Descriptive analysis of multiple response topographies of challenging behavior across two settings.

The team watched 23 different problem behaviors in two places, home and school. They used lag-sequential descriptive analysis. This means they timed what happened right before and right after each behavior.

All 14 clients had severe intellectual disability. Staff and parents kept their normal routines. No one ran an experiment. The goal was to see if simple ABC counts could guess the real function.

The lag-sequential guess matched the later brief experimental analysis in most cases. That is 86 percent accuracy.

Topography did not predict function. One girl’s head-hit, scream, and bite all served escape. Another boy’s same topographies got tangibles. You must assess, not assume.

Staats et al. (2000) later repeated the idea with stereotypy. They also saw one form, many functions. Both studies warn: treat the reinforcer, not the shape.

Jessel et al. (2019) took the speed goal further. Their 5-minute IISCA replaced the long descriptive watch. E et al. gave the first proof that quick data can work; Jessel made it even quicker.

Moss et al. (2009) sounds like a critic. They say contiguity is not contingency. True, but E et al. never claimed perfect cause. They used the pattern as a first draft, then tested it. The two papers fit together: watch, then verify.

You can start with a 20-minute naturalistic sample and already know where to aim your full FA. This saves client stress, staff time, and session space. When resources are tight, let descriptive lag-1 sequences write your first hypothesis, then run a brief test to confirm.