Units of analysis and kinetic structure of behavioral repertoires.
The BPI-01 tells you a severe behavior exists, but direct observation tells you which one to treat first.
01Research in Context
What this study did
Thompson et al. (1986) compared two ways to spot early severe behavior.
They used the BPI-01 rating scale filled out by caregivers.
They also watched the same children with direct observation.
All kids had developmental delays and were under three years old.
The team asked: do both tools flag the same top problems?
What they found
Caregivers and observers agreed most of the time on whether aggression, SIB, or stereotypy happened.
Agreement topped seventy-five percent.
But the two sources ranked the exact behaviors differently.
A movement that looked minor to observers often topped the caregiver scale.
How this fits with other research
DeRoma et al. (2004) asked the same question in middle-school students.
They found even lower agreement on both rank and function.
The pattern is clear: older youth, less concordance.
Arpone et al. (2022) repeated the idea with parents and researchers.
They also saw poor to moderate agreement, proving the gap lasts past toddlerhood.
Sturmey (1994) later showed that most rating scales share the same weak spots the 1986 paper hinted at.
Why it matters
If you write a behavior plan from only a checklist, you may target the wrong topography.
Start with the BPI-01 or QABF to screen, then run a short direct observation sample.
Match the two lists, pick the behavior that lands high on both, and you will write sharper, faster interventions.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After you score the BPI-01, watch the child for ten minutes and compare the top three behaviors on both lists.
02At a glance
03Original abstract
Severe problem behaviors, like aggression, self-injury, and repetitive behaviors, in people with intellectual and developmental disabilities often appear during early development and may persist without early intervention. The frequencies of self-injurious behavior, aggression, tantrums, property destruction and stereotyped behavior among 17 infants and toddlers at risk for developmental delays and severe behavior problems were assessed using two methods: 1) direct observation of responses during 10 s partial interval recording during analogue functional analysis and 2) the Behavior Problem Inventory-01 (BPI-01; Rojahn et al, 2001), an informant rating scale. Analogue functional analysis results suggested that the most common function for problem behavior was automatic reinforcement, followed by negative reinforcement in the form of escape from demands. Agreement across the two types of measurement systems as to occurrence of the behaviors reported on the BPI-01 and direct observations during analogue functional analyses was greater than 75% across aggression, self-injury, and stereotyped behavior. Agreement at a more molecular level of the ranking of the most commonly occurring specific behaviors was considerably lower. Results are discussed in terms of future research on identifying conditions that set the occasion for high levels of agreement between indirect and direct measurement systems for severe behavior problems.
Journal of the experimental analysis of behavior, 1986 · doi:10.1901/jeab.1986.46-219