Self-Report and Strategy Use
Lisa Huffman, Ph.D.
Kathryn L. Fletcher, Ph.D.
Norman W. Bray, Ph.D.
Mailing address: Department of Psychology and Civitan International Research Center, SC 313, University of Alabama at Birmingham, Birmingham, AL 35294. Phone: (205) 934-9768, FAX: (205) 975-6330. Send Internet email to: bray@cis.uab.edu
University of Alabama at Birmingham
University of Miami
Laura Barnes
Jodie Malmberg
Tanya Shunnarah
Lisa A. Grupe
University of Alabama at Birmingham
Developmental change in the ability to reflect on knowledge is a central issue in cognitive development. The study of diverse aspects of cognitive development often depends on children's verbalizations concerning their cognitions. As noted by Brown (1987), developmentalists know surprisingly little about the conditions under which children provide valid verbal reports of their cognitive processes. The validity of verbal reports can be divided into the two related issues of veridicality and reactivity (Russo, Johnson, & Stephens, 1989). Veridicality refers to the extent to which the report corresponds to an ideal criterion measure. Reactivity refers to the extent to which requiring a child to provide a verbal report influences the event reported or other measures associated with that event.
A major limitation of the previous research is that virtually all of the studies of veridicality and reactivity have used indirect criterion measures rather than a direct comparison of reported and observed overt strategies. Not only have the studies of veridicality and reactivity of self-report in children used indirect measures, they have not examined these issues over a wide range of different age and intelligence groups. The present study was designed to examine the veridicality and reactivity of self-reports of children in a task that allows a direct comparison between observed overt memory strategies and the self-report of strategies across a wide range of developmental and intellectual differences.
One hundred twenty-seven typical children, 32 each of 7- , 9-, 11-, and 17-years of age participated, providing a wide range of developmental differences. Sixty-three children with mild mental retardation were also included to provide a wide range of intellectual differences, 32 11-year-olds and 31 17-year-olds. Half of the children in each age/intelligence group were assigned to a self-report condition while the other half was assigned to a no self-report condition.
Self-reports were investigated in the context of the development of external orientation strategies, which include pointing to, holding, or moving objects in an attempt to remember. Because these strategies can be easily observed, they provide a clear criterion for comparison to the child's self-report. Children listened to a story about a journey through a "haunted house." During the journey, children heard sequences of one to seven sentences and, at recall, arranged the objects according to the description provided ("The apple is above the ghost; The key is on the blue side of the apple"). For children in the self-report group it was explained that after they finished placing the objects for each sequence of sentences, he or she would be asked what they were thinking or doing to help them remember where the objects were.
The veridicality of self-reports was assessed by computing correlations between the number of reports for a particular strategy and the number of trials on which the strategy was observed. For external but not for verbal strategies, there were strong positive relationships between the observed strategies and the children's self-reports. These results demonstrate that there were strong relationships between self-report and observation of a strategy, indicating veridicality for children's report of a strategy.
Are self-reports reactive? The short answer is no, based on analyses conducted on the effects of self-report on frequency of external and verbal strategy use and accuracy of recall. Results indicated that for pointing, holding, moving, verbal strategies and for recall accuracy, there were no main effects for the self-report condition as compared to children who were not asked to give trial-by-trial self-reports.
The results of the present study indicate that across a wide range of developmental and especially intellectual differences, children's self-report of strategy use corresponded closely with the strategies observed.