Sie sind auf Seite 1von 21

TUTORIAL SHEET 1 MA4704

Diameter Height Test


1 2.23 3.76 1
2 2.12 3.15 1
3 1.06 1.85 2
4 2.12 3.64 1
5 2.99 4.64 3
6 4.01 5.25 2
7 2.41 4.07 1
2.75 4.72 2
! 2.20 4.17 2
10 4.09 5.73 3

1. From the above data use the Minitab statistics packae to raph the !o""o#in accordin
to the suitab"e data$
i. %i& 'istoram %ii& (o) p"ot %iii& *ie chart %iv& *"ot +iameter ,s
'eiht
2. -sin the data !rom the +iameter co"umn above ca"cu"ate the !o""o#in usin a
ca"cu"ator$
i. %i& Mean %ii& Median %iii& .tandard deviation %iv& inter/
0uarti"e rane
b. Ans - Diameter:(i) 2.598 (ii) 2.320 (iii) 0.916 (iv) 1.125
3. -sin the same data above ca"cu"ate the !o""o#in !or each accordin to the suitab"e data
usin the Minitab statistics packae$
i. %i& Mean %ii& Median %iii& .tandard deviation %iv& inter/
0uarti"e rane
b. Ans - Height:(i) 4.098 (ii) 4.120 (iii) 1.103 (iv) 1.335

4. 1ive an e)p"anation and e)amp"e o! the !o""o#in terms$
%i& 2ontinuous variab"e %ii& +iscrete variab"e

5. 3n artic"e in Qa!it" #ngineering presents viscosit4 data !rom a batch chemica"
process. 3 samp"e o! these data are presented be"o# %read do#n5 then across&.

13.3 14.9 15.8 16.0
14.5 13.7 13.7 14.9
15.3 15.2 15.1 13.6
15.3 14.5 13.4 15.3
14.3 15.3 14.1 14.3
14.8 15.6 14.8 15.6
15.2 15.8 14.3 16.1
14.5 13.3 14.3 13.9
14.6 14.1 16.4 15.2
14.1 15.4 16.9 14.4
14.3 15.2 14.2 14.0

%i& 2onstruct a stem/and/"ea! disp"a4 !or the viscosit4 data.
%ii& 2onstruct a !re0uenc4 distribution and historam.
%iii& 2onvert the stem/and/"ea! p"ot in part %i& into an ordered stem/and/"ea! p"ot. -se
this raph to assist in "ocatin the median and the upper and "o#er 0uarti"es o! the
viscosit4 data.
%iv& 2a"cu"ate the !o""o#in$
%a& Mean %b& mode %c& median %d& .tandard deviation
%e& inter/0uarti"e rane
Ans: (a) 14.$61 (b) 14.3 (%) 14.$ (&) 0.846 (e) 1.1$5

%v& +o a raphica" and numerica" summar4 o! the data.

%-se Minitab to assist 4ou in ans#erin this 0uestion&

6. 1ive e)amp"es %a& 6here 4ou #ou"d use and
%b& 6here 4ou #ou"dn7t use the !o""o#in$
(ar 2hart5 'istoram5 *ie 2hart

7. 8he !o""o#in data ives the "enths !or a samp"e o! 25 dri"" bits$
25 29 23 27 25
23 22 25 22 28
28 28 17 24 30
19 17 23 21 24
15 20 26 19 23
%i& 2a"cu"ate the standard deviation o! the data above. Ans: 3.9$6
%ii& .ummari9e these data usin a !re0uenc4 tab"e usin 6 c"ass interva"s.
%iii& .ummari9e these data usin a re"ative !re0uenc4 tab"e.
%iv& 2a"cu"ate the cumu"ative !re0uenc4 tab"e.
%v& +ra# a historam !or the data.

"#rm$%ae& :ane ; "arest "enth < sma""est "enth.
6idth o! each interva" ; rane = number o! c"ass interva"s
%to the nearest #ho"e number&.
Fre0uenc4 ; proportion o! observations !a""in into a iven c"ass.
:e"ative !re0uenc4 e)presses the !re0uenc4 counts as a percentae o! the
tota" number o! cases.

S'ri(g 1!! E)am *$esti#(

+ ,a- 2ompare and contrast the mean and the mode as appropriate measures
o! centra" tendenc4 %"ocation paramatics&.
,.- +escribe brie!"4 ho# 4ou #ou"d construct
%i& %i& (ar chart
%ii& %ii& 'istoram.
6hen is a bar/chart a more appropriate raphica" summar4 than a historam
%use data to support 4our ans#er&.
,/ - 3 machine is desined to de"iver 128 ms o! po#der into a container. 8#o
such machines are insta""ed at the end o! production "ines 3 and ( respective"4.
3 production manaer athered the !o""o#in data !rom both "ines.

