Development and psychometrics of a freely available mechanical aptitude test

• Autores: Diana Bairaktarova, David Reeping
• Localización: The International journal of engineering education, ISSN-e 0949-149X, Vol. 35, no. 6 (Parte A), 2019, págs. 1839-1850
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Mechanical aptitude refers to individual differences in understanding and learning how simple machines work. Tests of mechanicalaptitude are predictive of performance in engineering jobs and capability to learn about mechanical processes. The advancement oftechnology has led to existing mechanical aptitude tests becoming dated. Commonly used tests are known to be gender-biased,limited in use (diagnostics tests), and are not freely available for use in educational settings. This work presents the development ofa Mechanical Aptitude Test (MAT). The mechanical aptitude items were designed and tested across four phases in large samples ofengineering and non-STEM students across four U.S. universities (n = 1,718). An item analysis of the last phase (n = 599) wasconducted to screen questions not meeting established criteria for item difficulty and item discrimination. After, a one-factorconfirmatory factor analysis was run with diagonalized weighted least squares. The one-factor confirmatory factor analysis fit wellwith exceptional fit indices (CFI = 0.994, TLI = 0.993, RMSEA = 0.02 90%CI[0.004,0.3], SRMR = 0.059), albeit a rejected modelchi-square, x2(34) = 146.939, p = 0.042. The current MAT scale consists of 17 multiple choice items, narrowed down from a largerbank of 68 items, covering topics related to mechanical insight, mechanical knowledge, shop geometry and measurement, and toolknowledge.