| Abstract: | The accuracy of interpretation of single subject data was evaluated by comparing raters' visual analysis of behavior change with statistical determination of behavior change using the split-middle methods of trend estimation. In addition, the effects of rater background (undergraduate, graduate student) and characteristics of the fictitious student (gender, label) on the ratings were determined. Results indicated that the accuracy of visual analysis was less than chance (47%). Rater and student characteristics, for the most part, did not affect the ratings. Implications of these results are discussed. © 1997 John Wiley & Sons, Inc. |
