A comparison of analysis methods to estimate contingency strength
Event-based counts give the truest picture of contingency; interval methods need schedule-specific tweaks or they mislead you.
01Research in Context
What this study did
Lloyd et al. (2018) tested four ways to measure how tightly a behavior and its consequence are linked.
They used computer-made data sets so they already knew the true contingency.
Then they ran event-based and three kinds of interval analyses to see which came closest to the real number.
What they found
Event-based methods hit the expected value every time.
Interval methods sometimes came close and sometimes missed by a lot; the error changed with the schedule of reinforcement.
Picking the wrong interval subtype could give you a false sense of a strong or weak contingency.
How this fits with other research
Moss et al. (2009) warned that mixing up contiguity and contingency leads to wrong answers; Lloyd now shows one way to stay accurate—stick with event counts.
Thomson (1974) compared time-sampling rules and also found schedule-specific errors, echoing Lloyd’s warning that interval fixes must match the schedule you face.
Prasher et al. (2007) meta-analysis said the method you pick changes later treatment success; Lloyd gives you a rule of thumb within that bigger picture.
Why it matters
When you graph a scatter plot or run a contingency space, choose event recording if you can. If you must use interval data, check the reinforcement schedule first and pick the matching formula. One small switch keeps your FA results clean and saves you from chasing phantom functions.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open your last scatter plot, note the schedule, and re-run the analysis with the matching interval formula—or switch to event counts if the data are still available.
02At a glance
03Original abstract
To date, several data analysis methods have been used to estimate contingency strength, yet few studies have compared these methods directly. To compare the relative precision and sensitivity of four analysis methods (i.e., exhaustive event-based, nonexhaustive event-based, concurrent interval, concurrent+lag interval), we applied all methods to a simulated data set in which several response-dependent and response-independent schedules of reinforcement were programmed. We evaluated the degree to which contingency strength estimates produced from each method (a) corresponded with expected values for response-dependent schedules and (b) showed sensitivity to parametric manipulations of response-independent reinforcement. Results indicated both event-based methods produced contingency strength estimates that aligned with expected values for response-dependent schedules, but differed in sensitivity to response-independent reinforcement. The precision of interval-based methods varied by analysis method (concurrent vs. concurrent+lag) and schedule type (continuous vs. partial), and showed similar sensitivities to response-independent reinforcement. Recommendations and considerations for measuring contingencies are identified.
Journal of Applied Behavior Analysis, 2018 · doi:10.1002/jaba.463