Assessment of Self-Injurious and Aggressive Behavior

This chapter described structural (or topographical) and functional assessment methodologies that may be relevant for SIB and aggressive behavior in individuals with intellectual disability, with a focus on procedures described in the behavioral research literature. Structural and functional assessment techniques are not mutually exclusive and can complement each other. Each one has important functions, depending on the assessment task at hand. While functional assessment and analysis are designed to identify critical variables that maintain a given problem behavior in order to derive a rational strategy, structural assessment instruments are required for single-subject experimentation, treatment outcome evaluations, survey research and other group studies, and screening. While we were able to touch on only the most relevant features of assessment, we hope that we made it apparent that a variety of approaches and methods exist to obtain relevant information about SIB and aggressive behavior, ranging from simple and inexpensive techniques to sophisticated and expensive high-tech methods. Research into assessment technologies has already produced a considerable number of promising instruments and offers the researcher and clinician many options from which to choose. However, many of the available assessment instruments have limited evidence of their psychometric properties and need more research to highlight their strengths and correct their weaknesses. Some of the psychometric weaknesses are, in part, inherent in the nature of problem behavior in persons with intellectual disabilities and, therefore, probably do not have a quick fix. For instance, behavior problems tend to be of low incidence (10% prevalence or less) and are not normally distributed in the population. Typically, they are positively skewed, which means that only a (relatively) small number of individuals exists with high frequency and highly severe problems, often with multiple behavior topographies, while the majority of individuals does not exhibit any behavior challenges at all. Low-incidence skewed-distribution phenomena unfortunately create technical challenges for the construction of assessment instruments and for statistical data analysis. For instance, it is an axiom of classical test theory that scales with fewer items have lower reliability than scales with more items. Rare but clinically important behaviors such as SIB and aggressive behavior are typically represented by only one item in a scale, which consequently makes it all but impossible to measure them reliably. This problem has been encountered again and again in specialized rating scales, psychopathology scales, and in standardized maladaptive behavior scales. In addition, inferential statistics require normal distribution of variables, an assumption that tend to get violated in group based on such instruments. There is an ongoing debate over the best, most accurate, and feasible strategy to identify functional properties of target behaviors. While indirect measures such as functional interviews and informant rating scales and checklists are much less expensive and avoid some of the problems of EFA, their critics point out that they suffer from the drawbacks common to all indirect behavior assessments, namely, that they tend to be subjective and often unreliable. Some functional assessment tools fare better than others as far as their reliability and validity are concerned, and additional research is needed to provide more detailed and comprehensive information about the psychometric properties of indirect methods. This is especially true for behavioral interviews, which are relatively inexpensive, flexible, and seem to be favored by staff (Ellingson et al., 1999). This is also true for many of the existing functional assessment rating scales which, compared with EFA, have great advantages as far as cost and risk containment are concerned. EFA, on the other hand, which was found to be effective and useful by many behavior specialists (Ellingson et al., 1999), is still often considered the gold standard and criterion against which functional assessment instruments must be compared for the documentation of their validity. However, the precision, cost, and potential risk of an EFA must be carefully weighted against the lessened immediate risk and cost, but potential imprecision of an indirect assessment.