Content-Irrelevant Item Attributes: A Framework to Classifying Items Based on Context and Referent

Wahyu Widhiarso ^1,2* Rolf Steyer ² Andrew Perossa ³

• ¹ Gadjah Mada University, Yogyakarta, Indonesia
• ² Friedrich Schiller University Jena, Jena, Thuringia, Germany
• ³ University of Toronto, Toronto, Ontario, Canada

Content-irrelevant items attributes (CIIAs) are characteristics of psychometric scale items that relate to how item stems are worded, rather than the construct they measure. For instance, an item can be framed from a first-hand (e.g., "How would you describe yourself?") or second-hand (e.g., "How would others describe you?") perspective. These attributes meaningfully change the way respondents interpret and answer scale items, so knowing what they are and how they impact data is essential to the construction of valid scales. The present paper serves as a taxonomy of known CIIAs and offers general suggestions on their use. Through this review, we hope to both introduce scale users to the intricacies of item design and offer experienced scale developers a much-needed resource on the types and uses of CIIAs. In doing so, we aim to contribute to the development of more effective, valid scales. We also aim to unify the research literature on item attributes under one taxonomy, to the benefit of scale developers and researchers alike.