Assessment & Research

Statistical inference in behavior analysis: Friend or foe?

Baron (1999) · The Behavior analyst 1999
★ The Verdict

The stats-vs-visual fight is still on—know your journal before you add p-values.

✓ Read this if BCBAs who write or review single-case studies.
✗ Skip if Clinicians who only read finished graphs and never publish.

01Research in Context

01

What this study did

Meyer (1999) reports a panel debate. The topic: should single-case work use p-values?

Experts spoke for and against statistical inference. The paper sums up the arguments.

02

What they found

No side won. The field is still split.

Some say stats help show an effect. Others say visual analysis is enough.

03

How this fits with other research

Michael (1974) and Iversen (2021) both say skip the stats. They argue strong visual proof beats p-values.

Reid et al. (1999) disagrees in the same year. They say you can pool single-case numbers without selling out.

Nasr et al. (2000) extends the worry. They warn that any average hides individual curves.

04

Why it matters

If you submit to journals, you need to pick a side. Some editors want p-values. Others reject them. Read the debate first so your stats match the gatekeepers.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Check the author guide of the journal you plan to submit to—note if p-values are required or banned.

02At a glance

Intervention
not applicable
Design
theoretical
Finding
not reported

03Original abstract

Behavior analysts are undecided about the proper role to be played by inferential statistics in behavioral research. The traditional view, as expressed in Sidman's Tactics of Scientific Research (1960), was that inferential statistics has no place within a science that focuses on the steady-state behavior of individual organisms. Despite this admonition, there have been steady inroads of statistical techniques into behavior analysis since then, as evidenced by publications in the Journal of the Experimental Analysis of Behavior. The issues raised by these developments were considered at a panel held at the 24th annual convention of the Association for Behavior Analysis, Orlando, Florida (May, 1998). The proceedings are reported in this and the following articles.

The Behavior analyst, 1999 · doi:10.1007/BF03391983