The effects of three levels of treatment integrity (100%, 50%, and 0%) on child compliance were evaluated in the context of the implementation of a three-step prompting procedure. Two typically developing preschool children participated in the study. After baseline data on compliance to one of three common demands were collected, a therapist implemented the threestep prompting procedure at three different integrity levels. One integrity level was associated with each demand. The effects of the integrity levels were examined using multielement designs. The results indicate that compliance varied according to the level of treatment integrity that was in place.DESCRIPTORS: noncompliance, preschool children, three-step prompting, treatment integrityTreatment integrity refers to the extent to which a treatment is implemented as designed. In research, behavioral interventions are likely to be implemented with near-perfect integrity, but in practice settings this is not always so. A number of studies have examined the effects of behavioral interventions when implemented at less than perfect levels of integrity. For example, Northup, Fisher, Kahng, Harrel, and Kurtz (1997) evaluated varying levels of integrity for a differential reinforcement plus time-out procedure. Appropriate behavior was reinforced on 100% of occasions, 50% of occasions, or 25% of occasions. In addition, time-out was implemented for aberrant behavior using these same values. Results showed that intervention effects were maintained at 100% integrity levels even when time-out was implemented at 50% integrity. Vollmer, Roane, Ringdahl, and Marcus (1999) examined differential reinforcement of alternative behavior (DRA) at varying integrity levels. They found that after exposure to DRA at 100% integrity (i.e., appropriate behavior was reinforced each time it occurred; problem behavior was never reinforced), lower levels of integrity did compromise intervention effects. However, participants exhibited a general bias toward appropriate behavior during many of the varying integrity levels, presumably because their recent history with the 100% integrity phase predisposed them to engage in appropriate behavior.Although these studies provide a start, more research on the effects of behavioral interventions at varying levels of integrity is needed. The purpose of this study was to examine the effects of varying levels of treatment integrity on child compliance in the context of the implementation of a multistep prompting procedure (Horner & Keilitz, 1975). This study extends the existing research on treatment integrity of behavioral interventions by conducting a parametric analysis of an antecedent-based intervention (prompting) as opposed to a consequencebased intervention (reinforcement).
METHOD
Participants and SettingJake (a 4-year-old boy) and Cara (a 4-yearold girl), the first 2 children nominated for the study, participated. Neither of the participants had a psychiatric diagnosis or a developmental Requests for reprints should be sent to David A. Wilder,