Double-blind Solomon Four-Group Design
Double-blind Solomon Four-Group Experimental Design · Also known as: double-blind S4GD, blinded Solomon design, double-blind four-group design, Solomon four-group with double-blind
The double-blind Solomon four-group design combines Richard Solomon's classic four-group structure — which isolates pretest sensitization effects — with double-blind blinding, ensuring that neither participants nor outcome assessors know group assignments. This combination yields high internal validity by controlling simultaneously for testing effects, expectancy bias, and experimenter influence, making it one of the most rigorous true experimental designs available.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use this design when you need to rule out both pretest sensitization (the testing effect) and expectancy or observer bias simultaneously — for example, in drug trials, educational intervention studies, or psychological treatment research where outcome assessments could be influenced by knowledge of group membership. It is appropriate when you can randomly assign participants to four conditions and can feasibly implement blinding for both participants and assessors. Do not use it when random assignment is impossible, when the outcome is self-evident to participants (making blind conditions implausible), when sample sizes are too small to power a 2×2 ANOVA adequately, or when resource constraints prevent running four separate groups. If pretest sensitization is not a plausible threat, a simpler double-blind RCT with a pretest–posttest design is more efficient.
Strengths & limitations
- Simultaneously controls for testing effects (pretest sensitization) and expectancy or observer bias, achieving exceptionally high internal validity.
- The 2×2 factorial structure allows direct estimation of the pretest-sensitization effect as an independent quantity, not just a nuisance to be eliminated.
- Double-blind blinding prevents performance demand effects in participants and measurement bias in assessors.
- Supports both within-group change analysis (groups 1 and 2) and between-group endpoint comparison (all four groups), providing multiple analytic angles.
- Well-accepted in clinical, psychological, and educational research as a gold-standard design when testing effects are a genuine concern.
- Requires approximately twice the sample size of a standard pretest–posttest RCT because participants are spread across four cells, each of which must be adequately powered.
- Logistical complexity is high: maintaining credible blinding across four groups simultaneously demands rigorous protocols and independent data management.
- The design is inefficient when pretest sensitization is not a realistic threat — in such cases the four-group structure adds cost without additional internal validity benefit.
- Double-blind blinding is not always feasible; for behavioral, educational, or surgical interventions, participants often infer their group assignment, degrading the blind.
- The 2×2 ANOVA interpretation requires that the interaction term be examined before interpreting main effects, which adds analytic complexity.
Frequently asked
How is the double-blind Solomon design different from a standard double-blind RCT?
A standard double-blind RCT uses two groups (treatment vs. control) and typically includes a pretest. The Solomon design adds two additional no-pretest groups, enabling the researcher to directly estimate whether the pretest itself affects posttest scores — something a two-group design cannot separate from the treatment effect. The double-blind component controls expectancy bias in both designs.
What statistical method should I use to analyze the results?
The standard approach is a 2×2 factorial ANOVA with Pretest (yes/no) and Treatment (active/control) as between-subjects factors and posttest score as the dependent variable. The interaction term tests whether the treatment effect differs depending on whether participants received a pretest — i.e., it directly quantifies pretest sensitization. Always inspect the interaction before interpreting main effects.
How large a sample do I need?
You need sufficient power in each of the four cells for the 2×2 ANOVA, including the interaction term (which typically requires more power than main effects). A commonly cited rough minimum is 20–25 participants per cell, giving a total of 80–100. Conduct a formal power analysis based on expected effect sizes before finalizing your sample target.
What should I do if maintaining a credible blind is difficult for my intervention?
If participants or assessors are likely to detect group assignment — common in behavioral, surgical, or educational interventions — consider whether single-blind blinding (assessors only) is achievable, and report this as a limitation. If blinding is not feasible at all, the Solomon four-group design without blinding is still valuable for controlling testing effects, but the design should then be labeled simply a Solomon four-group design rather than a double-blind variant.
When is it better to use a simpler design instead?
Use a simpler double-blind RCT (without the Solomon four-group structure) when pretest sensitization is not a plausible threat to validity — for instance, when the pretest uses an unrelated measure or when the measurement instrument is unlikely to alert participants to the study's focus. The four-group structure only pays for itself when the pretest is likely to be reactive.
Sources
- Solomon, R. L. (1949). An extension of control group design. Psychological Bulletin, 46(2), 137–150. DOI: 10.1037/h0062958 ↗
- Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi-Experimental Designs for Research. Rand McNally. ISBN: 978-0395307878
How to cite this page
ScholarGate. (2026, June 3). Double-blind Solomon Four-Group Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/double-blind-solomon-four-group-design
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Control Group Experimental DesignExperimental design↔ compare
- Factorial ExperimentExperimental design↔ compare
- Pretest-Posttest Experimental DesignExperimental design↔ compare
- Randomized Controlled TrialExperimental design↔ compare
- Solomon Four-Group DesignExperimental design↔ compare