What internal validity means
Internal validity is the confidence that a change in the dependent variable was caused by the treatment. It differs from external validity, the generalisability of results. Campbell and Stanley (1963) classified the main threats.
Main threats
- History: outside events during the study, such as a tobacco tax rise during a quit-smoking programme.
- Maturation: natural growth or fatigue over time.
- Testing: a pretest improves posttest scores.
- Instrumentation: measures or observer standards change.
- Statistical regression: extreme scorers drift toward the mean.
- Selection: groups differ from the start.
- Attrition: certain participants drop out.
- Interactions with selection, diffusion of treatment and compensatory rivalry.
Controls
Random assignment equalises groups, control groups provide comparison, and posttest-only or Solomon four-group designs check testing effects. Standardised measurement, blinding and attrition analysis also help. Where randomisation is impossible, quasi-experiments such as nonequivalent control group and interrupted time-series designs reduce threats.
References
- Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi-Experimental Designs for Research. Rand McNally.
- Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Houghton Mifflin.
- Rubin, A., & Babbie, E. R. (2016). Research Methods for Social Work (9th ed.). Cengage.