P = The general population
U = An unrandomized group
O = The post-intervention observation

Type:

Non Research Design

What it is:

The simple observation really is not a research design. In this design, an unrandomized group is arbitrarily selected from the general population and observed and recorded.

Method:

Step One

Choose an arbitrary group from the general population "P" and designate this group as the experimental group "U".

Step Two

Observe the experimental group "U" and record the observations "O"

Step Three

Summarize, analyze, and report the results.

Threats to Validity

History

Lack of control introduces potential history risk.

Maturation

Lack of control introduces potential maturation risk.

Testing Effects

Not a factor in single test designs.

Instrumentation

Probably insignificant. Depends on the method chosen.

Statistical Regression

Not a factor in a single test design.

Selection

Results obtained are largely a function of the group selected.

Attrition

Lack of control introduces potential attrition risk.

Interaction

Must always be considered.

Remarks:

The critical flaw is that the results obtained depend mostly on the specific group selected for observation.

 

—By Douglas E. Greathouse, MA.(2007)