The effect of validating and invalidating evidence on construct organization

Lawlor, Mary Philomena (1977) The effect of validating and invalidating evidence on construct organization. Masters thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (3MB)

Abstract

This study tested Bannister's (1963) serial invalidation hypothesis that repeated invalidation of one's judgments loosens construct organization. Sixty university students judged three sets of eight career roles on 16 different constructs. The first eight constructs were termed target constructs since they were the focus of validation and invalidation during treatment. The second eight were termed non-target constructs since no feedback was provided for judgments involving these constructs. Subjects rated eight career roles (pre-treatment), then rated eight more receiving feedback on the eight target construct judgments for each individual career role (treatment), and finally rated eight more without feedback (post-treatment). Treatment consisted of four combinations of validating and invalidating levels of both qualitative and quantitative feedback types. One group was assigned to each of the four feedback conditions. Qualitative feedback consisted of evaluative comments by the experimenter on the subjects' performance. Quantitative feedback consisted of fake ratings shown to the subjects, from which they could assess the accuracy of their performances. -- Data were analysed separately for subjects with strong pre-treatment relations among constructs and those with weaker pre-treatment construct relations, since these groups tend to modify construct organization differently. For subjects with strong construct relations the results indicated that those who received totally invalidating feedback loosened construct relations during and after treatment significantly more than those who received totally validating feedback. This result held for both target and non-target constructs. The mixed feedback groups closely resembled each other with average scores on both the strength and consistency of relations which uniformly fell between the validated and invalidated groups. However, for subjects with weaker relations among constructs, there were no significant differences on any measure or any condition. Analysis of the effect of feedback type supported an additive as opposed to an interactive contribution. The evidence from this study strongly supports the serial invalidation hypothesis, but only for subjects with strong construct relations. Implications for the study of thought disorder were discussed.

Actions (login required)

View Item