|
NCSU Institutional Repository >
NC State Theses and Dissertations >
Theses >
Please use this identifier to cite or link to this item:
http://www.lib.ncsu.edu/resolver/1840.16/1973
|
| Title: | Exploring the Dimensionality of Performance in the Army Special Forces |
| Authors: | Ellington, James Kemp |
| Advisors: | M.A. Wilson, Committee Member
J.W. Cunningham, Committee Chair
D.W. Drewes, Committee Member |
| Keywords: | performance dimensions
Army Special Forces
cluster analysis
factor analysis
attribute requirements
Q-sort method |
| Issue Date: | 14-Aug-2002 |
| Degree: | MS |
| Discipline: | Psychology |
| Abstract: | The purpose of this research was to explore the dimensionality of performance in the Army Special Forces (SF) using three separate data sources. A job analysis of the SF revealed 15 categories of performance common to all SF soldiers, in addition to 47 human attributes necessary to perform the behaviors associated with each category. A sample of SF soldiers was rated on their performance of the 15 categories using behaviorally anchored rating scales (BARS). Furthermore, a panel of subject matter experts (SME) rated each performance dimension on its requirement for each of the human attributes. These two data sources served in the development of two sets of performance dimensions, based on respective factor and cluster analyses of the data. A third data source included job analyst sorting data of the 15 performance categories into predefined numbers of groups. This data was also cluster analyzed to produce a third set of performance dimensions. Research questions focused predominantly on the number of performance dimensions that would be derived from the three data sources and their interpretability, and the extent to which agreement would exist between the three sources. Agreement was determined based on an index of cluster overlap for comparing two cluster solutions. Three dimensions were chosen as most interpretable from the performance ratings and a four-dimension solution was chosen based on the analyses of the attribute requirement ratings. The clusters derived from the analyst sorting data were difficult to interpret. The highest level of cluster overlap existed between the clusters derived from the performance ratings and the attribute requirement ratings. Overall, the three-dimension conception of performance was chosen as most meaningful. The dimensions were interpreted and named General Soldiering, Citizenship Performance, and SF Specific Performance. Comparisons were made with previously developed performance models, and recommendations were made for future research. |
| URI: | http://www.lib.ncsu.edu/resolver/1840.16/1973 |
| Appears in Collections: | Theses
|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
|
