NotesFAQContact Us
Collection
Advanced
Search Tips
Peer reviewed Peer reviewed
PDF on ERIC Download full text
ERIC Number: ED563061
Record Type: Non-Journal
Publication Date: 2013
Pages: 8
Abstractor: ERIC
Reference Count: 49
ISBN: N/A
ISSN: N/A
Effect Size Measure and Analysis of Single Subject Designs
Society for Research on Educational Effectiveness
One of the vexing problems in the analysis of SSD is in the assessment of the effect of intervention. Serial dependence notwithstanding, the linear model approach that has been advanced involves, in general, the fitting of regression lines (or curves) to the set of observations within each phase of the design and comparing the parameters of these lines (or curves). In the simplest case of an AB design, this involves fitting a regression lines to the observations in phases A and B and comparing the slopes and intercepts. In the event the slope is zero in the two phases, the comparison reduces to the comparison of the means or levels. If there is a trend but the slope is the same in the two phases, then the adjusted means (levels) are compared. A problem arises if the levels and the trends in the two phases are different. In this case the intercepts and slopes are compared across the phases. While statistically such comparisons are not problematic, the interpretations of treatment effects that arise from these comparisons are not intuitive. A "combination" of the slope and intercept parameters that yields a single measure of treatment effect could be more meaningful. Such a measure of treatment effect yields, in turn, an effect size that can be readily computed and interpreted. To this end, the study deals with: (1) Development of a measure of the effect of intervention which leads to an effect size measure; and (2) Development of statistical procedures for assessing intervention effects that take into account the serial dependence in single-subject designs. The statistical procedure described here that includes a definition of treatment effect and consequently an effect size provides a viable method for the analysis of single-subject designs. The advantage of the procedure outlined here is that a single measure of treatment effect and consequently a single effect size measure is available when there is both a level change and trend change in the phases.
Society for Research on Educational Effectiveness. 2040 Sheridan Road, Evanston, IL 60208. Tel: 202-495-0920; Fax: 202-640-4401; e-mail: inquiries@sree.org; Web site: http://www.sree.org
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Society for Research on Educational Effectiveness (SREE)