A comparison of analysis methods to estimate contingency strength.

Area of Science:

• Behavioral science
• Quantitative psychology
• Experimental analysis of behavior

Background:

• Estimating contingency strength is crucial in behavioral research.
• Various data analysis methods exist, but direct comparisons are scarce.

Purpose of the Study:

• To compare the precision and sensitivity of four contingency strength estimation methods.
• To evaluate methods using simulated data with programmed reinforcement schedules.

Main Methods:

• Applied exhaustive event-based, nonexhaustive event-based, concurrent interval, and concurrent+lag interval methods.
• Utilized simulated data with response-dependent and response-independent reinforcement schedules.
• Assessed correspondence with expected values and sensitivity to reinforcement manipulations.

Main Results:

• Both event-based methods aligned with expected values for response-dependent schedules.
• Event-based methods differed in sensitivity to response-independent reinforcement.
• Interval-based method precision varied by analysis type and schedule, with similar response-independent reinforcement sensitivity.

Conclusions:

• Event-based methods offer reliable estimates for response-dependent contingencies.
• Method selection requires consideration of schedule type and potential confounds.
• Recommendations are provided for optimal contingency measurement in behavioral research.