Chapter 13:

Single-n and Quasi-Experiments

I. Overview

II. Inferring causality in randomized laboratory experiments

A. Establishing temporal precedence

B. Establishing covariation

C. Controlling for irrelevant factors without keeping everything constant

III. Single-n experiments

A. Keeping nontreatment factors constant: The A-B design

B. Variations on the A-B design

1. The reversal design

2. Psychophysical designs

3. The multiple baseline design

C. Evaluation of single subject experiments

1. Internal validity.

2. Construct validity

3. External validity.

D. Conclusions about single subject experiments

IV. Quasi-experiments

A. The problem: To account for nontreatment factors

1. Identifying nontreatment factors: The value of Campbell and Stanley's spurious eight

2. Using logic to combat the spurious eight

B. The pretest-posttest design

C. Time series designs

1. Estimating the effects of threats to validity with a time series design

2. Problems in estimating the effects of threats to validity with a time series design

a. Inconsistent instrumentation effects

b. Inconsistent mortality effects

c. Inconsistent testing effects

d. Inconsistent maturation effects

e. History

f. The threat of regression

3. Eliminating, rather than estimating, threats to validity

4. Types of time series designs.

a. Reversal time series.

b. Two-group time series designs

D. The non-equivalent control group design

E. Conclusions about quasi-experimental designs

V. Concluding remarks

Summary

Key terms and issues

Exercises