that of second sample of ,;ire 4 ore [5 and 1. lhe combined varian.ce of the tWO samples is

oro",

99. For a highly skewed distribution, lbe best

rneasure of central "alu~ is _

tOO. Iftll. mean ofa distributioni. 15 and variance "25. Also given 1M! ~, '" 1, I"" third moment

about origin is _

101. Vlllianee of rneoian for a sample of slze n

from. normal P"l''-tllllion is _

102. Sl4IIrinrd deviation utilises mean.

t03. Variance of sample median. is _

Wan the variance of sample mean.

104. 11>< inequality that holds between ar'lwmetic mean (A.M.) and S!andard deviation (S.D.) is

lOS. Standard devlation of lirst n natural rnnnbe rs

is _

106. 11>< m eas ure of central tendency about which

Kurtosls is marked is ~_

U rheberrec htl lch ge5c h utztes Iv! aleri"

72

101. Standard dcvlllli.on gives more weight to ____ vruues.

108. StlIndard deviation is """ful in finding the

descriptive measures like and

_~ __ of a distribution.

109. Variation in data can graphically be perceived

by _

uo. If the sum of deviations from moo;". is not

zero. !hen a distribution will be _

Ill. II the frequencies on. either side of mode nrc not similarly distributed. the frequency

SECTION-C

PROOAAUNeo STAllSTlCS

distribution curve will be _

lIZ. Measure of skewnes; provides the _

and of asymmetry present in a

distribution.

113. Measure of Kurtosis shows the degree of ____ of a m:quency distributioo curve.

114. Standard devilllion of two values X, and X,

is equal 10 ~

115. (lIanY' of origin and scale of values for. set makes the calculation of standard deviation

Multiple Choice· Questions

Select the r;onect altemab\la out 01 gNBn ones:

Q. I Whi.ch of the following is not a measure of disperaion?

(~) mean deviation (b) quartile deviation (c) standard devialio.n

(d) .'''''1&<' <kviaJion from mean

Q. 1. Which of the following is a unitless measure ofdispo:r.;ion?

(a) :;tMdard deviation (b) mean deviation

(c) codficiCQ( of variation (d) range

Q. 3 Which one of (he given measures of dispersio .. il; considered best?

(a) SlI'Indard deviation

(b) runge

(c) vanapce

(d) coefficient of Y:uUtiOQ

Q. 4. r..,r comparison of tWO different series, !he best measure of dispersion. is:

(a) range

(b) mean deviation (c) standard dev!.ti01l (d) aene of the above

Q. 5. Ccrrecr formula for mean deviation from a constant A of • series in which me variate

valtH:$ x" x" •...• x. have frequencies Ii. fr

... , f. =pcctivcly is: .

(0) ..!.. ~).r.x,-A)

N,

(d) ..!.. 1:1" -<,-Ai

N,

where i = I. 2. ...• k andrli ~ N

Q. 6. Conect formula ror variance of n sample observations .fIt .:c 2' •••.• xill is:

(b) _I (1:xl-i')

n-l I

U rhcberrec htl lch gesc h utztes M ateri a

MEASURES OF DtSPERSIOII. SKEwNess AND KURTOSIS

1" I -I (d) - ':: ... /" - ~

II· i

Q. 1 TIle CO"""" n: lation between variance and standard devlation (S.D.) of. variable X is: (a) S.D." [Vo. (X)]l

(b) S.D." [Var (X)]ln

(e) S.D." Vft1 (X)

(d) none of the above

Q.8 Formula for coefficient of variation is,

(a) C.V.=~~lCO

mean

(b) C.V .• ~"JOO S.D.

(c) C.V", me:m~S.D. 100

(d) C,V.= __ I_OO--::-:::mean "S.D.

Q.9

Out of all measures of dispersion, the easiest ope to calculate ls:

(a) siandard deviation (b) range

(c) variance

(d) quartile deviation

Formula for range (R) of a set of vatues XI' X2"' •• ' X,. is:

(a} R= X"", - X..,. (b) R ~ IxmIo - x"",,1 (c) R=X_-X~, (d) R=X.-X,

Formuln for coefficlem of 1lUI8e of the oct of observations X •. Xl' ... "' X-n is:

(a) coeff, of range '" X,..., - X ....

