Content area
Full text
Abstract: Two studies regarding the graphic display of single case data are presented. First, measurement and graphical display of data from studies in highly-ranked journals in special education were assessed. Measurement of desirable behaviors rather than undesirable behaviors was common and most studies used sessions and percentages as x and y-values, respectively. Data presentation (graph ratios and number of data points and sessions per cm) was highly variable; average ratios did not align with previously-published recommendations. In the second study, 50 editorial board members of special education journals were surveyed to determine preferences for graphing ratios. Preferences did not align with recommended graphing practices and varied based on the number of sessions depicted on the graph.
Single case design (SCD) research plays a critical role in establishing evidence-based practices for individuals with developmental disabilities (Horner et al., 2005; Wong et al., 2015). Given its focus on the individual case and repeated measurement that is sensitive to change, SCD research aligns well with the goals of practice in special education and related fields (Ledford, 2018). Thus, assessing the nature and extent of SCD research available to support a specific practice is critical for improving outcomes for individuals with disabilities.
Visual analysis is the primary means by which SCD data are analyzed (Horner et al., 2005; Lane & Gast, 2013; What Works Clearinghouse [WWC], 2017). Visual analysis allows for formative decision-making (Ledford, Lane, & Severini, 2017) and can be used summatively to answer research questions. Specifically, SCD researchers should systematically assess for level, trend, and variability within and across conditions (WWC, 2017). Researchers should also examine consistency and immediacy of behavior change and degree of data overlap across conditions (Ledford et al., 2017; WWC, 2017). To facilitate accurate visual analysis, researchers should use appropriate visual display of their time series data (Morley & Adams, 1991; Tufte, 2001).
Several studies have shown that visual analysis of SCD data is dependent on a number of factors, including variability and magnitude of behavior change (Matyas & Greenwood, 1990), analyst expertise (Ledford & Wolery, 2013), and immediacy of behavior change (Lieberman, Yoder, Reichow, & Wolery, 2010). Factors that may increase the difficulty of SCD data analysis are the use of disparate graphing formats (Kubina, Kostewicz, Brennan, & King, 2017); number...