Blocked Solomon Four-Group Design
Blocked Solomon Four-Group Experimental Design · Also known as: Blocked S4G, randomized blocked Solomon design, Solomon four-group with blocking
The blocked Solomon four-group design combines Solomon's classic four-group structure — which disentangles pretest sensitization effects from treatment effects — with blocking on a known nuisance variable. Participants are first grouped into homogeneous blocks (e.g., by baseline ability, gender, or site), then randomly assigned within each block to one of four conditions: pretested treatment, pretested control, unpretested treatment, and unpretested control. This structure simultaneously controls for maturation, pretest reactivity, and block-level variance, making it one of the strongest quasi-controlled experimental frameworks 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 a blocked Solomon four-group design when (1) you need to detect a meaningful treatment effect AND suspect that administering a pretest may sensitize or alert participants in ways that alter their response to the treatment (pretest reactivity is a real concern in attitude, knowledge, or behavioral research); (2) a known nuisance variable (e.g., school, ward, ability tier) is likely to account for substantial outcome variance and you can form homogeneous blocks before randomization; and (3) you have enough participants to populate all four groups within every block (minimum 8 per block recommended for adequate power). Do NOT use this design when sample size is too small to sustain four cells per block — revert to a standard two-group blocked RCT; when pretest reactivity is not a plausible concern for your outcome measure, the Solomon extension adds participants and cost without benefit; or when repeated measurement of participants across conditions is infeasible.
Strengths & limitations
- Simultaneously controls for pretest sensitization, maturation, and block-level nuisance variance — addressing three major threats to internal validity in one design.
- Allows direct estimation of the pretest-by-treatment interaction, revealing whether the pretest modifies the treatment's effectiveness.
- Blocking reduces within-cell variance, increasing statistical power relative to an unblocked Solomon design with the same total N.
- Considered one of the most internally valid experimental structures available for behavioral and educational research.
- Requires substantially larger samples than a basic two-group design — four groups per block means at least 8–12 participants per block for minimally adequate power.
- Logistically complex: withholding the pretest from two groups can be difficult in applied or clinical settings where baseline assessment is standard practice.
- The blocking variable must be identified and measured before randomization; post hoc blocking is not valid.
- Analysis requires understanding of factorial ANOVA or ANCOVA with blocking factors, which is more demanding than a simple t-test or paired-samples analysis.
Frequently asked
Is the blocked Solomon four-group design different from a 2×2 factorial design?
Structurally it produces a 2×2 cell structure (treatment yes/no × pretest yes/no), which is analyzed like a factorial ANOVA. The Solomon design is sometimes described as a special case of the factorial design. The key distinction is conceptual: the Solomon design treats pretest administration as an explicit independent variable to test measurement reactivity — not merely a covariate or nuisance. The blocked version adds a third factor (block) to remove known nuisance variance.
How many participants do I need per block?
A minimum of 8 participants per block (2 per cell) is the absolute floor, but this gives very low power. For 80% power to detect a medium effect (d ≈ 0.5), aim for at least 20–25 participants per block. Conduct an a priori power analysis for a 2×2 factorial ANOVA to determine block size for your specific effect size and alpha level.
What if I cannot withhold the pretest from half my participants?
If ethical or institutional constraints require all participants to complete a baseline assessment, the Solomon design is not feasible. In that case, use a standard blocked RCT with the pretest as a covariate in ANCOVA, which partially (but not fully) addresses reactivity. The Solomon design is most defensible when the pretest is a research instrument, not a clinical necessity.
Can I use covariates in addition to blocking?
Yes. A blocked Solomon design can be analyzed as an ANCOVA that includes both block as a factor and additional continuous covariates (e.g., age, baseline score for the pretested groups). This hybrid ANCOVA approach is appropriate when some nuisance variables are categorical (handled by blocking) and others are continuous (handled by covariance adjustment).
When should I drop to a simpler design instead?
Drop to a simpler two-group blocked RCT when (a) pretest reactivity is not plausible for your specific outcome (e.g., physiological biomarkers are unlikely to be reactive to measurement), (b) your total sample is too small to populate four groups per block with adequate power, or (c) the logistical cost of four conditions exceeds the inferential benefit. The Solomon extension is worth its cost only when pretest sensitization is a genuine methodological concern.
Sources
- Solomon, R. L. (1949). An extension of control group design. Psychological Bulletin, 46(2), 137–150. DOI: 10.1037/h0062958 ↗
- Kirk, R. E. (2013). Experimental Design: Procedures for the Behavioral Sciences (4th ed.). Sage Publications. ISBN: 978-1412974455
How to cite this page
ScholarGate. (2026, June 3). Blocked Solomon Four-Group Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/blocked-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.
- Blocked Randomized Controlled TrialExperimental design↔ compare
- 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