Jennifer Jablonski

Construct-worded vs. Mixed-item Formats: a Psychometric Comparison and Theoretical Model Jennifer Jablonski

As social scientists, we are often in the position of administering measures for clinical or research purposes. Research has shown that psychometric advantages of measures using a construct-worded as opposed to a mixed-item format include improved reliability and validity, the elimination of factorial artifacts, more consistent item means, and lower item mean standard deviations. This book further explores these findings by comparing the widely-used mixed-worded State-Trait Anxiety Inventory (STAI) and a new construct-worded revised version (STAI-R). Although the psychometric advantages of construct-worded measures have been well-noted, no previous efforts have been made to explain why they occur. By exploring the cognitive processes involved in responding to construct-worded and opposite-worded items, this book provides the first systematic attempt to describe why the psychometric properties of these item types differ. The eye-opening findings presented in this book will help social scientists gain new insights into the rating scales they select and the results they obtain.