Abstract:
This 1977 study was undertaken to establish validity and reliability for the Industrial Arts Aptitude Battery: Power Mechanics Test. The concurrent validation method was used. In this method a test which has shown validity is compared against the test to be validated. If a high relationship is found between the two, validity may be inferred. The Bennett Mechanical Comprehension and the DAT Mechanical Reasoning had both shown predictive validity for auto mechanics training in previous studies. The researcher elected to use the Kuder-Richardson formula 20 method to establish reliability. Six tests were administered to two groups of twenty students each. The experimental group consisted of twenty students in an auto mechanics training program at the Flint Hills Area Vo-Tech School in Emporia, Kansas. The control group consisted of twenty general education students from Emporia High School. The six tests were: (1) the Industrial Arts Aptitude Battery: Power Mechanics Test, Form A; (2) the DAT Mechanical Reasoning, Form L; (3) the DAT Abstract Reasoning, Form L; (4) the DAT Space Relations, Form L; (5) the Bennett Mechanical Comprehension, Form S; and (6) the Otis-Lennon Mental Ability Test, Form J. Intercorre1ations were calculated between all testing instruments using the Pearson Product-Moment Method. The Industrial Arts Aptitude Battery: Power Mechanics Test was compared against the DAT Mechanical Reasoning and the Bennett Mechanical Comprehension. The correlation between the Power Mechanics Test and the DAT Mechanical Reasoning (.82) was significant at the .01 level. The correlation between the Power Mechanics Test and the Bennett (.71) was also significant at the .01 level. Both relationships were considered high. The reliability coefficient was .97 for the Power Mechanics Test. The high relationship between the Industrial Arts Aptitude Battery: Power Mechanics Test, the DAT Mechanical Reasoning and Bennett Mechanical Comprehension establishes validity for the Power Mechanics Test. A reliability coefficient of .97 shows that 97 percent of the variance measured was due to the actual trait measured.
