Effects of Plotting a Second Observer's Data on A-B-A-B Graphs When Observer Disagreement is Present

摘要

Plotting data from a second observer have been suggested as an alternative to presenting only mean agreement and ranges across sessions. One argument against this method of reporting is that plotting these data would not change analysis decisions by persons analyzing graphs. To empirically test this argument, we asked graduate students to analyze graphs with and without data from a second observer plotted alongside primary data. Results show that when these data are plotted: (a) observers do not change analysis decisions when both or neither paths show a functional relation, (b) observers change analysis decisions when only one path shows a functional relation, though this varied across graphs and participants, and (c) observers who did not change analysis decisions when only one path showed a functional relation often rated their decision with less confidence than when both paths showed a functional relation or when only summary statistics are reported. Findings suggest plotting data from a second observer may be helpful in making analysis decisions for A-B-A-B graphs.