Direct observation as a means of assessing frequency of maladaptive behavior in youths with severe emotional and behavioral disorder

Abstract

Objective: The purpose of this pilot study was to identify the incidence of maladaptive behavior in youths enrolled in community living and support training programs. Direct observation and partial-interval time sampling was used to determine occurrence of behaviors to develop effective strategies for group participation.

Method: Participants were 30 youths ages 10 to 17 and diagnosed with behavioral disorders. Four observers recorded incidence of eight maladaptive behaviors during three group occupations using direct observation across time and setting, twice per week for 8 weeks.

Results: Intraclass correlation coefficients (ICCs) were used to describe the interrater reliability of four observers in recording frequency of maladaptive behaviors. Interrater reliability was high to moderate for the most commonly demonstrated behaviors (ICC = 0.83-0.89). Lower agreement was observed in behaviors that were less frequent, such as violent episodes and sexual inappropriateness (ICC = 0.66-0.68).

Conclusion: Direct observation across time and setting demonstrated good-to-moderate interrater reliability in identifying frequency of behaviors that may interfere with group participation. Observation provides the opportunity to document occupation-based assessment without having to depend on youths' willingness and ability to answer questions.