X ..... + X.."

(b) coeff, of range .. X""' + X .. , X.",-X ...

Q.I0

Q.ll

X .... (c) coeff of range " X"".

73

X ,-X.

(d) coeff, of range" .. " .....

X"""

Q. 12 Coefficiefll of quartile deviation ls given by the formula:

(a) ceeff. of Q. D, = QJ + QI Q,-Q.

(b) coeff, of Q.D_. QJ + Q.

Q,-Q,

(0) cooIT. of Q.O. ~ Q, - Q, Q,-o.

(d) cooIT. of Q.D .• Q, - Qj Q; +Q,

Q. 13 For asymmetrical distributioQ, M J ± Q. D.

ICOVr=rs:

(n) 25 per cent, of the observations (b) 50 per cent of the observations (e) 7S per tent of the absent.ti,ons (d) 100 per cent of the observations

M.= Median and Q.D." Qua:tiledeviation Q. 14 Quartile deviation or semi inter-quartile deviation ts given by the formula: .

(a) Q.D.=o.+Q,

2

(b) Q.D.=~-Q,

(el Q, D.= (Q) -Q,)/2 (d) Q.D.= (0. -Qd/4

Q. IS Mean deviation is mi nlrnum When devinliau. are taken from:

{a) mean

(bj median

(C:) mode

(<1)=

Q.16 Sum of squares ofthe deviations Is minimum when dc·viations are taken from:

(a) mean

(b) median

(c) mode

(d) lPO

U rheberrcc htl lch gcsc h utztes M ateri a

14

Q. 17 If a conSlant v.loe 5 is subtracted from each obse.rvation of a set. the varian"" is:

(a) reduced by 5

(b) reduced by 25

(0) unDJlcrcd

(d) mcreased by 25

Q. 18 lf each observation of. set is divided by 10. the S.D. of !he new observation. is:

I

(a) j(jth of S.D. of original ob s.

I

(b) 100 th of S.D. of original obs,

(e) not changed

(dl 10 times of S.D. of original obs

Q. I' If each value of a set is divided by c, !hen the variance (ill of !he original samplo values from coded value, <Lt,. dx1 •...• <Lt. is obt.ined by the foemul.:

(0) .. ' =_I_L:{<LtI_iix)l xc n-l ,

(h) 5',. -.!.... L:(<Lt,-iix)' xc' n-l I

(e) s' m _1-L:{<LtI-iix)' ~.!.

n~,1 i C

1 I ~( -)' I (d) S =_. ~£... dx,-dx"l

II-I 1 .c'

Q. 10 Which oflhe following fO!11lulafor standard deviation of " f""lueney distri bunon i. not correct?

PROGRAMMED STATlsncs

where r.di = N 31Id oIhcr notations "'" as usual.

Q. n If" constant 'c' is sublnu:ted from each observatlcn and then divided by d. then the formula for variance of f""lucnc.y di,tri· botioo having k groups is:

(b) ,,1 ,,-1-[EJi(X;-X'l']><d2

N-I ,.1 .

[ .']

,t tft'-;

( 'J 1 - .1 '"I. ,I c., ) d'

c 0 ~ -f_. l~l ---- ><

N ,., N

(d) None of the above

where xi = Xi -c and N ="I:.Jj

d .,

Q. II The empirical relationship between quartile deviation (Q.D.) and s1ru1dard deviation ill Donna! distribution is:

(a) 3 Q.D, .. 2 S.D.

(b) 4 Q.D ... 3 S.D.

(e) 6 Q.D. ~. 5 S.D.

(d) 5 Q.D. ~ 4 S.D.

Q. 23 11,. relationship between mean devialion (M.D.) and s1ru1dMd devialion is:

(a) J M.D., = 2 S.D.

(b) 5 M.D. = 4 S.D.

