Purpose: Designed to measure abilities related to mechanical comprehension.
Population: Individual applicants for mechanically-related jobs or schooling.
Score: Percentile score based on the appropriate normative population.
Time: (30) minutes.
Author: George K. Bennett.
Publisher: The Psychological Corporation.
Description: The Bennett Mechanical Comprehension Test (BMCT) measures a complex set of abilities composed of three primary facets or constructs: 1) mechanical information, 2) spatial visualization, and 3) mechanical reasoning or understanding. The knowledge measured by this test is based on common experiences, and special training seems to have little effect on test scores.
Scoring: The BMCT is scored by hand with a scoring template. The raw score is simply
the number correct. Raw scores are converted to percentile scores based on the appropriate normative population. Scoring instructions for the BMCT are clear and easy to master. Scoring the 68 items should take an experienced scorer no longer than 2 minutes. Interpretation of the BMCT is based on an objective raw score (number correct) that is converted to a percentile score. The percentile score represents the percent of a similar population that would have scored at or below the raw score attained by the candidate.
Reliability: The manual for the BMCT presents split-half reliability coefficients (corrected by the Spearman-Brown formula) of .81 to .93, with a median of .86. The standard errors of measurement range from 3.0 to 3.8. The manual indicates that the reliability of the BMCT is low because of the lower mean item intercorrelations, due to the variety of topics covered in the 68 items. Test-retest reliability data are not presented in the 1969 manual.
Validity: Content validity is not discussed in the manual because there is no specific job or curriculum against which to compare the test items. Construct validity is also not specifically addressed in the 1969 manual. A table of correlations of the BMCT with other standardized tests from an industrial data base is represented. For applicants for skilled trade jobs in an automobile company, the BMCT correlates .46 with the Personnel Test (mixed forms) and .55, .48, and .59 with the Wesman Personnel Classification Test (Forms A and B) Verbal, Numerical, and Verbal and Numerical scores, respectively. For applicants for electrical inspector trainee jobs in an aviation company, the BMCT correlated .57 and .68 with the Verbal and Numerical scales of the Personnel Tests for Industry (Forms A and B) and .15 and .04 with the Numbers and Names subtests of the Minnesota Clerical Tests.
Norms: To develop Forms S and T, 180 experimental items (95 existing items, 43 original items that had been modified, and 42 new items) were administered to 706 male students in Grades 11 and 12 in Philadelphia, Pennsylvania, and Kansas City, Missouri.
Suggested Uses: Recommended uses include the prediction of job performance.