|
Now, you can buy | |
|
Add/share/save | |
| |
|
| |
| |
|
Separate Pre-Post Samples design
Explanations > Social Research > Design > Separate Pre-Post Samples design
Description | Example | Discussion | See also
Description
This method is used when groups are being examined but where successive measurements cannot be sure of measuring the same people. For example entire companies or gangs.
The approach is simply to take a different sample, from both the target and control groups, for pre-tests and post-tests.
In design notation, this can be written as:
| N1 | O | ||
| N1 | X | O | |
| N2 | O | ||
| N2 | O |
Example
A researcher is examining the effects of a management method in temp agencies. She selects two secretarial agencies for study, with N1 as the treatment agency and N2 as the control. An initial survey of a sample of temporary workers is done before the managers in the treatment agency are coached over several months. The subsequent post-test is of a different sample shows the retention rate has increased by a significant amount as compared with the control agency.
Discussion
The assumption on which this test is founded is that people within each of the overall groups being assessed (the target and control) are relatively similar.
Different samples are taken from the control group also to measure at the same time as the treatment groups and hence counter any effects such as learning and wider population shifts over time.
The Separate Pre-Post Samples design is a quasi-experiment.
See also
|
|