Psychomotor Abilities Tests as Predictors of Training Performance

Article excerpt


The present study examined the predictive and incremental validity of three psychomotor ability measures (manual dexterity, finger dexterity, and motor coordination) in Canadian Forces personnel (n = 209) being trained in technical and mechanical occupations. For both types of occupations combined, manual dexterity predicted training performance (r =.18), as it did for the Technical (r = .22) and Mechanical (r = .17) families, separately. Finger Dexterity (r = .02) and Motor Coordination (r = .05) did not predict training performance for either the combined group or for each family by itself. The addition of the three psychomotor measures increased validity beyond what was predicted by cognitive ability in the combined occupations (delta R^sup 2^ = .05); however, only Manual Dexterity (beta = .26) made a significant contribution to the regression model. Similar, though nonsignificant, increases in predictive validity occurred within each family. The results from this study suggest that specific abilities, when determined through a job analysis, improve predictions based solely on cognitive ability.


La presente etude s'est penchee sur la validite predictive et la validite differentielle de trois mesures de la capacite psychomotrice (dexterite manuelle, dexterite digitale, coordination motrice) chez des membres du personnel des Forces canadiennes (n = 209) qui recevaient une formation reliee a des emplois de nature technique et mecanique. Pour les deux types d'emploi combines, la dexterite manuelle a permis de predire le rendement en tours d'instruction (r =.18) comme ce fut le cas, d'une part, pour la categorie Technique (r = .22) et d'autre part, pour la categorie Mecanique (r = .17). La dexterite digitale (r = .02) et la coordination motrice (r = .05) n'ont pas permis de predire le rendement en tours d'instruction, ni pour le groupe combine, ni pour chaque categorie prise separement. Le fait d'ajouter les trois mesures psychomotrices a permis d'accroitre la validite et la faire depasser celle predite par les capacites cognitives dans le cas des emplois combines (delta R^sup 2^ = .05); toutefois, seule la dexterite manuelle (beta = .26) a contribue de facon signi

ficative au module de regression. L'accroissement de la validite predictive, semblable quoique non significative, a ate observe dans chaque categorie. Les resultats de cette etude laissent croire que des capacites precises, lorsqu'elles sont definies au moyen d'une analyse des emplois, ameliorent les predictions qui se fondent uniquement sur les capacites cognitives.

Selection and classification of personnel into training programs or jobs should be based on valid predictors that are related to job requirements (Childs, Baughman, & Keil, 1997). General mental ability or general cognitive ability (GMA, or g) predicts performance across a broad spectrum of jobs (Schmidt & Hunter, 1998). Noncognitive abilities (e.g., psychomotor and perceptual ability) that have been identified through job analytic procedures may improve the predictive validity provided by measures of g (e.g., Lubinski & Dawis, 1992; Schmidt & Hunter, 1998; Schmidt, Ones, & Hunter, 1992; Wise, McHenry, & Campbell, 1990).

Project A, a seven-year research program to improve the selection and classification into entry-- level occupations of the United States (U.S.) Army, established that separate components of job performance show unique patterns of relationships with various predictor measures (Campbell, 1990). Different mixes of skills, interests, temperament, and background may be needed to obtain optimal prediction across jobs (Wise et al., 1990). For example, while GMA was the best predictor of performance across nine different military occupations, the addition of specific ability tests improved prediction with respect to the job components of Core Technical Proficiency (i.e., job tasks specific to an occupation) and General Soldiering Proficiency (i. …