Keywords
cognitive psychology, psychometrics, working memory, number series
Select the category the research project fits.
Social Sciences/Humanities
Is this submission part of ICaP/PW (Introductory Composition at Purdue/Professional Writing)?
No
Abstract
In intellectual ability assessments, series completion tasks are used as a measure of inductive reasoning. Individual differences in working memory capacity and general fluid intelligence are positively correlated with success on these tests, but research on how and why test-takers make mistakes is limited. The current research explored item-level performance on a frequently used number series test, focusing on the incorrect choices subjects made if they selected an incorrect response option. We also analyzed self-report strategies that test-takers used, specifically constructive matching and response elimination. A large sample of young adults from three universities completed working memory tasks, number series, and a retrospective strategy questionnaire. Working memory was correlated with number series performance, as expected, but only weakly correlated with number series strategy use. Individuals self-reporting higher use of a constructive matching strategy got more items correct on number series, and those self-reporting higher use of response elimination got fewer items correct on number series. Number series items varied in success rates, and on specific items, individuals were likely to pick a specific incorrect item. The strategy and error results suggest that, similar to other reasoning tests, the relationship between working memory and number series is accounted for general control processes.
Recommended Citation
Goulden, Katherine, "Working memory and reasoning: Analyses of errors and strategies in number series performance" (2019). Purdue Undergraduate Research Conference. 19.
https://docs.lib.purdue.edu/purc/2019/Posters/19
Working memory and reasoning: Analyses of errors and strategies in number series performance
In intellectual ability assessments, series completion tasks are used as a measure of inductive reasoning. Individual differences in working memory capacity and general fluid intelligence are positively correlated with success on these tests, but research on how and why test-takers make mistakes is limited. The current research explored item-level performance on a frequently used number series test, focusing on the incorrect choices subjects made if they selected an incorrect response option. We also analyzed self-report strategies that test-takers used, specifically constructive matching and response elimination. A large sample of young adults from three universities completed working memory tasks, number series, and a retrospective strategy questionnaire. Working memory was correlated with number series performance, as expected, but only weakly correlated with number series strategy use. Individuals self-reporting higher use of a constructive matching strategy got more items correct on number series, and those self-reporting higher use of response elimination got fewer items correct on number series. Number series items varied in success rates, and on specific items, individuals were likely to pick a specific incorrect item. The strategy and error results suggest that, similar to other reasoning tests, the relationship between working memory and number series is accounted for general control processes.