0r#1$/ti#(
Li(e A
1285 128.55 127.55 1285 127.55 1285 128.55 1295 127.
0r#1$/ti#(
Li(e 2
1275 1285 129.55 128.55 1305 1295 1285 1265 127.


>ou are re0uired to
%i& 2onstruct a bo)/p"ot o! the data !rom each production "ine.
%ii& -se the bo)/p"ots to comment on the t#o machines.
%iii& 8he maintenance supervisor c"aims that because both machines p"ace
on averae the same amount o! po#der into each container there is no
need to overhau" either machine. 6rite a brie! report !or the operation7s
manaer in response to the maintenance supervisor7
S#%$ti#(s t# T$t#ria% Sheet 1 3 MA4704 4i%% .e
'#ste1 here #( Th$rs1a5 21st "e.r$ar5 2002+
TUTORIAL SOLUTIO6S SHEET 1 MA4704
1. Minitab .tatistics *ackae$
%i& 'istoram < 8o dra# a historam o! the ?uantitative data i.e. diameter
and heiht5 4ou o to the 1:3*' menu on the M@A- (3:5 then o
do#n to historam and c"ick on it. From here a bo) appears #here 4ou
enter diameter or heiht b4 doub"e c"ickin on the chosen one5 then
c"ick BC.
/ +iameter
4.25 3.75 3.25 2.75 2.25 1.75 1.25 0.75
4
3
2
1
0
Diameter
F
r
e
q
u
e
n
c
y
/ 'eiht
5.75 5.25 4.75 4.25 3.75 3.25 2.75 2.25 1.75
3
2
1
0
Height
F
r
e
q
u
e
n
c
y
%ii& (o)/p"ot < 8o dra# a bo)/p"ot o! 4our 0uantitative data i.e. diameter
and heiht5 4ou o to the 1raph menu on the M@A- (3:5 then o
do#n to bo)/p"ot and c"ick on it. From here a bo) appears #here 4ou
enter diameter or heiht b4 doub"e c"ickin on the chosen one5 then
c"ick BC.
/ +iameter
4
3
2
1
D
i
a
m
e
t
e
r
/ 'eiht
6
5
4
3
2
H
e
i
g
h
t
%iii& *ie/chart < 8o dra# a pie/chart o! 4our 0ua"itative data i.e. test5 4ou o
to the 1:3*' menu on the M@A- (3:5 then o do#n to pie/chart
and c"ick on it. From here a bo) appears #here 4ou enter test in the
2'3:8 +383 DA bo) b4 doub"e c"ickin on it5 then c"ick BC.
/ *ie/chart
1 (4, 40.0%)
3 (2, 20.0%) 2 (4, 40.0%)
Pie hart !" #e$t
%iv& *"ot +iameter ,s 'eiht / 8o p"ot diameter vs heiht5 4ou o to the
1:3*' menu on the M@A- (3:5 then o do#n to p"ot and c"ick on
it. From here a bo) appears #here 4ou enter diameter and heiht in the
E and > bo)es b4 doub"e c"ickin on them5 then c"ick BC.
/ *"ot +iameter ,s 'eiht
6 5 4 3 2
4
3
2
1
Height
D
i
a
m
e
t
e
r

2. %i& Mean '
'
n
i
;
i;1
n

) ; data
i ; 1F10
n ; samp"e si9e
; %2.23 G 2.12 G 1.06 G 2.12 G 2.99 G 4.01 G 2.41 G 2.75 G 2.20 G 4.09&
10
; 25.98 ; 2.598
10
%ii& Median < 2a"cu"ate the median b4 arranin the va"ues in increasin order.
D! there is an odd number o! observations5 the median is the midd"e number5
!ind the number in the position %k G 1&5 #here k is the number o!
2
va"ues 4ou have. D! there is an even number o! observations the median is the
mean o! the t#o observations occup4in the midd"e position5 mean o! the
observations in position k and k G1 #here k is the number o! va"ues 4ou have.
2 2

k ; 10
k ; 5 k G1 ; 6
2 2
Aumbers in increasin order/
Median ; %2.23 G 2.41& = 2 ; 2.32

%iii& .tandard +eviation
( )
s
' '
n
i
i
n
;

