Graphing Data and Visual Analysis
Keith C. Radley, Evan H. Dart · Oxford University Press eBooks · 2019
As previously described, single-case design has several advantages in the evaluation of evidence-based practices and for the evaluation of the effects of interventions in applied settings. Following collection of data, data are typically graphed in order to determine the effects of an intervention on student behavior. However, recent research has determined that the manner in which graphs are constructed is likely to impact the decisions that visual analysts make regarding the effect of an intervention. As such, it is important that graphs be constructed in a manner that minimizes potential for error. This chapter describes quality indicators for graphs, and discusses analysis- and aesthetic-altering elements of graphs. In particular, the chapter describes two analysis-altering elements that must be considered when constructing graphs: scaling of the y-axis and the data points per x- to y-ratio (DPPXYR). Finally, the chapter describes how to conduct visual analysis. Six elements are discussed: changes in level, trend, and variability, consistency across similar conditions, overlap across adjacent phases, and immediacy of intervention effects.