Abstract
Issues of normativity (responding in a typical or average fashion) and desirability (the tendency for raters to endorse positive characteristics rather than neutral or more negative ones) are common in areas of the social sciences that frequently utilize profile correlations to measure dyadic similarity. They have implications for family scholars as well. In the present study, a pre-existing data set was used to make an initial, though limited, investigation into potential confounds of normativity and desirability for macrolevel observational assessments of family interaction. An empirical example is presented using q-sort ratings of family interaction, with variance in observational assessments decomposed into component parts. High levels of both normativity and desirability were found, indicating possible problems in terms of both reliability and validity of assessment. While the results provide an interesting beginning, they are limited due to the use of a q-sort methodology as well as an instrument with limited background and use. These limitations are discussed, as well as alternative interpretations for normativity and desirability and implications for future research.