Module Three    Selecting a Sample

Definitions
Sample of Convenience
Other Non-Probability Samples
Probability Samples 
Steps in Probability Sampling 
Volunteer Samples

Definitions 

A. Population
B. Sample
C. External Validity   

Sample of Convenience 

A. The researchers need to be sure they are an appropriate group = representative of the group to whom the results will be generalized 

B. The researchers need to report the demographics. This enables the consumer of the research to evaluate the generalizability of the results to his/her    population or sample of interest.

Three examples of variables that are commonly called demographic are: ages, gender, occupation, salary, city, and state of residence.

C. The results are not generalizable to other groups with much confidence; however, this is not as big a problem in exploratory research

Other Non-Probability Samples 

A. The researchers need to be sure they are an appropriate group = representative of the group to whom the results will be generalized 

B. The researchers need to report the demographics. This enables the consumer of the research to evaluate the generalizability of the results to his/her population or sample of interest. 
 

Check: Can you identify which paradigm is most likely to use non-probability samples. 
Give 2 reasons why researchers from this paradigm prefer non-probability samples. 
Give 2 specific examples of when a researcher from any paradigm might use this approach.

Probability Samples 

A. Simple Random Sampling -- all individuals in the defined population have an equal and independent chance of being selected as a member of the sample 

B. Linear Systematic Sampling -- select every nth person on list; can not use unless list is in random order 

C. Stratified Random Sampling -- assures that certain subgroups are represented either equally or in the proportions that exist in the target sample; must use the same proportions if want to compute means on entire sample 

D. Cluster Random Sampling -- unit of sampling is not the individual but rather a naturally occurring group of individuals; may not be able to use    conventional formulas for computing statistics and it is less accurate than simple random sampling

Steps in Probability Sampling 

A. Define the target population -- make this reasonable (not all first graders in the United States) 

B. Select the sampling technique 

C. Report the completion / return / response rate 

In the post-positivist paradigm, the best sample is: 
 

Representative of the population from which it was drawn = sampling procedure 

Of sufficient size that the researcher can be confident that if he drew another sample from the same population using the same procedure, his/her results would be the same 

Of sufficient size to allow for attrition 

Random = increases generalizability and permits use of inferential statistics 

Sampling Error -- difference between the characteristics of a sample and the characteristics of the population from which the sample was drawn. Sampling error increases as sample size decreases.

Volunteer Samples 

A. All samples are volunteer, however purely volunteer samples are likely to differ from non-volunteers 

B. Complicates the interpretation of results and their generalizability 

C. They should compare on as many variables as possible with target population 

D. Research indicates that volunteers are usually:

1. Better educated 

2. Of higher social class 

3. More intelligent 

4. Higher in need for social approval 

5. More sociable 

6. More unconventional 

7. Female 

8. Less authoritarian 

9. Less conforming