2
1
1

'
; mean
) ; data
n ; samp"e si9e
i ; 1F.10

s
2
; %2.23 < 2.598&
2
G %2.12 < 2.598&
2
G %1.06 < 2.598&
2
G %2.12 < 2.598&
2
G %2.99 < 2.598&
2
G %4.01 < 2.598&
2
G %2.41 < 2.598&
2
G %2.75 < 2.598&
2
G
%2.20 < 2.598&
2
G %4.09 < 2.598&
2
HHHHHHHHHHHHHHHH
10 / 1
s
2
; 7.54816 ; 0.838684
9
s ;
0.838684
; 0.916
%iv& Dnter/0uarti"e :ane < 8o ca"cu"ate the inter/0uarti"e rane #e !irst must
ca"cu"ate ?1 and ?3. 2a"cu"ate b4 arranin the va"ues in increasin order.
?1 is the va"ues at position %n G 1& and ?3 is the va"ues at position 3%n G 1&.
4 4
Diameter
1 1.06
2 2.12
3 2.12
4 2.20
5 2.23
6 2.41
7 2.75
2.99
! 4.01
1
0
4.09
6here n is the number o! va"ues. D! the position is an inteer5 interpo"ation is
used. Fina""45 #e ca"cu"ate the inter/0uarti"e rane %D?:&5 D?: ; ?3 < ?1.
?1 ; %10 G 1& ; 2.75
4
?1 is bet#een the second and third va"ues5 three/0uarters o! the #a4 up.
?1 ; 2.12 G 0.75%2.12 < 2.12& ; 2.12
?3 ; 3%10 G 1& ; 8.25
4
?3 is bet#een the eihth and ninth va"ues5 0uarter o! the #a4 up.
?3 ; 2.99 G 0.25%4.01 < 2.99& ; 3.245
D?: ; ?3 < ?1 ; 3.245 < 2.12 ; 1.125
3. Minitab .tatistics *ackae$
+escriptive .tatistics / 4ou o to the .838 menu on the M@A- (3:5 then o
do#n to (3.D2 .838D.8D2. and then o to +D.*I3> +@.2:D*8D,@
.838D.8D2.. From here a bo) appears #here 4ou enter diameter or heiht
b4 doub"e c"ickin on the chosen one5 then c"ick BC. >our resu"t appear in
the session #indo#5 !rom #hich 4ou can ca"cu"ate 4our inter/0uarti"e rane
and dra# a report !rom them.
/ +iameter
Descriptive Statistics
Variable N Mean Median TrMean StDev
Diameter 10 2.598 2.320 2.604 0.916
Variable SE Mean Minimum Maximum 1 3
Diameter 0.290 1.060 4.090 2.120 3.245
/ 'eiht
Descriptive Statistics
Variable N Mean Median TrMean StDev
!ei"#t 10 4.098 4.120 4.1$5 1.103
Variable SE Mean Minimum Maximum 1 3
!ei"#t 0.349 1.850 5.$30 3.51$ 4.85
4. ?uantitative data < can be 0uanti!ied i.e. the4 are associated #ith speci!ic
numerica" va"ues.
?ua"itative data < can on"4 be separated into di!!erent cateories.
%i& 2ontinuous ,ariab"e
?uantitative data5 random variab"es that can assume a va"ue correspondin
to an4 point on some interva"5 #ithout aps.
@)amp"e$ "enth o! a dri"" bit
%ii& +iscrete ,ariab"e
?uantitative data5 usua""4 associated #ith data that invo"ves countin5 ma4
have aps.
@)amp"e$ Aumber o! de!ects in a manu!acturin process.
5. %i& .tem and Iea! +isp"a4
-sin the Minitab .tatistics *ackae5 4ou o to the 1:3*' menu on
the
M@A- (3:5 then o do#n to stem/and "ea!. From here a bo) appears
#here 4ou enter the co"umn 4our data is in b4 doub"e c"ickin on the
chosen one5 then c"ick BC.

