Comparing functional assessment methodologies: a quantitative synthesis.

Prasher et al. (2007) pooled every study that compared two kinds of functional assessment. One kind uses brief tests where the team manipulates the environment. The other kind uses interviews, checklists, or simple watching without any changes.

They then looked at how well each kind predicted the best treatment later. They also checked if the form of the behavior (hitting vs yelling vs rocking) matched the reason behind it.

Experimental assessments led to stronger treatments. When teams ran short tests, the later plans cut problem behavior more than plans built on interviews alone.

The review also showed that certain behaviors usually pair with certain functions. For example, self-injury often earns escape, while disruption often earns attention.

Jessel et al. (2022) extends this work. They show you can keep the power of experimental tests but make them safer. Open-contingency formats cut dangerous episodes without losing clear answers.

Wunderlich et al. (2022) adds a warning. The neat function-to-behavior links Prasher et al. (2007) found do NOT hold for stereotypy. Subtype labels from FA did not predict which stereotypy treatment worked best.

Petry et al. (2007), published the same year, ran a small classroom study. They found paired-choice quizzes matched full FA results for most kids. This single-case data lines up with the larger trend: quicker experimental tools still give valid leads.

Pick the strongest assessment you can safely run. If time, staffing, or risk block a full FA, swap in an open-contingency or paired-choice version instead of jumping straight to an interview. You keep the treatment boost shown in the meta-analysis while staying within real-world limits. Always re-check; stereotypy and other automatic behaviors may need extra rounds.