The Application of Configural Frequency Analysis Using Stirling’s Formula to Single Case Designs
MetadataShow full item record
Single case experimental designs (SCEDs) have found their place among a range of fields including psychology, education, and medicine. SCEDs provide rigorous experimental evaluation of treatment effects. Currently, SCED evaluation methods, such as visual analysis and effect size estimation statistics, provide evidence for determining treatment effects. Although useful, best practices for SCED data analysis are debated. Configural frequency analysis (CFA) is introduced as a statistical method for the analysis of data from SCEDs. This research compared CFA statistical significance results to non-overlap SCED analysis methods that do and do not provide for statistical significance, as well as visual analysis methods. As there are currently no agreed upon best methods for the analysis of data obtained by SCEDs, it was important to explore additional statistical methods to aid researchers choosing to utilize SCEDs. CFA was compared to 5 non-overlap treatment effect evaluation methods that provide statistical significance and effect size values, 5 methods that only provide for effect size values, and visual analysis performed by 6 doctoral (PhD) students trained in SCEDs. A review of 23 years of The Journal of Behavior Therapy and The Journal of Behavior Modification resulted in 168 SCED data sets for comparison methods. Graphs were analyzed using each non-overlap effect size procedure as well as CFA. Results suggest that CFA aligned well with existing statistical significance calculations. Visual analysis appeared to align with simple non-overlap effect size methods rather than with CFA calculations, hinting at the importance of including statistical significance when evaluating treatment effects. Overall, this research found that CFA performed well when compared to other SCED data analysis techniques.
Patience, Marc A (2015). The Application of Configural Frequency Analysis Using Stirling’s Formula to Single Case Designs. Doctoral dissertation, Texas A & M University. Available electronically from