By Drs. John Ashburn, Jr. and Mike Cherry
There are a tremendous number of coaching assessment tools currently in use. Commonly, these tools are broadly divided into assessments that measure various cognitive functions, and those that measure personality traits (Passmore, 2012). In this review we’ll be focusing on the latter and, more specifically, reviewing an article that analyzed the structure and elements of the Hogan Personality Inventory (HPI). The HPI is of particular interest to the writer because it is commonly used in my coaching population of interest, i.e., healthcare executives (Dr. Boysen, personal communication, May 28, 2021). This particular article was chosen because it reviews how higher order concepts can themselves be broken down into more specific elements, which itself reflects the scaling of the functionality of the assessment measure from broadband (i.e., applying generally to a wide client base) to narrowband (applying to a greater depth, but more narrow, client base).
The title of the article is the “Subdimensional Structure of the Hogan Personality Inventory” (Salgado, et. al., 2013). The HPI is based on the Five-Factor Model of personality (FFM) (the FFM itself will not be reviewed herein; for more information please see Costa and McCrae, 1992). Importantly, each of the five higher order factors can be broken down into sub-factors. The article reviewed a study of 200 applicants who were administered the Hogan Personality Inventory as a part of the job application process. Of note, the study was completed in Spain, and the job applications were for production worker positions with a multinational company in the wind power industry. The authors sought to identify and review the sub-factors of the HPI. To do this, they aggregated the results of the 200 applicants and performed various statistical analyses to unpack each of the higher order HPI factors into their component parts. The authors found that the major five factors of the HPI can be broken down into thirteen subdimensions, and that these are themselves made up of more specific homogenous item composites (“HIC’s”).
The study is significant for several reasons. As noted in the article (Salgado, et. al., 2013, p. 278), despite a large number of studies being completed on the HPI and its relationship to the FFM, there is little research on the HPI and the sub-dimensions of the FFM. This gap in the literature is significant for several reasons: 1) identification of the broader HPI structure would assist with understanding the extent to which the variables captured in the HPI go beyond the framework of the FFM; 2) the possibility that a better identification of the lower-order facets of the “Big Five” (i.e., those personality elements measured by the FFM) would facilitate the generation of improved tools to predict occupational performance in specific jobs; 3) there is considerable debate over the lower-order facets of the Big Five themselves (which research on the HPI could potentially inform); 4) identification of the HPI lower-order facets would assist in better understanding those who score in the middle range of the HPI, and 5) identification of the HPI lower-order facets would assist in teasing out the potential problem of multicollinearity, i.e., the extent to which the various scales share variance.
The study is noteworthy for generating findings that address some the issues noted above. While the statistical analysis itself is complex, the overall findings are straightforward (Salgado, et. al., 2013, see figure 1 p. 283). Ultimately, the authors identify thirteen sub-dimensions measured by the HPI, with several adjectives describing each sub-dimension. For example, the HPI higher-order element of adjustment/emotional stability has two sub-elements (i.e., emotional control and no anxiety), each of which has several facets (i.e., emotional control – empathy, even-tempered, and no somatic complaints; no anxiety – not anxious, no guilt, trusting, calmness, and good attachment). While there is some variability in the specificity of the descriptors, each is more specific and easier to measure from a behavioral perspective than the higher-order element of adjustment/emotional control. The results also lend themselves to establishing construct validity for the HPI which, as noted in class (Cherry, 2021), is an essential measure of any coaching assessment tool and which helps build on the research base of the HPI. Researchers in either the HPI or FFM would benefit from a review of the work the study authors completed.
The writer believes the study to be a valuable addition to the extant literature. Overall, the article is well-written and understandable. The statistics section is quite complicated, and because of the type of research involved there are numerous correlational matrices included for review. The basic analytic method utilized was an exploratory factor analysis, which involves uncovering (through reviewing the degree to which variables are correlated/overlapping) the underlying factors of a larger data set. While the section appears to be well-written, it is inherently complicated and is beyond the scope of most coaches (or psychologists) to objectively evaluate. Given the goals of the current study, the idea of doing an exploratory factor analysis seems reasonable; for those looking to review the in-depth statistical or methodological choices of the study, consultation with a statistician or research psychologist is recommended.
The study is also noteworthy for several generalizability limitations. While there is some cross-cultural research on both the FFM and HPI, there are outstanding concerns about the universality of the frameworks. Given that the study was completed in Spain, questions about foundational assumptions of the FFM have to be considered. Further, as the applications were for production worker positions, consideration has to be given to the possibility that individuals in different age and education strata would generate different results.
While not relevant for this particular article focusing on the HPI and FFM, there could be further research based on the data collected for this study. For example, the authors note (Salgado, et. al., 2013, p. 279) that in addition to the HPI, a cognitive measure (which was not specified) was administered as well. Evaluation of the relationships between the results of the HPI and the cognitive test could potentially build on the existing literature on this topic. For example, one study using twins found that there were statistically significant correlations between several facets (i.e., positive for openness to experience and agreeableness; negative for neuroticism) of another measure of the FFM, the shortened NEO-FFI, and intellectual quotient (IQ) (Bartels et. al, 2012). As the high-order FFM elements are essentially similar in the HPI and the NEO-FFI, the current study may be able to inform the findings of the Bartels et. al. study. If this were done consideration of several elements, including the similarity or difference of the tools used to measure IQ, as well as cultural factors, would have to specifically be addressed to reduce sources of error.
Bartels, M., van Weegen, F. I., van Beijsterveldt, C. E. M., Carlier, M., Polderman, T. J. C., Hoekstra, R. A., & Boomsma, D. I. (2012). The five factor model of personality and intelligence: A twin study on the relationship between the two constructs. Personality and Individual Differences, 53(4), 368-373. https://doe.org/10.1016/j.paid.2012.02.007
Cherry, M. (2021). 549 Week 2 Class Lecture Notes. Coaching Assessment and Research. Lewis University.
Costa, P.T., & McCrae, R.R. (1992). NEO-PI-R, Professional Manual. Odessa, FL: Psychological Assessment Resources
Passmore, J. (2012). Psychometrics in coaching: using psychological and psychometric tools for development (2nd ed.). Kogan Page.
Salgado, J.F., Moscoso, S., & Alonso, P. (2013). Subdimensional Structure of the Hogan Personality Inventory. International Journal of Selection and Assessment, 21(3), 277-285. https://doi.org/10.1111/ijsa.12037