Character Stem-and-Leaf Display
Stem%and%lea& '& (5 N ) 44
*ea& +nit ) 0.10
3 13 334
$ 13 6$$9
18 14 01112333334
,8- 14 55568899
18 15 1222233334
8 15 6688
4 16 014
1 16 9
%ii& Fre0uenc4 distribution
-sin the Minitab .tatistics *ackae5 4ou o to the .838 menu on the
M@A- (3:5 then o do#n to 83(I@. and then 83II>. From here
a
bo) appears #here 4ou enter the co"umn 4our data is in b4 doub"e
c"ickin
on the chosen one5 then c"ick on the bo) #ith *@:2@A8. on it to et
percents and then c"ick BC.
Summary Statistics for Discrete Variables
(5 ('unt .er/ent
13.3 2 4.55
13.4 1 2.2$
13.6 1 2.2$
13.$ 2 4.55
13.9 1 2.2$
14.0 1 2.2$
14.1 3 6.82
14.2 1 2.2$
14.3 5 11.36
14.4 1 2.2$
14.5 3 6.82
14.6 1 2.2$
14.8 2 4.55
14.9 2 4.55
15.1 1 2.2$
15.2 4 9.09
15.3 4 9.09
15.4 1 2.2$
15.6 2 4.55
15.8 2 4.55
16.0 1 2.2$
16.1 1 2.2$
16.4 1 2.2$
16.9 1 2.2$
N) 44
'istoram
8o dra# a historam o! the ?uantitative data5 4ou o to the 1:3*' menu on
the M@A- (3:5 then o do#n to historam and c"ick on it. From here a bo)
appears #here 4ou enter #hichever co"umn 4our data is in b4 doub"e c"ickin
on the chosen one5 then c"ick BC.
17.25 16.75 16.25 15.75 15.25 14.75 14.25 13.75 13.25
10
5
0
5
F
r
e
q
u
e
n
c
y
%iii& Brdered .tem/and/Iea! *"ot
/ arranes the "eaves b4 manitude5 #hich is produced automatica""4 b4
Minitab.
%iv& +escriptive .tatistics / 4ou o to the .838 menu on the M@A- (3:5
then o do#n to (3.D2 .838D.8D2. and then o to +D.*I3>
+@.2:D*8D,@ .838D.8D2.. From here a bo) appears #here 4ou
enter
the co"umn in #hich 4our data is contained b4 doub"e c"ickin on the
chosen one5 then c"ick BC. >our resu"t appear in the session #indo#5
!rom #hich 4ou can ca"cu"ate 4our inter/0uarti"e rane and dra# a
report
!rom them.
Descriptive Statistics
Variable N Mean Median TrMean StDev
(5 44 14.$61 14.$00 14.$40 0.846
Variable SE Mean Minimum Maximum 1 3
(5 0.12$ 13.300 16.900 14.125 15.300
>ou can a"so ca"cu"ate these usin a ca"cu"ator.

H
%a& Mean '
'
n
i
;
i;1
n

) ; data
i ; 1F10
n ; samp"e si9e
; 649.5 ; 14.761
44
,b& Mode o! a set o! va"ues is the va"ue that occurs most !re0uent"4.
6e can ca"cu"ate this b4 countin the number o! times a va"ue appears or
b4 usin the Minitab .tatistics *ackae5 4ou o to the .838 menu on the
M@A- (3:5 then o do#n to 83(I@. and then 83II>. From here a
bo) appears #here 4ou enter the co"umn 4our data is in b4 doub"e
c"ickin
on the chosen one5 then c"ick BC. 8his ives 4ou a !re0uenc4
distribution.
Mode ; 14.3 #hich occurs 5 times.

%c& Median < 2a"cu"ate the median b4 arranin the va"ues in increasin order.
D! there is an odd number o! observations5 the median is the midd"e
number5 !ind the number in the position %k G 1&5 #here k is the
2
number o! va"ues 4ou have. D! there is an even number o! observations the
median is the mean o! the t#o observations occup4in the midd"e position5
mean o! the observations in position k and k G1 #here k is the
2 2
number o! va"ues 4ou have.

k ; 44
k ; 22 k G1 ; 23
2 2
Aumbers in increasin order/
1 13.3 12 14.2 23 14.8 34 15.3
2 13.3 13 14.3 24 14.8 35 15.3
3 13.4 14 14.3 25 14.9 36 15.4
4 13.6 15 14.3 26 14.9 37 15.6
5 13.7 16 14.3 27 15.1 3 15.6
6 13.7 17 14.3 2 15.2 3! 15.8
7 13.9 1 14.4 2! 15.2 40 15.8
14 1! 14.5 30 15.2 41 16
! 14.1 20 14.5 31 15.2 42 16.1
10 14.1 21 14.5 32 15.3 43 16.4
11 14.1 22 14.6 33 15.3 44 16.9
Median ; %14.6 G 14.8& = 2 ; 14.7

%d& .tandard +eviation
( )
s
' '
n
i
i
n
;

