Using Risk Ratios to Quantify Potential Behavior-Environment Relations

Joslyn et al. (2024) wrote a how-to paper. They show you how to turn everyday ABC data into risk ratios.

The authors walk through five real cases. Each case shows the math step-by-step.

No new experiment was run. The goal is to give clinicians a quick number that says how much more likely a behavior is in one condition versus another.

The tutorial shows that risk ratios are easy to compute. You only need counts from your baseline and test sessions.

A ratio of 1 means no change. A ratio of 2 means the behavior is twice as likely in the test condition.

The paper gives simple rules: ratio above 2 or below 0.5 flags a strong link worth following up.

Manolov (2026) also wants to make single-case stats simple. That paper reviews free websites that give p-values. Joslyn adds risk ratios, so you now have two friendly options: a ratio or a website p-value.

Caron (2024) warns that log ratios can break when you have zeros. Joslyn’s risk ratio can break too if you divide by zero, but the tutorial shows how to add a tiny count to fix it. Both papers agree: never add big fake numbers like 0.5.

Jacobs (2019) and Weaver et al. (2019) push randomization tests instead. Those tests give a p-value but need more data points. Risk ratios work with as few as five sessions, so you can start sooner and still stay objective.

You can calculate a risk ratio today with any ABC sheet. The number tells you, at a glance, if the behavior really changes with the environment. Use it to check your visual guess before you write the FBA or start an intervention.