Multiple functions of problem behaviors: assessment and intervention.
When one behavior works for two reasons, teach two different mands—one for each reason.
01Research in Context
What this study did
Researchers watched one boy with intellectual disability hit himself and others. They saw the hits happened for two reasons. Sometimes he wanted to escape hard tasks. Other times he wanted toys or snacks.
They ran a full functional analysis. It proved one behavior can serve two functions. Then they taught two different signs. One sign meant 'break please.' The other meant 'toy please.'
What they found
Problem behavior only dropped to low, safe levels after both signs were taught. Using just one sign left the second need unmet. Hits soon returned.
The team showed clinicians must match each function with its own communication response.
How this fits with other research
Donahoe et al. (2000) ran the direct follow-up. They compared one generic mand against two specific mands. Only the two-mand version wiped out problem behavior. This is a clean replication of the 1994 claim.
Weber et al. (2024) looked at 1,600 clinic cases. They found FCT works best when behavior has a single function. When escape is part of a mixed function, success drops. Their big-data view extends the 1994 warning.
Nasr et al. (2000) sharpened the rule. They showed kids only use the mand that fits the current need. If the 'escape' need is absent, the 'break' sign stays quiet. This supports teaching separate, function-matched responses.
Why it matters
Next time your functional analysis shows two reinforcers, write two separate communication goals. Program one mand for escape and one for access. Train both to mastery before you thin reinforcement. This simple step turns a good FCT program into a great one.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Check your client's latest FA. If two functions show up, add a second mand goal today.
02At a glance
03Original abstract
Three individuals with severe intellectual disabilities participated in separate analyses of problem behavior. In each case, a functional analysis was conducted under two parallel conditions. In one condition, self-injury or aggression resulted in escape from difficult tasks; in the second condition, the same problem behavior resulted in access to preferred items. Results indicated that the problem behaviors for each participant were maintained by both types of contingencies. Functional communication training was then delivered first in one condition and then in the second. After each participant was trained in a functionally equivalent mode of communication for one condition, levels of problem behavior decreased in that condition but not in the untrained condition. Only after separate communication forms were trained in both conditions was problem behavior reduced to clinically acceptable levels. These results document three examples of problem behaviors under multiple control, and emphasize the need to organize interventions that address different contingencies of reinforcement that maintain the same problem behavior.
Journal of applied behavior analysis, 1994 · doi:10.1901/jaba.1994.27-279