Modeling Teachers' Judgment of Giftedness: a methodological inquiry of biased judgment1

Abstract

Previous empirical studies concerning the quality of teacher judgment in the domain of identifying gifted children have provided equivocal results. In this paper the process of identifying gifted children performed by teachers is conceptualised as a subjective decision strategy. Teachers' decisions are assumed to be based on students' characteristics that constitute the teachers' probabilistic concept of giftedness. Using a sample of 58 teachers experienced in the education of gifted school students, their preferences of relevant characteristics of giftedness for diagnostic use are compared to normative models and biased tendencies of how to use those characteristics diagnostically. The teachers' subjective concept of giftedness is characterised by aspects of cognitive competence and motivational preparedness for learning. The teachers' diagnostic strategies follow to some extent the normative Bayesian model of weighting evidence but are substantially influenced by representation bias and by forms of confirmation bias. The “base rate fallacy” biases teachers' diagnostic information processing, too. Other models of the diagnostic decision process which could be used as alternatives of the cue utilisation approach applied in this paper are discussed. 1 Data presented in this paper were collected when the author worked at the University of Munich. Thanks go to Kurt Heller, director of the Department of Educational Psychology, and to the Bavarian and Baden‐Wuerttembergian ministries of Education for their support of this study. Data evaluation and text processing were completed at the Max Planck Institute, with the help of Jane Dörfer and the statistical expertise provided by Eva‐Maria Fronk.