(c) 6 M.D. " 5 S.D.

(0) M.D. = S.D.

Q. Z4 The ernpirical relanon between quartile devlation (Q.D.J and mean deviation (M.D.) from mean is:

U rheberrcc htl lch gesc h utztes M ateri"

MEASURES Of OlsPERSION, SKEWNESS NIO KURTOSIS

(a) 3 Q,D. ,: 5 M,D,

(b) 6 Q.D 3 M.D,

(e) 5 Q.D 6 M.D.

(d) 6 Q.D, ,;, 5 M.D.

Q. 2S An empirical relation between standard deviation. mean deviation about mean and qu a rtile deviation is:

(a) 4 S.D 6 M.D. ,;, 5 Q.D.

(b) 4 S,D 5 M.D 6 Q,D.

(e) 6 S.D. = 5 M.D 4 S.D.

(d) 5 S,D. = 4 M.D. '" 6 Q,D.

Q. 26 TIle empirical relation between flmg" (N) and standard deviation is:

(a) R = 3 S.D,

(b) R= 25.0.

(c) R= 65.D.

(d) 8 = 4 S,D.

Q. 27 An empirical relationship between range and quartile deviation about mean is:

(a) R =4 Q.O.

(b) R .. 9 Q.O.

(e) R '" 6Q.D.

(d) nona of the obove

Q. 28 An empirical relation bcl.w""", range and mean deviation is:

(a) R", to M.D, (b) 2R '" S M,D, (e) 38 = 5 M.D. (d) 28 = 15 M,D.

Q. 29 Which measure of dispersion ensures highest degree of reliability?

(a) range

(b) mean deviation (c) quartile deviation (d) standard deviation

Q. 30 Wbich measure of dispersion ensures lowes! degree of reliabi lity?

(a) .. ange

(b) mean deviation (e) quartile deviation Cd) Sland.rd deviation

Q. 31 There are 1"'0 populations consisting of N, and Nt units, lutving mean. X, ond Xl'

75

variances .. ; and .. i respectively, utlbeir poo led mean. be X". Also. supposing X, -X" .. d, and X,-X" = d1. The formula for pooled variance is:

I N,(o; +dn+Nl(ai +di)

(0) a a N, +N)

1 N,(ar + dn + N,{al + di)

(b) " = N, +Nl

.. 2 _ N,,,: + N,df + N2o! + N,di

(e) Nt +N,

... = N,a~ + N, .. j + N,dt + NIdi

(d) • N, +N,

Q. 32 Avora,e wages of workers of a fllClOry are Rs. 550.00 per month and Ill •• landlord deviation of wageS is 110. The coefficient of variation is:

(a) C.V. = 30 per cc.nt (b) C.V. = 15 per cent (e) C,V. = 500 per cent (d) C.V. '" 20 per cent

Q. 33 If the mean deviation of a distribution is 20.20, tile standard deviation of the distribution is:

Ca) 15.15· (b) 25.25 (e) 30.30

(d) none of the above

Q. 34 If tho mean and sUtndard deviation of A and B are as, XA = 15,0. X. = 20.0 and"-~ = 25 and oi = I(i. which of Ille two series is more consistent

(D) .ericsA

(b) series B

(e) series A and B are equally consistent Cd) none of the ebove

Q. 3S !C the standard deviation of a distribution is rs. the qumtile deviation ofllle distribution is:

J rbeberrec htl ieh gosc h utztes M atcrla

76

(a) IH) (b) 12.5 (e) !OJ)

(d) none of the above

Q. 36 [flho quanil~ devi Olion of a series is 60, !he mean deviation of llIi. series is:

(oj n

(b) 48

(e) SO

(d) 75

Q.37 In a discrete set of valees, tnecorreet relalion between mean deviation and standard devintion is:

(~) M.D. > S.D. (b) M.D. < S.D. (c) M.D. S S.D. (d) M.D. ~ S.D.

