Browsing by Author "Cowgill, Marc C."

Now showing 1 - 2 of 2
  • Thumbnail Image
    Behavioral specificity and reliability in job analysis and job specification
    Cowgill, Marc C. (Virginia Tech, 1991-12-05)
    Job analysis, narrowly defined, refers to the collection of data describing job-related behaviors and the characteristics of the job environment. Job specification refers to the process of inferring required traits or abilities necessary for a desired level of job performance. Differences in the judgmental processes involved in these two functions were explored by (a) investigating the potential schema- or stereotype-based nature of job specification ratings, and (b) assessing the relationship between behavioral specificity and interrater reliability. These concerns were investigated through the use of 3 groups of subject raters: one group possessing extensive job knowledge, one group possessing some degree of job familiarity, and one group possessing little or no job knowledge. All subjects completed a job analysis instrument (the Job Element Inventory) and a job specification instrument (the Threshold Traits Analysis; TTA). Contrary to predictions, little evidence was uncovered to suggest extensive schema-usage on the part of TTA raters. In addition, the 2 instruments achieved similar levels of interrater reliability among the 3 subject groups. However, marginal support was found for the notion that behaviorally specific items generate higher reliability the less-specific items, and in replication of previous findings, job-naive raters were found unable to achieve the reliability of subject matter experts. Suggestions for future research are offered.
  • Thumbnail Image
    A Genetic Algorithm Approach to Cluster Analysis
    Cowgill, Marc C.; Harvey, Robert J.; Watson, Layne T. (Department of Computer Science, Virginia Polytechnic Institute & State University, 1998-08-01)
    A common problem in the social and agricultural sciences is to find clusters in experi- mental data; the standard attack is a deterministic search terminating in a locally optimal clustering. We propose here a genetic algorithm (GA) for performing cluster analysis. GAs have been used profitably in a variety of contexts in which it is either impractical or impossible to directly solve for a globally optimal solution to complex numerical problems. In the present case, our GA clustering tech- nique attempted to maximize a variance-ratio (VR) based goodness-of-fit criterion defined in terms of external cluster isolation and internal cluster homogeneity. Although our GA-based clustering algorithm cannot guarantee to recover the cluster solution that exhibits the global maximum of this fitness function, it does explicitly work toward this goal (in marked contrast to existing clustering al- gorithms, especially hierarchical agglomerative ones such as Ward’s method). Using both constrained and unconstrained simulated datasets, Monte Carlo results showed that in some conditions the ge- netic clustering algorithm did indeed surpass the performance of conventional clustering techniques (Ward’s and K-means) in terms of an internal (VR) criterion. Suggestions for future refinement and study are offered.