2
1
1

'
; mean
) ; data
n ; samp"e si9e
i ; 1F.10
s
2
; 30.91
44 / 1
s
2
; 0.718837209

s ;
0.718837209
; 0.846
%e& Dnter/0uarti"e :ane < 8o ca"cu"ate the inter/0uarti"e rane #e !irst must
ca"cu"ate ?1 and ?3. 2a"cu"ate b4 arranin the va"ues in increasin
order.
?1 is the va"ues at position %n G 1&
4
?3 is the va"ues at position 3%n G 1&.
4
6here n is the number o! va"ues. D! the position is an inteer5 interpo"ation
is used. Fina""45 #e ca"cu"ate the inter/0uarti"e rane %D?:&5
D?: ; ?3 < ?1.
?1 ; %44 G 1& ; 11.25
4
?1 is bet#een the e"eventh and t#e"!th va"ues5 0uarter o! the #a4 up.
?1 ; 14.1 G 0.25%14.2 < 14.1& ; 14.125
?3 ; 3%44 G 1& ; 33.75
4
?3 is bet#een the thirt4/third and thirt4/!ourth va"ues5 three/0uarters o!
the #a4 up.
?3 ; 15.3 G 0.75%15.3 < 15.3& ; 15.3
D?: ; ?3 < ?1 ; 15.3 < 14.125 ; 1.175
%v& 1raphica" and Aumerica" .ummar4
A 8here are 44 samp"es in the data set.
Mean 8he mean diameter is 14.761.
Median 8he median diameter is 14.7. 8he mean is "arer then the
median indicatin the distribution ma4 be ske#ed to the
riht.
8r Mean 8his is about 90J trimmed mean. Dt is the averae o! the
data va"ues a!ter removin the sma""est 5J %rounded to the
nearest inteer& and the "arest 5J. .ince 5J o! 44 is 2.2
or 25 the sma""est 2 and the "arest 2 va"ues are removed
and the mean o! the remainin 40 va"ues are ca"cu"ated.
8he trimmed mean o! 14.740 is s"iht"4 more than the
untrimmed.
.t+ev 8he samp"e standard deviation is 0.846
.@ Mean 8he standard deviation o! the mean is the samp"e standard
deviation %0.846& divided b4 the s0uare root o! the samp"e
si9e %44&. 6e use this measure to make in!erences.
Min 8he minimum diameter is 13.3.
Ma) 8he ma)imum diameter is 16.9.
?1 3bout 25J o! the diameters are "ess than 14.125.
?2 3bout 75J o! the diameters are "ess then 15.3 and about
25J o! the diameters are reater than 14.125.
17 16 15 14 13
5
%!&'(!t !" 5
3bout 50J o! the batch5 have viscosit4 bet#een 14.125 and 15.3K the
median is about 14.7. 8he area at either side o! the median seem to be
divided even"45 the #hisker to the riht hand side is "oner than the
#hisker to the "e!t hand side indicatin the distribution is ske#ed to#ards
the hiher viscosit4. 8he bo)/p"ot did not detect an4 out"iers.
17.0 16.5 16.0 15.5 15.0 14.5 14.0 13.5
10
5
0
5
F
r
e
q
u
e
n
c
y
Hi$t!gram !" 5
8he historam roups viscosit4 into c"asses o! #idth 0.5. 8he !irst c"ass is
centered on 13.5 and runs !rom 13.3 to "ess than 13.7. the second c"ass is
centered at 14.0 and runs !rom 13.7 to "ess than 14.35 and so on. 8he
distribution seems to be centered at about 15.0 and is s"iht"4 ske#ed
to#ard hiher viscosit4.
Character Stem-and-Leaf Display
Stem%and%lea& '& (5 N ) 44
*ea& +nit ) 0.10
3 13 334
$ 13 6$$9
18 14 01112333334
,8- 14 55568899
18 15 1222233334
8 15 6688
4 16 014
1 16 9
8he A ; 44 indicates that 44 va"ues are in the stem and "ea! disp"a4. 8he
"ea! unit stated in the disp"a4 is 0.10K #hich means the stem unit is 1. 8he
sma""est va"ue5 #ith a stem o! 13 and a "ea! o! 45 have a viscosit4 o! 13.4K 3
batches have a viscosit4 i! 13.4. 8he "arest va"ue have a viscosit4 o! 16.9.
8he center is at about 14.9.
2onc"usion$
From the in!ormation athered and ana"4sis #e conc"ude that the batch
!rom the chemica" process seems to sho# that most o! the process is
producin batches #ith hih viscosit4.
6+
,a- 7#$%1 Use ,.- 7#$%1(8t Use
2ar 9hart ?ua"itative +ata$
8est resu"ts in a !actor4
Ienth o! dri"" bits
Hist#gram ?uantitative +ata$
Ienth o! dri"" bits
8est resu"ts in a !actor4
0ie 9hart ?ua"itative +ata$
8est resu"ts in a !actor4
Ienths o! dri"" bits
7. %i& .tandard +eviation
( )
s
' '
n
i
i
n
;

