Transcript Probability sampling

Probability sampling
-
Is the random selection of elements from the
population.
1. Simple random sampling
A- Identify the accessible population.
B- The development of the sampling frame
C- Enumeration of the all elements.
D- Selection of the sample elements.
Advantages:
1. Little knowledge of population is needed.
2. Most unbiased of probability methods.
3. Easy to analyze data and compute
errors.
Disadvantages:
1. A complete listing of population is necessary.
2. Time consuming.
3. Expensive.
2. Stratified random sampling:
-The population is divided into homogeneous
subgroups, or strata, according to some
variable or variables of importance to research
study, then a simple random sample is taken
from each of these subgroups.
- Proportional stratified sampling:
Involves obtaining a sample from each
stratum that is in proportion to the size of
that stratum in the total population.
Disproportional stratified sampling: used
whenever comparisons are sought
between strata of greatly unequal
membership size.
 Advantages:
a. Increase the precision and
representiveness of the sample.
b. Assures adequate number of cases for
subgroups.
 Disadvantages:
a. Requites accurate knowledge of
population.
b. Costly.
c. Statistics more complicated.
3. Cluster sampling:
-Is a random sampling of units “multistage
sampling”
 Advantages:
a. Saves time and money.
b. Arrangements made with small number
of sampling units.
 Disadvantages:
a. Larger sampling errors than other probability
samples.
b. Statistics are more complicated.
4. Systematic sampling:
The selection of every kth case from some list or
group.
K=N
n
N: the size of the population.
n: the size of the desired sample.
K: the sampling interval width.
Evaluation of probability sampling:
The superiority of probability sampling lies in its
avoidance of conscious or unconscious biases.
The great drawbacks of probability sampling are
its expense and inconvenience.
The larger the sample, the more representative of
the population it is likely to be.
The larger the sample, the smaller the sampling
error.
Guidelines for critiquing sampling
plans.
1. Is the target or accessible population identified
and described? Are the eligibility criteria clearly
specified?
2. Given the research problem and resource
limitations, is the target population
appropriately designated? Would a more
limited population specification have controlled
for important sources of extraneous variation
not covered by thy research design?
3. Are the sample selection procedures
clearly described? Does the report make
clear whether probability or nonprobalitiy
sampling was used?
4. How were subjects recruited into the
sample? Does the method suggest
potential biases?
5. Is the sampling plan one that is likely to
have produced a representative sample?
6. Did some factor other than the sampling plan
itself (such as a low rate of response) affect
the representative ness of the sample?
7. If the sampling plan is relatively weak (such as
in the case of a convenience sample), are
potential sample biases identified?
8. Are the size and key characteristics of the
sample described?
9. Is the sample sufficiently large?
10. Was the sample size justified on the
basis of a power analysis? Is another
rationale for the sample size presented?
11. If the sampling plan is relatively weak
(e.g., use of a small nonprobaility sample),
can the use of such a design be justified
on the basis of homogeneity of the
population the key variables?
12. To whom can the study results be
generalized?
Can the results of the study reasonably be
generalized to a broader population than
the one from which the subjects were
sampled? Does the report discuss
limitations on the study’s generalizability?