Master of Arts (MA)
Visual analysis is the “Gold Standard” for single-subject data because of two assumptions: a low Type I error rate and consistency across raters. However, research has shown it less reliable and accurate than desired. Autocorrelation, variability, trend, lack of obvious mean shift, and differences in the physical presentation of graphs contribute to inconsistencies and higher error rates. Statistical analysis has been advocated as a judgmental aid to visual analysis, but an appropriate statistic has not been found. In the present study, the accuracy of Hierarchical Linear Modeling was compared to raters’ visual analysis of previously published data using Receiver Operating Characteristic curves. The statistic was established as a potentially useful judgmental aid; however, definite conclusions were hindered by low power.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Godbold, Elizabeth S,, "Hierarchical linear modeling against the "Gold Standard" of visual analysis in single-subject design" (2008). LSU Master's Theses. 4211.
Frank M. Gresham