Reasons to Randomize: Sort-Specific Effects on Construct Content in Cross-Sectional Repertory Grid Studies

Abstract

Several articles have discussed the effect of variations in the repertory grid procedure on measures derived from the data it generates. Largely unexplored, in contrast, is the effect of a more subtle variation inherent in the grid procedure itself: the specific combination of elements used in a dyad or triad during elicitation. This study investigates the differential effects of a specific element combination on the elicited constructs by means of data-driven content analysis. Three independent repertory grid samples were collected, one using nutrition styles (N = 70) and two using occupational profiles as elements (both N = 60). For the latter two studies, the sorting technique (dyad vs. triad) was varied. A pronounced specific effect of the sorts on elicited construct content was found across all three studies.

ACKNOWLEDGMENTS

Parts of this study were presented at the ICPCP in Hertfordshire in 2015 and at the EPCA conference in Padua in 2016.

SUPPLEMENTAL MATERIAL

Supplemental data for this article can be accessed on the publisher's website at: http://dx.doi.org/10.1080/10720537.2017.1299654