2
1
1

'
; mean
) ; data
n ; samp"e si9e
i ; 1F.10
Mean
'
'
n
i
;
i;1
n
:
583
25
2332 = .
s :
379 44
24
.
:
1581 . ; 3.976
%ii&=%iii&=%iv& :ane ; "arest "enth < sma""est "enth.
: 30 < 15 ; 15
6idth o! each interva" ; rane = number o! c"ass interva"s
; 15 = 6 ; 3 %to the nearest #ho"e number&.
9%ass I(ter;a% "re<$e(/5 Re%ati;e "re<$e(/5 9$m$%ati;e Re%+ "re<
14.5 < 17.5 3 12 12
17.5 < 20.5 3 12 24
20.5 < 23.5 7 28 52
23.5 < 26.5 6 24 76
26.5 < 29.5 5 20 96
29.5 < 32.5 1 4 100
T#ta% ( : 25 100=
8o ca"cu"ate the cumu"ative re"ative !re0uenc45 the !irst !iure is the !irst
!iure in the re"ative !re0uenc4 co"umn and then !or the second !iure5 #e add
the !irst !iure in the re"ative !re0uenc4 to the second5 and then !or the third #e
add this number to third va"ue in the re"ative !re0uenc4 co"umn and so on and
so on.
%v& 'istoram < 8o dra# a historam o! the ?uantitative data i.e. "enths 5 4ou o
to the 1:3*' menu on the M@A- (3:5 then o do#n to historam and
c"ick on it. From here a bo) appears #here 4ou enter the co"umn containin
4our data b4 doub"e c"ickin on the chosen one5 then c"ick BC.
3 historam o! the data$
30 2) 26 24 22 20 1) 16 14
6
5
4
3
2
1
0
*
F
r
e
q
u
e
n
c
y
E)am *$esti#(3 S'ri(g 1!!
8. %a& 8he mean or averae is the sum o! the observations divided b4 the tota"
number o! observations. Dt is used #hen #e et a be""/shaped curve !or the
data. Dt is the centra" tendenc4 in the distribution5 or its "ocation.
8he mode o! a set o! data is de!ined as the number #hich occurs most
!re0uent"4.
8he mean is not necessari"4 the !i!tieth percenti"e o! the distribution %the
median&5 and it is not necessari"4 the most "ike"4 va"ue o! the variab"e
% #hich is ca""ed the mode&. 8he mean simp"4 determines the "ocation o!
the distribution.
6e usua""4 use the mean or the median to describe the center o! our
distribution. 8he mean is sensitive to e)treme va"ues #hereas the median
is una!!ected b4 e)treme va"ues.
D! the historam is s4mmetric this imp"ies that the mean is e0ua" to the
median #hich is e0ua" to the mode. 8he most common"4 used measure o!
centra"it4 is !or a s4mmetric distribution is the mean.
D! the historam is as4mmetric i.e. ske#ed this imp"ies #e #i"" use the
median as our measure o! centra"it4 since it #i"" be una!!ected b4 the
va"ues that are causin the historam to be ske#ed.
6hen #e have a "e!t/ske#ed distribution the mean is "ess than the median.
6hen #e have a riht/ske#ed distribution the mean is reater than the
median.
%b& 2onstructin$
%i& (ar 2hart <
+ra# a vertica" a)is about t#o and a ha"! centimeters or one inch in
!rom the "e!t/hand side o! the raph paper and a hori9onta" one a
simi"ar distance !rom the !oot o! the paper. 8hese a)es #i""
intersect. 8he point o! intersection on the vertica" a)is is 9ero. 8he
hori9onta" one ma4 be started at an4 convenient point near the
vertica" a)is.
2hoose sca"es #hich #i"" sho# 4our diaram to the best advantae5
and #i"" inc"ude a"" 4our data.
Iabe" 4our a)es #ith appropriate tit"e and mark in numerica" sca"e
va"ues.
3"" bars in a bar diaram shou"d be o! the same #idth.
%ii& 'istoram <
8o dra# a historam5 #e must !irst roup our data into interva"s.
8hen #e p"ace a number "ine on the hori9onta" a)is.
8he number "ine shou"d start at the sma""est va"ue in the !irst
interva" and continue unti" the "arest va"ue in the "ast interva".
Bn the vertica" sca"e5 #e mark o!! count va"ues or proportion
va"ues startin #ith 9ero.
8hen !or each interva"5 #e make a bar #ith the #idth coverin the
interva" and heiht e0ua" to the correspondin count or proportion.
3 bar chart is a more appropriate raphica" summar4 then a
historam #hen #e have cateorica" data such as car
manu!acturers. 8his imp"ies #e can "ist the di!!erent brands a"on
the hori9onta" a)is.
%c& %i& (o)/p"ot /
6e need to ca"cu"ate the median and upper and "o#er 0uarti"es
be!ore #e start to dra# a bo)/p"ot. 8hen5 dra# a bo) bet#een the
upper and "o#er 0uarti"es5 thus indicatin #here the midd"e 50J
o! the data !a"". 'ori9onta" "ines ca""ed #hiskers are then e)tended
!rom the midd"e o! the sides o! the bo) to the minimum and to the
ma)imum. 8he median is then marked #ith a vertica" "ine inside
the bo).
12*
12)
127
+
i
n
e

