Good Question? How Subtle Changes in Question Wording Alters the Interpretation of Measures That Assess the Cognitive Symptoms of Anxiety and Depression.

Lisa Ann Alexander

Advisor: Patrick E McKnight, PhD, Department of Psychology

Committee Members: Todd Kashdan, Jeff Steuwig

David J. King Hall, #2064
November 29, 2017, 01:00 PM to 03:00 PM

Abstract:

Subtle changes in item wording partially account for the correlation between anxiety and depression. The way individuals process information makes self-reported symptom scores particularly susceptible to the item wording characteristics of direction and valence. Using a fully-crossed experimental design, I examined the influence of direction and valence on the correlation between the cognitive symptoms of anxiety and depression. The experimental conditions used minimally altered items from several of the most common measures in research and clinical practice. My analyses demonstrated that items with similar direction and valence consistently increased the correlation between anxiety and depression, whereas items with different direction and valence almost always decreased this correlation. Further, direction and valence accounted for a significant proportion of variance in the correlation between anxiety and depression beyond the effects of the construct. Finally, direction and valence accounted for 16% of the variance in the total scores, whereas the construct (i.e., anxiety and depression) accounted for roughly 0% of this variance. These results likely were not due to confounds created by differential endorsement rates of items within the experimental conditions. When combined, this evidence indicates that item wording is a relevant variable that must be taken into consideration when examining the correlation between anxiety and depression.