Description
Combines pretest-posttest and posttest-only groups to estimate an intervention's effect and whether the pretest itself changed responses.
Pretests can sensitise participants: answering questions about privacy before using a feature may make them notice privacy cues. Solomon's design uses four groups: pretest-treatment-posttest, pretest-control-posttest, treatment-posttest only and control-posttest only. A 2x2 analysis of posttest scores (treatment by pretested) estimates the treatment effect, the pretest effect and their interaction. It is the classic way to detect testing effects and remains relevant whenever UX research measures attitudes or knowledge before and after an intervention, such as onboarding, education or trust-building features.
Attributes
- Discipline
- Research Design & Rigor
- Loss type
- Extinct
- Goal
- Prove causation
- Stage
- Evaluation
- Data
- Quantitative
- Effort
- Weeks
Steps
- Define the intervention, the pretest and posttest measures.
- Randomly assign participants to the four groups, balancing group sizes.
- Administer pretests to groups 1 and 2 only; deliver the intervention to groups 1 and 3.
- Administer posttests to all groups under identical conditions.
- Analyse posttest scores with a 2x2 ANOVA (treatment x pretest) and check the interaction for sensitisation.
- Use pretest data from groups 1 and 2 for gain analysis only if no sensitisation is found.
Sources
- Solomon, An extension of control group design (1949), Psychological Bulletin
- Campbell & Stanley, Experimental and Quasi-Experimental Designs for Research (1966)
Also known as
- Solomon design
- Pretest sensitisation control
- Pretest-posttest quasi-experiments
- One-group pretest-posttest
- Non-equivalent control group pretest-posttest design
- NECGPP
- Before-after study with comparison group
- Static-group comparison
- Posttest-only control group
- One-shot case study
- Multiple time-series
- Pre-experimental and quasi-experimental designs (Campbell and Stanley)
Categories
- Research Operations
- Research Design & Rigor