Q. 38 If the first 15 per cent observations of a series are 20 or le ss and last 15 per cent oh~;uions of'" series are 50 oemoee, the quartile deviation (semi inteT-qu:u:tile devi lui on) is:

(a) 15'

(b) 3S

(c) 15

(cl) )0

Q. 39 The mean and sW>danl ~ialiop ofa set of values are 2S and 5. respectively. If • constant value 5 is added to each vatuc, the coefficient of variation of the new set of values lir

'(a) 250 per cent (b) 600 per cent (e) 20 per cent (d) 16.6 per cent

Q. 40 If the mean oh series is !O and its coefficient of variation is 40 per cent, the variance of the series is:

(0) 4

(b) 8

(ej 12

(d) none of tbe above

Q.41 which of rhe following measures of dispersion can aLL"in 'a negative value'

PROG1WdMEO STATISTICS

(oj range

(b) mean clev,;"tion (e) standanl deviation (d) v..nBllce

Q. 41 A set of values is.aid to be relati.vely unifurm if it has:

(n) high dispersion (b) zero dispmi.Oll (e) little dispersion

(d) nesotive dispersion

Q. 43 The mean and standanl deviation of a ,set of values from. a nonna! distribution arc 66 and 4. respectively. The range in wbich almost 95 per cent values lie i.l:

(a) 62 to 70

(b) 621074

(c) S8 to 74

(d) 66 BIId 74

Q. 44 The measure of dispersion wbich ignores signs of the deviaiicns {rom a central value

is;

(aJ rang.

(b) quanile·dcvrntion (e) standard deviation (d) mean deviation

Q. 4S Wbicb measure of dispersion is leasl affected by extreme values?

(a) range

(b) mean deviation (e) standard deviation (d) quartile deviation

Q.46 Which measure of dispersion i. most affected by extreme values?

(a) range

(b) mean deviation (e) standard deviatioo (d) quartile deviation

Q. 41 R>lnge ofa set of values is 6S and max.imum vclue in the se riCll is 83.'The minimum value of tile series is!

(a) 14

(b) 9

(c) 18

(d) none of the above

U rhcberrec htl lch gesc h (jutes M ateri"

MEASURES OF OtSP,5RSIOII, SKEWNESS AND KURTOSIS

Q,48 If the minimum v3100 in a set is 9 and its range is 57.lhema.imum value of the set is (Il) 33

(b) 66

(c) 48

(d) none or the above,

Q. 49 U'the val ues nf a set are measured io ern, the uni I of variance will be:

(3) no unit

(b) cm

(c) em'

(d) em'

Q.50 Whi,.:h measure of dispersion, has a different unit other than the: unit of measurement of values:

(0) range

(b) mean deviation (e) standard deviation (d) variance

Q.51 The average of the sum of squares of the deviations about mean IS called:

(a) variance

(b) absolute deviation (e) ,lOnd.rd deviation (d) mean deviation

Q. 52 Quartile devi.{i on is equal 10: (a) interquaztile range

(b) dOD ble lhe inlerq uartile range (e) half of the lnrerquartlle runge (d) none of the above

Q. S3 Which measure of dispersion can be calcutated in case of open end in!erval$1

(a) range

(b) standard deviation

(c) coefficieru of variation (d) quurtil e deviation

Q. 54 Which One property nut of the foIklwiog does not hold good in case of standard deviation?

(a) II i. distorted by extreme val"<s

(b) It is nOI wry .c·nsitive, 10 sarnp ling nuclu3tion, US compared 10 other measures.

