Mary Jackson Pitts, Ph.D.

 

Hit Counter

mpitts@astate.edu

Home ] Up ]

Sampling

A sample "is a model or representation of the population"

Two types of samples

1)Non- probability

convenience samples,

similar/dissimilar case sample,

typical case samples,

critical case samples,

snowballing samples,

quota samples.

Non-probability samples are often used in exploratory research, when you cannot pin down a sample any other way.

Probability samples: They have external validity. They can be generalized to a larger population.

There are five major types of probability samples:

Simple random samples,

systematic,

stratified,

cluster, and

multistage.

 

 

 

 

 

Sample size:

1) efficient sample size,

2) implications of the design for efficient sample size,

3) implications of the sample size and design for subpopulation analysis,

4) adjustments for ineligibles and nonresponses,

5) cost of the data collection, and

6) credibility.

 

 

 

Sampling bias

An awareness of nonresponse bias, sampling bias, and sampling variability is also needed during the administration of the questionnaire and during the evaluation.

 

Nonsampling bias:

 

Sampling bias:

 

Sampling variability:

 

 

 

 

Sampling error

 

Standard Deviation can be calculated.

SD is the difference between the sample average and the population average.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Size:

We can ask how confident the client must be that the sample can be trusted to represent the population?

A confidence level of 95 percent is most often wanted.