Michael K. Lindell
Abstract
Recent research has suggested that outcome feedback is not the optimal form of feedback for learning complex ingerence tasks. The present experiment was designed to test the effects of outcome feeedback against cognitively oriented feedback in a range of linear tasks.
Preliminary analysis of the methodology of multiple cue judgment studies reveals two shortcomings. First use of the correlation coefficient, ra, as the basic dependent measure ignores inter subject differences in elevation and dispersion. This defect can be remedied by the use of a more general measure, Cronbackh's Acc2 index. Second, use of the R2s index as a measure of cognitive control is criticized as a likely source of artifacts. Solution of this problem involves the use of the varience of the residuals s2Zs rather than the more commonly usedR2s index.
The data show support for the superiority of cognitive over outcome feedback. Though found even in the simpler tasks, this superiority is greatest in the more complex tasks. Further, the effect of cue weight is shown to be a more important determinant of judgmental accuracy than cure intercorrelation.
The implications of these results for clinical judgment and for interpersonal learning are noted.
Ph.D. in Psychology
University of Colorado Boulder
1975
Committee Members
Kenneth R. Hammond
John Forward
Thomas Stewart
Daniel Bailey
Peter Polson
Michael K. Lindell is currently an emeritus professor at the Texas A&M University Department of Landscape Architecture and Urban Planning, an affiliate professor at the University of Washington Department of Urban Design and Planning, an affiliate professor at the Boise State University Department of Geosciences, and an affiliate Professor at the Oregon State University School of Civil and Construction Engineering. His primary research interest is household preparedness and response to environmental hazards.