(e,) II is a, unitl ... s rnensure of dispersion (d) 11 is 3 most "sed measure of dispersion

77

Q. SS If each value of a series is divided by S, its coefficient of variationis reduced by:

(,,) 0 per cent

(b) Sper cent

(c) 10 per cent

(d) 20 per cent

Q.56 If each value of. series is multiplied by 10. the coefficient of variation will be inCl'CllSCll by:

(a) 5 per cent (h) 10 per cent (e) IS per cent (d) 0 per cent

Q. 57 If a constant value 10 is subtracted from =h valuc of a series. the coefficient of variation will be,

(0) decreased incomparison to original vall!(:

(b) increased in comparison to original value

(e) sa me as original value (d) Moe of the above

Q. 58 If each value of •• eries is multiplied by a constant ~ct, ,he coefficient of vari_iltion. :as compared to original value i.:

(a) increased

(b) decreased

(e) unaltered

(d) zero

Q.59 lf each vnlueofaselisdividcdby a constant . Ii' • the coefficient of \'ilriali on will be:

(a) same a" original. value

(b) less than original value

(c) more than original .3JOe

(d) none of the above

Q. 60 For e positive skewed distribul ion. which of !he following incqul1Uty Imld,,?

(a) median> mode

(b) mode> mean

(e) mean» median

(cl) mean> mouo

Q. 61 Far , neBlilively skewed disuibllilon, the correct i "''<I" .Iily is,

(,) n1OUe. < median

(b) mean < median

U rhcberrec htl lch gesc h (jutes M ateri"

18

(c) mean < mode

(d) none of the above

Q. 62 For a positive skewed frequency curve, the inequalhy that holds is:

(a) Q, + Q, ,. lQ,

(b) Q, + Q, ,. 2Q~

(e) Q. + Q, ,. Q,

(d) QJ - Q. ,. Q,

Q. 63 For a negat! vety skewed frequeacy distribution curve, the, third central moment, (a) ~,.O

(b) It, <0

(e) It, = 0

(d) J.I, does not exist

Q. 64 For a symmetrical distribution. the coeffi· cient of skewness:

<a) «, = ]

(I) «;;_ 3

(e) «) = 0

(d) a, =-1

Q. 65 For R [eptokurtic frequency curve. the measure of Kurtosis,

(al a. =0

(b) Cl.,.-3

(c) «, < I

(d) Cl." 3

Q. 66 In case of a positive skewed distribution. the relation beiweea mean. median and nlOde that holds i.,

<a) median c- mean" mode (b) mean c- median ,. mode (e) mean = median = mode (d) none of the ebove

Q. 67 If 3 moderd!cly skewed dislribution has mean 30 and mode 36. the median of the distribution is:

(a) 30

(b) 28

(e) 32

(d) none of the above

Q. 68 Ifamodernt.lysk.~di5lributioohasmean 4() And median equal to 30. the mode of Ille distribution js~

PROORA.llMEO ST"TlS'TICS

(n) 10 (b) 35 (e) 20 (d) zero

Q. 69 First and third qunrtiles of • frequency distribution are 30 and 75. Also its coefficient of skewness is 0.6. The median of !he frequency distribution is:

(a) 40

(b) 39

(e) 38

(d) 41

Q.70 If the mean, 'landan! deviation and coefficient of skewn ess of a frequency dlstribuoon are 60, 45 and -0.4, rer;peclively, the mode of the frequency dislribution is:

Ca) 80 (b) 82 (e) 78 (d) 68

Q. 71 For a moderately .kew di.<tribution. the empirical relalion between mean (M). median (Md) and mode (Mol is:

(a) 3(M-Mo)cM-Md

(b) 3{Md-M)=Mo-M

(e) 3(M-MJ)=M-Mo

(d) 2(Mo - M) = 3(MJ - M)

Q. 72 If Ihe firs! quartile Q, = I 5 and !hird quartile Q, = 25. lIIe coefficient of quartile deviation is;

(a) 4

(b) 114

(c) 513

(d) 3/S

Q. 73 If the ftrSt quartile Q, = 20 and third quartile Q, = 50. the quartile deviation is:

(al 3S

(b) 15

(e) 2.5

(d) 0,8

Q. 74 For • negatively skewed distribution. the correct relation beiween mean, medi .. and mode i .• :

Urbeberrechtllch gcschutztcs Material

MEASURES OF DISPERSION, Si<EwNess AND KuRTOSIS

(a) mean" median", mode (b) median < mean < mode (e) mean < median < mode (d) mode < mean c median

