Section 5: Assessing Program Impact

Selecting a Design

The prospect of selecting a design can seem daunting. Worksheet 5.1, located at the end of this section, has been prepared to assist the evaluator to narrow the choices. Answers to the questions on Worksheet 5.1 can be compared with the statements listed below. These answers will guide evaluators to the most appropriate choices.

• If a comparison group called for, use either a true- or quasi-experimental design. If a comparison group is unnecessary, a non-experimental design can be used.
• If the choice is to use a pretest and posttest, the following sections apply:

WITH a comparison group, go to

• true experimental design, or
• quasi-experimental designs
  • non-equivalent control group design with pre- and posttest

WITHOUT a comparison group, go to

• non-experimental designs
  • single group pretest-posttest design

• If there are multiple assessment points, the following sections apply:

WITH a comparison group, go to

• true experimental design, or
• quasi-experimental design
  • time series with non-equivalent control group

WITHOUT a comparison group, go to

• non-experimental design
  • single group time series design

The questions and answers above can be referred to during examination of the following descriptions of evaluation designs. The pros and cons of each relevant approach should be discussed with the program's evaluation team. The design must also be compatible with resource constraints, the characteristics of the target population, and the proposed measurement instruments and procedures.