Size matters! Effect sizes for single-case experiments evaluating exposure in vivo for pain-related fear.

A single-case experiment (SCE) can be used to evaluate the effect of an intervention for a single person. Because a growing amount of research in the behavioural sciences uses SCEs, it is important to develop methods to accurately analyse and report them. One of these methods is the randomisation test, which is based on the random assignment of measurement occasions to treatments. Taken together, the randomisation test and the random assignment improve both the statistical conclusion validity and the internal validity of SCEs. Although a statistical significance test (like a randomisation test) provides interesting information to rule out the null hypothesis of no treatment effect, crucial additional information is provided by a measure of effect size. Our project consists of two objectives concerning effect sizes for SCEs. First, we will examine randomisation tests that use effect sizes as test statistics. Second, we will construct confidence intervals for effect sizes based on the randomisation test rationale. We will apply these two new analytic methods on single-case data collected for the evaluation of exposure in vivo therapy for chronic pain patients. The current project will focus on systematic reviews of SCEs for this kind of therapy and this type of patients, and on simulation studies, based on the results of the systematic reviews, to assess the performance of these two methods in a more general setting.