Within-session patterns of responding during functional analyses: the role of establishing operations in clarifying behavioral function.

Kahng et al. (1999) watched how problem behavior changed minute-by-minute inside single FA sessions. They looked for jumps or drops that lined up with when reinforcers were turned on or off. The goal was to see if these tiny swings could clear up 'undifferentiated' FAs that had no clear winner.

When they tracked the moment-to-moment data, the picture snapped into focus. Behavior rose as soon as the reinforcer was removed and fell the instant it returned. These within-session shifts matched the true function and let the team move to treatment.

Jolliffe et al. (1999) published the same year and said FA results are shaky; test-retest scores were poor and different scorers picked different functions. The two papers seem to clash, but they don't: T et al. showed the whole tool can wobble, while S et al. gave a fix—zoom in on EO swings to steady the read-out.

Heald et al. (2020) later asked whether FAs just 'create' new functions. They found six of eight kids kept the same function even when extra reinforcer tests were run, backing S et al.'s idea that moment-to-minute patterns reveal real, pre-existing functions rather than lab artifacts.

Griffith et al. (2021) trimmed FA sessions to five minutes and still found the same escape function. That efficiency boost pairs nicely with S et al.'s message: you can run short sessions and still spot the EO-driven ups and downs that matter.

Next time your multielement graph looks flat, don't scrap the assessment. Graph the responses in one-minute bins instead. If you see a spike right after the reinforcer is removed and a drop when it returns, you have evidence for that function. This quick add-on can save you from extra sessions or a full re-do.