Bridging Disciplinary Boundaries (January 11 - 14, 2007)
The positive relationship between HC and intentions to remain employed (IRE) among CW workers has been confirmed (Ellett, 2000; Ellett, Ellett, & Rugutt, 2003; Landsman, 2001). A core finding is that professional committment is the primary correlate of IRE. Another way to explore the complex role that HC plays in IRE is to develop linear combinations of various dimensions of HC to differentiate extreme groups classified by an IRE measure. This kind of study can inform one about the complexities of HC measures and their ability to differentiate between a priori IRE groups (an element of construct validation).
Methodology: The entire population (N=2190)of public CW direct services workers in Georgia was surveyed by mail in 2005. Participants completed two measures: the Revised Human Caring Inventory (RHCI) and Intent to Remain Employed (IRE)measure. Details of the development and reliabilities of these two measures can be found in Ellett (2000) and Ellis (2006). The RHCI and the IRE response format was a four-point Likert scale. Respondents were first divided into upper (n=222) and lower (n=216) quartiles, extreme IRE total score groups. The IRE was conceptualized as the dependent variable and six factored dimensions of the RHCI (Ellis, 2006)were conceptualized as an independent variable set. The DFA computed two kinds of information of interest: 1)the extent to which each RHCI dimension contributed to a best linear combination of the variables separating the two IRE groups and classifying cases; and 2)the percentages of correct, a priori classification of cases into the two groups.
Results: 786 surveys (36%) were available for analysis. The DFA identified a single best linear discriminant function with standardized canonical discriminant coefficients clearly strongest for the RHCI Professional Committment(.912),Receptivity(.619),and Personal Attachment(-.445)dimensions. The strongest classification coefficients for both lower and upper groups respectively were for Respect for Clients (2.57, 2.44), Committment to Clients (1.33, 1.36), Receptivity (1.19, 1.48). 77.2% of the total cases were correctly classified a priori. The percentage of cases correctly classified a priori in the lower and upper quartile groups was 74.5% and 79.7%.
Implications: The results provide criterion-related validity evidence for the RHCI and show that the RHCI is somewhat more accurate as a predictor for those with strong IRE in CW. One possible use of the RHCI is to screen new employees, with particular attention paid to Respect for Clients, Receptivity, and Committment to Clients scores. Those rating high on these dimensions would predictably be those with the strongest IRE. Future studies should be completed to link the RHCI to other important CW organizational variables (e.g., organizational culture and leadership, work morale, quality of supervision).