Files and Templates

This template, developed by Dr. Wayne Fisher (University of Nebraska Medical Center), his colleagues, and students, allows you to evaluate whether an intervention had a significant effect based on the degree to which intervention data points exceed the baseline mean and trend in a desired direction (increasing behavior for skills and decreasing behavior for behavior reduction interventions). Using the binomial distribution, it examines whether a sufficient number of intervention data points are above (for skills) or below (for behavior reduction) baseline mean and trend lines to judge the treatment effect as significant. This tool has been shown to produce more acceptable Type-I and Type-II error rates than the General Linear Model, Interrupted Time Series Analysis, and several visual analysis methods.  It plots the results on a graph for you and provides a table for determining statistical significance.

Article link:
J Appl Behav Anal. 2003 Fall; 36(3): 387–406. doi:  10.1901/jaba.2003.36-387

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1284456/