Cognitive and Psychometric Analysis of Analogical Problem Solving

ISBN-13: 9780387973210 / Angielski / Miękka / 1990 / 237 str.

Isaac I. Bejar; Roger Chaffin; Susan Embretson

Opis

If one were to conduct an analysis of any profession the "ability to think analogically" is more than likely to be one of the requirements for success, be it an architectural studio, a research laboratory, a legal office, or a nuclear plant. Cognitive scientists are aware of the prominence of analogical reasoning in all forms of reasoning and learning, and have devoted substantial effort to ascer taining its nature. Test builders, like cognitive scientists, are aware of the cen trality of analogical reasoning and figure, correctly, that a test that samples a student's ability to think analogically may well be a good predictor of success in a variety of fields. This book is the result of a project to investigate analogical reasoning from both an individual differences and a cognitive perspective. The book is directed to both researchers and practitioners concerned with the nature and measurement of analogical reasoning. Cognitive scientists, linguists, psycholinguists, and natural language researchers will find the seman tic taxonomy and accompanying empirical results food for thought. Test devel opers will fmd it reassuring that performance on verbal analogy items is not just a reflection of the size of a person's vocabulary, and that tests can be designed according to principles, rather than assembled to satisfy a set of statistical speci fications. Psychometricians will find that content and response modelling can go together and that there are distinct benefits in approaching psychometric re sponse modelling from that integrative perspective."

Spis treści

1: Introduction.- Objectives of the Project and Overview of the Report.- 2: Theories of Memory Representation and Analogical Reasoning.- A Normative Model.- Words and Concepts.- Representation of Word Meaning.- Prototype Theory.- Feature Models.- Semantic Networks.- Schemata, Frames, and Scripts.- Summary.- Relations Between Concepts.- Relation Expressions.- Novel Relations.- Relation Similarity.- Summary.- Comparison of Relations.- Relation Element Theory.- Reasoning with Analogies.- Psychometric Models.- Spearman’s Model.- Sternberg’s Models.- Evans’ Model.- Pellegrino and Collaborators.- Embretson (Whitely) Models.- Relation Element Theory.- Reasoning Models.- Rumelhart and Abrahamson’s Model.- Gentner’s Structural Mapping Theory.- Holyoak and Thagard’s ACME Model.- Summary.- 3: A Taxonomy of Semantic Relations.- The Taxonomy.- Completeness of the Taxonomy.- A Higher-Order Classification of Semantic Relations.- Intensional and Pragmatic Relations.- The 10 Relation Families.- Precedents for the Dichotomy.- Empirical Evidence for the Dichotomy.- Previous Studies.- A Reanalysis of Whitely’s 1977 Data.- A New Sorting Study.- A Comparison of GRE Candidates With Different Majors.- Factor Analytic Evidence.- Earlier Work on Cognitive Dictionaries.- Uses of the Taxonomy.- Item Development.- Computer Generation of Items.- Theoretical Primitives.- Research.- 4: Description of the Item Pool.- Distribution of Deltas and r-Biserial.- Distribution According to Test Development Taxonomies.- Distribution According to Relational Taxonomy.- Relationship of the Taxonomy of Semantic Relations to Other Test Development Taxonomies.- Summary.- 5: The Effect of Vocabulary Level and Rationale Complexity on Item Difficulty.- Data and Procedures.- Regression of Delta on Stem and Key Frequency for All Items.- The Role of Complexity in Difficulty.- Summary.- 6: The Relationship Between Delta and r-Biserial.- The Negative Relationship As an Artifact.- The Relationship of Delta and r-Biserial for Other GRE Items.- Other Data Sets.- Recomputing r-Biserial.- Polyserial Analysis.- Summary.- 7: Expert Analyses of Analogy Items.- Method.- Results.- Data Analysis.- Prediction of Item Difficulty.- Raters’ Model of Item Difficulty.- Prediction of Delta.- Raters’ Performance and Type of Analogy.- Summary.- 8: Cognitive Processing and Item Difficulty.- Relationship Between Processing Demands and Psychometric Difficulty.- Subjects and Method for Experiments 1 and 2.- Experiment 1.- Experiment 2.- Lexical Overlap.- Practical Value of Process Variables in Modeling Difficulty.- Summary.- 9: Cross Validation: Analysis of Pretested Items.- Establishing Criteria.- Regression Results.- A Practical Assessment of Utility.- Predicting Low r-Biserials.- Summary.- 10: Summary and Conclusions.- The Study in Perspective.- Systematic Item Writing.- Psychological Response Modeling.- Processing Models.- Practical Considerations.- Further Research.- References.- Author Index.