The big idea
Both single-case and group designs can show that an independent variable causes a change. They differ in what they compare. A group design compares groups of people, usually by randomly assigning participants to treatment and control conditions and comparing average outcomes with statistics. A single-case design compares one person’s behavior across conditions over time (see D.4).
Behavior analysts favor single-case designs because practitioners treat individuals, and the question in practice is whether this intervention works for this person. Group designs remain the better tool for some questions, especially questions about populations. The exam expects you to know the strengths of each, not to dismiss either one.
Breaking it down
Strengths of single-case designs
- Effects are visible for each person. An average can hide participants who did not improve or who got worse. Single-case data show every individual’s response.
- Repeated measurement shows the process. Many data points per phase reveal trend, variability and the speed of change, not just a before-and-after score.
- Each participant is their own control. Differences between people cannot explain the result.
- Flexibility. The analyst can adjust a procedure, extend a phase or add a condition as data come in, which matches how practice works.
- Feasible in applied settings. A study can run in a clinic or classroom with a few clients and no untreated control group.
- Focus on meaningful change. Visual analysis tends to detect large, practically important effects rather than merely detectable ones.
- Generality through replication. Findings are extended by direct and systematic replication across people, settings and behaviors (see D.2).
Strengths of group designs
- Population estimates. A large random sample shows how an intervention affects a population on average and what proportion of people benefit, which matters for policy and funding decisions.
- Random assignment controls many threats to internal validity at once, including selection.
- Detecting small effects. Statistical tests can detect average effects too small to see on a graph.
- Questions that need between-group comparisons. When an effect is irreversible, takes years to appear, or cannot be turned on and off within one person, comparing groups may be the only option.
- Credibility in other fields. Medicine, education policy and funding agencies often weigh randomized controlled trials heavily.
Limits, side by side
| Concern | Single-case designs | Group designs |
|---|---|---|
| Individual outcomes | Shown directly | Often hidden in the average |
| Number of measures | Many per condition | Often few, such as a pretest and posttest |
| Generalizing to a population | Built slowly through replication | Built in through sampling |
| Small effects | Hard to detect visually | Detectable with large samples |
| Differences between people | Examined case by case | Treated as error and averaged out |
Seeing it in practice
- A behavior analyst wants to know whether a new transition routine reduces crying at drop-off for Leilani, age 4. A reversal or multiple-baseline design with Leilani answers the question directly, and the data guide her plan as it unfolds.
- A state agency asks whether funding a parent-training program will reduce out-of-home placements across the state. A randomized group trial with hundreds of families answers that population question better.
- A group trial shows that a social-skills curriculum produces a small average improvement. A clinician considering it for one client may still run a single-case evaluation, because the average says little about whether this client will benefit.
Common mix-ups
- Single-case designs are not less scientific. Both kinds of design can show experimental control. They answer different questions.
- Group results do not automatically apply to individuals. An average effect describes the population, not any particular person in it.
- Single-case is not a case study. It uses baselines, repeated measurement and replication; an anecdotal report does not.
- Statistical significance is not social significance. A reliable average difference can still be too small to matter for anyone.
Exam tip: If the question concerns whether an intervention works for a particular person, or needs ongoing data to guide decisions, think single-case. If it concerns the average effect in a population, very small effects, or outcomes that cannot be reversed or repeated within one person, think group design.
Quick recap
- Single-case strengths: individual effects, repeated measurement, within-person comparison, flexibility, feasibility in practice.
- Group strengths: population estimates, random assignment, detection of small effects, between-group questions.
- Single-case research builds generality through direct and systematic replication.
- Neither is better in general; the question decides.
Further reading
- Sidman, M. (1960). Tactics of scientific research: Evaluating experimental data in psychology. Basic Books.
- Horner, R. H., Carr, E. G., Halle, J., McGee, G., Odom, S., & Wolery, M. (2005). The use of single-subject research to identify evidence-based practice in special education. Exceptional Children, 71(2), 165-179.
- Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.), chapter 10.
Key terms
- Single-case experimental design (Single-subject design, Single-case design)
- A research design in which each participant's behavior is measured repeatedly under different conditions and compared with their own baseline, so each person serves as their own control. Experimental control is shown through prediction, verification and replication.
- Group design (Between-groups design)
- A research design that compares the outcomes of groups of participants, usually with random assignment to treatment and control conditions, few measures per person and inferential statistics to judge differences between group averages.
- Direct replication
- Repeating an experiment under the same conditions as the original. Repeating it with the same participant shows the effect is reliable; repeating it with new but similar participants shows the effect is not unique to one person.
- Systematic replication
- Repeating an experiment while deliberately varying one or more of its features, such as the setting, the participants, the target behavior or the person delivering the procedure. Each success extends the range of conditions under which the finding holds.
- External validity
- The degree to which a study's results hold beyond the specific participants, settings, behaviors and procedural details of that study. In single-case research it is established mainly through direct and systematic replication.
Check yourself
5 questions on D.5. Choose an answer to see whether you're right and why.
Show the answer
Answer: C. The question is about an average effect and a proportion across a population, which is what group designs with random assignment are built to estimate. A single-case design would show clearly whether the program works for the students studied, but not what share of 3,000 students would benefit.
Show the answer
Answer: D. Averages can hide individuals who did not respond or who got worse. Single-case designs show each person's data, which is what practitioners need when treating one client. Random assignment and population estimates are strengths of group designs.
Show the answer
Answer: D. With large samples, statistical tests can detect average effects too small to see in visual analysis. The other three choices describe strengths of single-case designs: individual data, flexibility and within-person comparison.
Show the answer
Answer: B. Because data are collected and inspected continuously, single-case researchers can modify procedures as they go, much as practitioners do. Random assignment, large samples and statistical control of between-person differences are features of group designs.
Show the answer
Answer: A. A reversal design needs the effect to come and go with the intervention. A two-year program with a lasting outcome cannot be withdrawn and repeated within one child, so comparing groups who did and did not receive it is a better fit. Group designs still need a comparison condition.
Adds a step to your study record. Saved in this browser only.