Download Multistage cluster sampling

Chapter 7
The Logic of Sampling
Probability sampling is the primary method for
selecting large, representative samples for
social science research.
Two types of Sampling
1)
Nonprobability sample
2)
Probability sample
Nonprobability Sampling





does not represent the population
qualitative based
study specific characteristics
poor reliability
good validity
Purposive or judgmental sampling


select sample on the basis of your own
knowledge of the population and the purpose
of the study
selecting deviant cases for study is another
example of purposive sampling
Snowball sampling




form of accidental sampling
convenient sampling
appropriate when the members of a special
population are difficult to locate
E.g. : prostitutes, steroid users, rape victims,
abusers, homeless
Snowball sampling


researcher collect data on the few members of
the target population they can locate, and then
ask those individuals to provide the
information needed to locate other members of
that population who they happen to know
exploratory purposes
Quota sampling


address the issue of representativeness
begins with a matrix or table describing the
characteristics of the target population -> you
need to know what proportion of the
population is for example male and female as
well as what proportions of each gender fall
into various age categories, education levels,
ethnic groups, etc.
Quota sampling


then you collect the data from people having
all the characteristics of a given cell
then assign weight to all the people in a given
cell that is appropriate to their portion of the
total population
Topic: Single Parents
White 5%
Male
20-30 years
Divorced
White 20%
Female
20-30 years
Divorced
Black 5%
Male
20-30 years
Divorced
Black 10%
Female
20-30 years
Divorced
Hispanic 10%
Female
20-30 years
Single
Hispanic 10%
Male
20-30 years
Single
Asian 10%
Male
20-30 years
Single
Asian 10%
Female
20-30 years
Divorced
White 20%
Female
15-20 years
Single
Informants



is a member of the group who can talk directly
about the group
select informants some what typical of the
groups you’re studying
informants will sometimes be marginal or
atypical within their group
The Theory and Logic of Probability
Sampling



generalize
representative
quantitative
bias

those selected are not typical or representative
of the larger population they have been chosen
from – can be unintended
representativeness and probability of
selection

representative -> sample will be representative
of the population from which it is selected if
the aggregate characteristic of the sample
closely approximates those same aggregates
characterized in the population
basic principle

all members of the population have an equal
chance of being selected in the sample ->
EPSEM (Equal Probability of Selection
Method)
2 Advantages



more representative than other types- biases
are avoided
estimate the accuracy or representative of
samples
based on random selection procedure
Element



unit about which information is collected and
that provides the basis of analysis -> people,
certain types of people
element used for sample selection
unit of analysis used for data analysis
Population

theoretically specified aggregation of the
elements in a study
Study population

aggregation of elements from which the
sample is actually selected
Random selection

purpose of sampling -> to select a set of
elements from a population in such a way that
descriptions of those elements (stats)
accurately portray the parameters of the total
population from which the elements are
selected
Key: Random Selection

each element has an equal chance of selection
independent of any other event in the selection
process
A sampling unit


element or set of elements considered for
selection in some stage of sampling
serves as a check on bias
Parameter


probability theory -> basis for estimating the
parameters of a population
parameters -> summary description of a given
variable in a population
Normal curve

probability theory enables us to estimate the
sampling error – the degree of error to be
expected for a given sample design -> standard
deviation
Confidence level and interval

- statistics fall within a specified interval from
the parameter
Sampling frame


the list or quasi list of elements from which a
probability sample is selected
Ex. : if a sample of students is selected from a
student roster, the roster is the frame
Types of Sampling Design
Simple random sampling
Basic sampling method assumed
Use sampling frame
↓
Researcher gives each element a number
↓
A table of random numbers used to select
elements (Appendix B)
Systematic sampling






every kth element in the total list is chosen
(systematically) for inclusion in the sample
should select the first element at random
ex. Select a random number between 1 and 10
the element having that number is included in the
sample, plus every tenth element following it
systematic sample with a random start
sampling interval is the standard distance between
elements selected in the sample
Stratified sampling



obtains greater degree of representativenessdecreasing the probable sampling error
rather than selecting a sample from the total
population at large, the researcher ensures that
appropriate numbers of elements are drawn
from homogeneous subsets of that population
can stratify, for example, by class, gender
Stratified sampling


the ultimate function of stratification is to organize
the population into homogeneous subsets (with
heterogeneity between subsets), and to select the
appropriate number of elements from each
the choice of stratification variables typically depends
on what variables are available – you should be
concerned primarily with those that are related to
variables you want to represent accurately
Multistage cluster sampling


cluster sampling may be used when it’s either
impossible or impractical to compile an
exhaustive list of the elements composing the
target population
involves repeating two basic steps: listing and
sampling
Multistage cluster sampling




the list of primary sampling units (churches,
blocks) is compiled and stratified for sampling
then a sample of those units is selected
the selected primary sampling units are then
listed and perhaps stratified
the list of secondary sampling units is then
sampled
Multistage cluster sampling



the general guideline for cluster design is to
maximize the number of elements within each cluster,
this scientific guideline must be balanced against an
administrative constraint
the efficiency of cluster sampling is based on the
ability to minimize the listing of population elements
by initially selecting clusters, you need only list the
elements composing the selected clusters