,
130
12*
12)
127
126
+
i
n
e

%
%ii& For Iine 3 about 50J o! the production is bet#een 127.5 and
128.55 !or Iine ( about 50J o! the production is bet#een 127
and129.255 both "ine productions have a median o! 128.
*roduction Iine 35 appears to run more smooth"4 then *roduction
Iine ( as 3 has e0ua" distance at both sides o! the median un"ike (
#here there is s"iht"4 more area above the median then be"o#. Dn
*roduction Iine 3 and (5 the #hiskers appear to have the same
"enth #hich imp"ies both distributions are not ske#ed. 8he
#hiskers on"4 e)tend to the "ast data va"ue #ithin 1.5 D?:5 and
an4thin outside this is considered an out"ier. But"iers are marked
individua""45 sometimes b4 an o. (oth bo) p"ots did not detect an4
out"iers.
%iii& Iine 3 appears to have a mean o! 128 #hich imp"ies it hits its
taret de"iver4 the maLorit4 o! the time. Iine (5 ho#ever has a
mean o! 128.11 #hich is s"iht"4 over the taret de"iver4 and so in
bu"k production #ou"d "ose a "ot o! its po#er b4 over!i""in the
containers. 8here!ore5 D #ou"d recommend that *roduction Iine (
needs to be overhau"ed.
TUTORIAL SHEET 2 MA4704
1. 8he !o""o#in data are the Loint temperatures o! the B/rins %derees F& !or each test
!irin or actua" "aunch o! the space shutt"e rocket motor$
84 49 61 40 83 67 45 66
70 69 80 58 68 60 67 72
73 70 57 63 70 78 52 67
53 67 75 61 70 81 76 79
75 76 58 31
%a& 2ompute the samp"e mean and samp"e standard deviation.
Ans: (ean ) 65.86 *tan&ar& Deviati+n ) 12.16
%b& Find the median and the upper and "o#er 0uarti"es.
Ans: (e&ian ) 6$.50 ,--er Qarti!e ) $5.00 .+/er Qarti!e ) 58.50
%c& .et aside the sma""est observation %31
o
F& and recomputed the 0uantities in parts %a&
and %b&.
Ans: (ean ) 66.86 *tan&ar& Deviati+n ) 10.$4 (e&ian ) 68.00
,--er Qarti!e ) $5.00 .+/er Qarti!e ) 60.00
2omment on 4our !indins. 'o# Mdi!!erentN are the other temperatures !rom this
sma""est va"ue.
2. For an e"ectronic manu!acturin process5 it current"4 has speci!ications o! 100
+
10
mi""iamperes. 8he process mean and standard deviation are 107.0 and 1.5
respective"4. 2a"cu"ate the process capabi"it4 and interpret this va"ue #hen the process
demands a minimum 2p o! 2.0. 6i"" the product be acceptedO Ans: 0- 1 2.22
3. 3n enineerin compan4 4ou supp"4 demands a minimum 2p o! 1.33. Bne component
the4 re0uire5 has a taret "enth o! 80mm. 8he product speci!ication de!ines an
acceptab"e error o! G=/ 5mm. 8he process standard deviation is ; 1.5. 6i"" the product
be accepted5 ive a reason !or 4our ans#er. Ans: 0- 1 1.111
4. 1ive a de!inition o! the !o""o#in$ %a& .ina" %b& Aoise.
5. 3 manu!acturer o! coi" sprins produces a standard "enth to be 10cm. Dt is areed that a
5J probabi"it4 o! a 84pe 1 @rror %sina" $ noise ; G=/ 1.96& is acceptab"e. 8he standard
deviations is 1.86cm. 3 samp"e o! 6 is taken !rom the production "ine and the !o""o#in
measurements #ere taken$
)1 ; 9 )2; 12 )3;7 )4;9 )5;8
)6;11
2a"cu"ate the sina" to noise ratio and ive a conc"usion to 4our !indins.
Ans: *igna! : 2+ise 3ati+ 1 0.8$84
6. %a& 3ssume that P has a standard norma" distribution. -se the tab"es to determine
the
va"ue !or 9 that so"ves the !o""o#in$
%i& *%/9 Q P Q 9& ; 0.95 Ans: 4 1 56- 1.96
%ii& *%/9 Q P Q 9& ; 0.99 Ans: 4 1 56- 2.5$5
%iii& *%/9 Q P Q 9& ; 0.9973 Ans: 4 1 56- 3
%b& 3ssume that E is norma""4 distributed #ith a mean o! 6 and a standard deviation o! 3.
+etermine the va"ue !or ) that so"ves each o! the !o""o#in$
%i& *%E R )& ; 0.5 Ans: ' 1 $.5
%ii& *%E R )& ; 0.95 Ans: ' 1 8.85
%iii& *%) Q E Q 9& ; 0.2 Ans: ' 1 $.08
%iv& *%3 Q E Q )& ; 0.8 Ans: ' 1 $.9

