Statistical Power of Piecewise Regression Analyses of Single-Case Experimental Studies Addressing Behavior Problems

In intervention research, single-case experimental designs are an important way to gain insights into the causes of individual changes that yield high internal validity. They commonly applied examine effectiveness classroom-based interventions reduce problem behavior in schools. At same time, there is no consensus on good design characteristics when dealing with behavioral problems Moreover, specific challenges arise concerning appropriate approaches analyzing data. Our study addresses interplay between test power piecewise regression analysis and specifications research designs. Here, we focus influence following designs: number measurement times, initial frequency behavior, effect, data trend. We conducted a Monte-Carlo study. First, simulated datasets were created conditions based reviews published studies. Following, analyzed using Poisson-regression models, was investigated. results indicate regressions have potential adequately identifying effects for strongly related study: Few especially phase A, low frequencies make it impossible detect even large effects. Research times show robust power. The gained highly relevant researchers field, as decisions during early stage conceptualizing planning studies may impact chance identify existing effect process correctly.