Q,75 [ftbe modeof a frequency distribetlon Mo= 16 and ils mean X: .. 16, the mediAn of ill.!: distribution is:

(II) zero

(b) 16

(e) sz

(d) 8

Q. 76 ln C$) of po!itivesk""'Cd <li .s tribulion, 1IIe extreme values lie in the

(,,) left (.Oil

(b) righlillil

(e) middle

(d) anywh<fm

Q. 77 The extreme value. in a negotivdy skewed distribulion lie in the!

(n) middJe

(b) righl tail

(e) left !ail

(d) whole CUrve

Q, 78 The relation between variaree .nd standard dev ialion is:

(.) variance is the sqUW'C root of standard deviation

(b) standard deviation is the square of the variance

(e) variance is equal to standard deviation (d) square oflhc .• tandard deviation is equal to variance

Q. 79 Which of Ihe following statements is true for a measure of disp<l'Sion?

(n) mean deviation does not follow algebraic rule

(b) range" a crudest measure

(e) coefficient of variation is II relative measure

(d) all the above statements

Q .. 80 For a set of vnl Des:

(a) mean deviation is alway. len than standard oil.'"Viation

(b) mean deviationis always greate, than slllndard deviation

79

(e) mean deviution is .Iway. equal 10 standard deviation

(d) none of the above

Q. 81 Variance of rhe foUowing frequency distri· bution,

Clru~~. Freq".nQ

2·4 . 2 .

~6 5

&8 4

8·10 I

is approximately equal to:

(II) 2.!i

(b) 2.9

(e) 5.0

(d) none of the above.

Q. 82 Fo, the data given is question 81, mean deviations .boul median is:

(a) 1.43

(b) 1.00

(e) 2.43

(d) 6

Q. 83 Coefficienl of variation for the datn given in question 81 is:

ta) 48.3] pe' cent (b) 206.90"", cent (c) 195.17 per cent (d) 30.03 per cent

Q. 84 The range of values for the frequency distribution given in question 8) is: .

(a) 2

(b) 10

(c) 8

(d) 6

Q. lIS For the data given in question S I, the coefflclent of quartile deviation is:

(a) 4.385

(b) 0.228

(e) 2.6

(d) IL4

Q, 86 The range of the set of values, IS. 12. 27. 6, 9.18.21 is:

(.) 21

(b) 4.5

(c) 0.64

(d) 3

U rhcberrec htlich gesc h utztes M ate,i a

80

Q. 87 The eoelflcient ofrangefor Lbe .. lues given in question 86 is:

(n) 1.571

(b) 4.500

(c) 0.636

(d) 0.222

Q. 88 'The coefficient of skew""", Olf a series A is O. J 5 and Ihal of series 8 0.062. Which of the Iwo series ts less 'xc"'?

(a) series a

(Ii) series B

(c) no decision

(d) none of the above

Q.89 If the coefficient of Kurtosis 1,) of a dlstrihotion i, zero, the f.equency curve is:

(a) leptokurtk'

(b) plai)'xunic

(e) mesoxunle

(d) ony of the above

Q. 90 If for. dlstribetion, coefficient of Runosis y) < O. the frequency curve il:

(~) leptokurtie

(b) pt,uykunic

(e) mesokurtlc

(d) any of the above

Q. 91 The wi ue of coefficient of Kurtosis PI can be:

<all ess than 3 (b) grontc: than 3 (cJ <qui to 3

(d) <lit thcabcwe

Q. 91 The .11I!Idnrd d.viation of a set of values will be:

(al positive when th. values are positive (b) positive when thovalue.s are negalive (c) alway. positive

(d) all the above

Q.93 Sum ofsquareorllled~iatio"si.minimum when the deviation, are talen from:

(a) mean

(b) median

(e) mode

(d) an arbitral}' value,

Q. 94 The re lalien.hip between the variance of media" (V (med» and variance of mean [V

(mean)] of a normal population ls: (a) V (mcd) < V (mean)'

(b) V (med) '" V (mean)

(e) V (med) > V (mean)

(d) 1.57 V (med) = V (mean)

Q, 95 Kllnnsi, in frequency distribution is adjudged around:

(0) .=000 quartile

(b) arithmetic mean

(c) qua6ratic mean

(d) mode.

