The purpose of this study was to examine the distinctiveness of interest inventory scores. The researcher studied the difference between scores on the Self-Directed Search (SDS) in an effort to determine when a difference between two scores represents a significant and meaningful difference. Researchers commonly use the standard error of measurement (SEM) to determine confidence intervals for individuals’ true scores. The SEM for the SDS equals eight points, which means two scores must be separated by at least eight points to be considered distinct. Over time the SEM has become known as the “rule of eight”, and practitioners use it as a guideline for interpreting SDS results. However, researchers determined the SEM from a statistical formula, and no study has empirically examined this guideline. This study examined the distinctiveness of interest inventory scores by calculating the difference between individuals’ highest two SDS scores, while comparing congruence between two concurrent measures of vocational interest, both taken from the SDS.
SDS data was collected for 2397, (1497 female and 900 male), undergraduate students enrolled in the exploratory major at a large Midwestern university between 1996 and 2002. Primary-code distinction represented the absolute difference between the top two SDS scores. Expressed vocational interest and inventoried interest were compared to determine whether or not a congruent match existed between the two. Congruence results were grouped by level of primary-code distinction in an effort to determine when distinction scores represent a meaningful difference.
Descriptive statistics suggest a positive relationship exists between primary-code distinction and congruence. Furthermore, distinction scores of four points appear to distinguish between individuals’ top two SDS scores. Logistic regression confirmed the existence of a significant positive relationship between primary-code distinction and congruence, such that a one-point increase in primary-code distinction increases the likelihood of finding congruence between expressed and inventoried interests by 8%. Using these results, the researcher concluded that the “rule of eight” should be replaced with the “guideline of four”, and that test manuals and interpretative routines for practitioners be updated to reflect this distinction.