Beruflich Dokumente
Kultur Dokumente
Advantages
cost
speed
accuracy
convenient
2 stages:
1. Divide the target population into sampling
units
E.g households, trees, water samples, cities
Sampling methods
Aim
To obtain a sample that is representative of
the target population
Categories:
non-probability sampling
probabibility sampling
Non-probability sampling
procedures in which the units that make
up the sample are collected with no
specific probability structure
E.g.
units are self-selected (volunteers)
units are the most easily accessible
units are selected on economic grounds
(payment for participation)
units are chosen without no obvious design
Purposive sampling
Quota sampling
Convenience sampling
Snowball sampling
Self-selection
Purposive sampling
Participants selected subjectively
Researcher pick a sample he believes
is representative to the population of
interest
Respondents selected using judgment
of interviewers
Snowball sampling
Used when trying to interview hard to
reach groups e.g unemployed people
First, contact a few potential residents
Next, interview them and ask if they
know anybody else with same
characteristics you are looking for
Convenience sampling
the most convenient population is
chosen, which may be the researchers
work colleagues, friends, students from a
nearby college
used to save time and resources
procedure
select people most conveniently available to
interviewers e.g ask shoppers, or passersby to participate in a product taste
Quota sampling
interviews, mail surveys and telephone surveys
are often used in conjunctioned with quota
sampling
based on defining the distribution of
characteristics required in the sample
e.g if sample requires 50 men and 50
women, a quota sample will be selected until
the right number of each type has been
surveyed
can be extended to cover several
characteristics (males <50 years old;
females with children)
Self-selection
Self-explanatory respondents
themselves decide whether to take part
in the survey or not
Probability sampling
basis: selection of sampling units to
make up the sample based on defining
the chance that each unit in the sample
frame will be included
concerned with the probability of each
sample being chosen
each sampling unit has the same chance
of being included as every other
sampling unit
systematic sampling
cluster sampling
multi-stage sampling
simple random
select units by random sampling from
frame
use random number tables, or
use short computer program to generate
6 random numbers between 1 16
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
10480
15011
01536
02011
81647
91646
69179
14194
62590
36207
20969
99570
91291
90700
22368
46573
25595
85393
30995
89198
27982
53402
93965
34095
52666
19174
39615
99505
24130
48360
22527
97265
76393
64809
15179
24830
49340
32081
30680
19655
63348
58629
42167
93093
06243
61680
07856
16376
39440
53537
71341
57004
00849
74917
97758
16379
37570
39975
81837
16656
06121
91782
60468
81305
49684
60672
14110
06927
01263
54613
77921
06907
11008
42751
27756
53498
18602
70659
90655
15053
21916
81825
44394
42880
99562
72905
56420
69994
98872
31016
71194
18738
44013
48840
63213
21069
10634
12952
96301
91977
05463
07972
18876
20922
94595
56869
69014
60045
18425
84903
42508
32307
89579
14342
63661
10281
74553
18103
57740
84378
25331
12566
58678
44947
05585
56941
10
85475
36857
53342
53988
53060
59533
38867
62300
08158
17983
16439
11458
18593
64952
Col./Lin
e
Stratified sampling
apply when: target population consists of
a series of separate 'sub-population',
each may have different values for the
properties we are studying
process of splitting - stratification
total population (N) divide into a set of L
(mutually exclusive sub-populations
N=(N1+N2+...+NL)
Systematic sampling
simpler to design and administer
decide on sample size n from population size
N
choose starting point by selecting rth unit from 1
end of sequence, r is less than k, and is
chosen randomly
take the rest of the sample by adding k to r,
where k is an integer number to equal to N/n
do this repeatedly until we reach end of
sequence
Cluster sampling
Population of interest is divided into
representative clusters of individuals
Sample selected randomly from
clusters
Conducted when it is impossible or
impractical to draw from other types of
sampling
Multi-stage sampling
formula
assume population mean Y is to be
etimated from sample mean y by a
simple random sample of n0 units. If n0 is
much smaller than N, n0 is given by:
n0 = [tS/d]2
http://www.raosoft.com/s
amplesize.html
sampling terms
population - finite (or infinite) set of
'objects' whose properties are to be
studied in a survey
target population the population whose
properties are estimated via a sample;
usually the same as the 'total' population
sample a subset of the target
population chose so as to be
representative of that population
sampling unit
a member of the sample frame
a member of the sample
probability sample
any method of selecting a sample such that
each sampling unit has a specific probability
of being chosen
non-probability
a method in which sample units are collected
with no sample specific probability structure