Q, 96 The variance. of first n natural. numbers is:

(a) (nl +1)/12

(h) (n+I)'/12

(e) (n' ~1)/12 (d) (2nl ~ 1)/s

Q. 97 If. random variable X has mean 3 and standard deviation 5, then the vorirulce of a variable y;!OO 2X - 5 illl:

(a) 45

(h) 100

(e) 15

(d) 40

Q. 98 Kunosis and skewness of a frequency distri. bution curve arc bound by the. relation:

(a) !hey alwa)'. eoextst

(b) either of the two can exist olon. (c) their measures arc always positive (d) their measures arc, always negative.

Q. 99 All values in a sample are same. Then their variance is;

(a,) wo

(h) one

(e) not calculable (d) all the above

Q, 100 Calculati.on of pool ed variance of two series of sizes n I and n, requires:

(a) means of individual series (b) variances of individual.cries (c) pooled menn

(d) oil the .~e

U rheberrcc htl lch gesc h utztes M ateri"

- Business Statistics by [S P Gupta]Hochgeladen vonnexusvon
- MCQs Basic Statistics 1Hochgeladen vonMuhammad Imdadullah
- 40 solved MCQs of STA630- Research MethodsHochgeladen vonPower Girls
- Programmed Statistics by b l Agarwal PDF FreeHochgeladen vonRonak
- Mcq Testing of HypothesisHochgeladen vonJosé Juan Góngora Cortés
- Statistics and Data With RHochgeladen vonabrax65
- Fundamental of applied statistics by SC GuptaHochgeladen vonTff Bbjh
- Methods of Time SeriesHochgeladen voncul239
- Business StatisticsHochgeladen vonLokesh Baskaran
- Sampling Theory and MethodsHochgeladen vonagustin_mx
- MCQs on Correlation and Regression AnalysisHochgeladen vonMuhammad Imdadullah
- Statistics Mcq'S I.Com Part 2Hochgeladen vonShafique Ur Rehman
- Statistic Exam (MCQ)Hochgeladen vonzephyryu
- Operations Research - JK SharmaHochgeladen vonS V S Vardhan
- Fundamental of Stat by S C GUPTAHochgeladen vonPushpa Banerjee
- Index NumbersHochgeladen vonK.Prasanth Kumar
- Differential Equations (Golden Maths Series) - N. P. BaliHochgeladen vonCalibio Nhavene
- MCQ statsHochgeladen vonRahul Rein
- William G. Cochran Sampling TechniquesHochgeladen vonMuslim Ukm
- An Introduction to Multivariate Statistical Analysis.pdfHochgeladen vonRomán A. Pérez
- Applied+Multivariate+Statistical+Analysis+6E【课后习题答案】Hochgeladen vonxyunchao
- Applied Multivariate Statistical Analysis 6th EditionHochgeladen vonAbheera Hazra
- Mood - Introduction to the Theory of StatisticsHochgeladen vonFelipe Ortiz
- S.C. Gupta, V.K. Kapoor Fundamentals of Mathematical Statistics a Modern Approach, 10th Edition 2000Hochgeladen vonNikhil Singh
- Probability, Statistics and Random Processes 1Hochgeladen vonbaali4
- Applied Business Statistics, Methods and Excel-based Applications.pdfHochgeladen vonwalidkhalid1970
- Applied Regression Analysis - DraperHochgeladen vonrohail_taimour5089
- MATHEMATICAL STATISTICS (Keith Knight).pdfHochgeladen vonvictor_chung_23
- Statistics BookHochgeladen vonrajans1976
- A Quick Approach to Statistics by G.R.pashAHochgeladen vonAshfaq Khan

## Viel mehr als nur Dokumente.

Entdecken, was Scribd alles zu bieten hat, inklusive Bücher und Hörbücher von großen Verlagen.

Jederzeit kündbar.