Contents

CHAPTER 9: Single-Case Designs and Small-n Research 299Key ConceptThe Multiple-Baseline Design The multiple-baseline design also makes use ofbaseline and treatment stages, but not by withdrawing a treatment as in theABAB design. As the name suggests, researchers establish several baselineswhen using a multiple-baseline design. The multiple-baseline design demonstratesthe effect of a treatment by showing that behaviors in more than onebaseline change following the introduction of a treatment.One example of the multiple-baseline design is to treat one person’s behaviorin different situations. In this case, the first step in the multiple-baselinedesign is to record behavior (such as the aggressiveness of a child) as it normallyoccurs in several situations (such as at home, in the classroom, and at anafter-school daycare facility). The researcher establishes the baseline frequencyof the behavior in each situation (i.e., multiple baselines). Next the treatmentis introduced in one of the situations (e.g., at home), but not in the other situations.The researcher continues to monitor behavior in all of the situations.A critical feature of the multiple-baseline design is that treatment is appliedto only one baseline at a time. The behavior in the treated situation shouldimprove; the behavior in the baseline situations should not improve. The nextstep is to apply the treatment in a second situation (treatment may continuein the first situation as well) but leave the third situation as a continuing baseline.Behavior should change only in the treated situation, not in the baselinesituation. The final step is to administer the treatment in the third situation;again, the behavior should change when the treatment is administered in thethird situation. The key evidence for the effectiveness of a treatment in themultiple-baseline design is the demonstration that behavior changes onlywhen the treatment is introduced.There are several variations on the multiple-baseline design, depending onwhether multiple baselines are established for different individuals, for differentbehaviors in the same individual, or for the same individual in differentsituations. Although they sound complex, multiple-baseline designs are frequentlyused and easily understood. We will describe each type of multiplebaselinedesign using an applied research example.In the multiple-baseline design across individuals, baselines are first establishedfor different individuals. When the behavior of each individual hasstabilized, an intervention is introduced for one individual, then for anotherindividual, later for another, and so on. As in all multiple-baseline designs, thetreatment is introduced at a different time for each baseline (in this case, for eachindividual). If the treatment is effective, then a change in behavior will occurimmediately following the application of the treatment in each individual.An example of the use of a multiple-baseline design across individuals comesfrom the field of sports psychology. Allison and Ayllon (1980) were interestedin evaluating the effectiveness of a coaching method that involved several behavioraltechniques on the acquisition of specific football, tennis, and gymnasticskills. Although they found that the method was effective for each sport, wewill describe their test of the effectiveness of behavioral coaching for the acquisitionof a football skill. The participants in this experiment were second-stringmembers of a citywide football program chosen because they “completelylacked fundamental football skills” (p. 299).