7. 8he "enth o! an inLected/mo"ded p"astic case that ho"ds manetic tape is norma""4
distributed #ith a mean "enth o! 90.2 mi""imeters and a standard deviation o! 0.1
mi""imeter.
%a& 6hat is the probabi"it4 that a part is "oner then 90.3 mi""imeters or shorter than 89.7
mi""imetersO Ans: 0.158$
%b& 6hat shou"d the process mean be set at to obtain the reatest number o! parts
bet#een 89.7 and 90.3 mi""imetersO Ans: 90
%c& D! the parts are not bet#een 89.7 and 90.3 mi""imeters are scrapped5 #hat is the 4ie"d
!or the process mean that 4ou se"ected in part %b&O Ans: 0.99$3
S'ri(g 1!!! E)am *$esti#(
8.%a& 3ssume that the "enth o! catheter manu!actured !or the 'ea"th 2are
industr4 are norma""4 distributed #ith a mean o! 100 mm and a standard
deviation o! 1mm. +ra# a rouh sketch and then ca"cu"ate the
correspondin probabi"ities !or the !o""o#in measurements occurrin on an
individua" catheter
%i& (et#een 100 and 101.5 mms. Ans: 0.4332
%ii& Iess than 98.6 mm Ans: 0.0808
%iii& (et#een 98.5 and 101 mm Ans: 0.$$45
%iv& Iess than 100.5 mms. 3ns$ 0.6915
%b& 3ssume that 9 scores are norma""4 distributed #ith a mean o! 9ero and a
standard deviation o! 1.
%i& D! p%0 Q 9 Q a& ; 0.388 !ind a Ans: a 1 1.215
D! p%/b Q 9 Q /b& ; 0.7698 !ind b Ans: b 1 0.295
D! p%9 Q c& ; 0.1841 !ind c Ans: % 1 0.9
%c& 8he mean time bet#een !ai"ures !or protot4pes o! computer printers is
kno#n to be appro)imate"4 norma""4 distributed #ith mean

and variance

2
. 3 proLect enineer has !ound that 80J o! the printer protot4pes "ast
!or at "east 24 hours continuous use and that 6J "ast !or 44 or more hours
continuous use. 2a"cu"ate the mean time bet#een !ai"ures and the standard
deviation based on the proLect enineer7s resu"ts.

%d& 8he e"ectronic components o! the printer !o""o# an e)ponentia" distribution
#ith a mean "i!e o! 200 hours. 6hat is the probabi"it4 that the e"ectronics #i""
%a& !ai" in the !irst t#ent4 !our hours.
%b& .urvive !or more than 44 hours.

S#%$ti#(s t# T$t#ria% Sheet 2 3 MA4704 4i%% .e
'#ste1 here #( Th$rs1a5 2th "e.r$ar5 2002+

Das könnte Ihnen auch gefallen