Publication details
Researching response scale format effects in questionnaires: Using the Height Inventory
| Authors | |
|---|---|
| Year of publication | 2024 |
| Type | Appeared in Conference without Proceedings |
| MU Faculty or unit | |
| Citation | |
| Attached files | |
| Description | Self-report measures of attitudes and personality characteristics mainly use items with Likert-type response scales (LS) where respondents select an answer from a range of ordered options (e.g., agree–disagree). Such response scales can differ in several formal attributes – number of options, presence of verbal anchors, their extremity, or orientation (so-called reversed items). These may affect the reliability and validity of responses and total scores (Furr, 2011). To assess response-format effects, our team extensively exploits the unique properties of our Height Inventory (Rečka, 2018). The idea of measuring height using a psychological questionnaire is not entirely original (van der Linden, 2016; Kam et al., 2021). However, we elaborated a new methodological approach utilizing self-reported height to assess the criterion validity of observed scores and latent traits using latent variable models and measurement invariance analysis. So far, we used this approach to study the effects of LS length, the presence and extremity of its verbal anchors, and reversed-key items. We also assessed the performance of the Visual Analogue Scale compared to LS and the effects of speedy vs. careful responding. Some of our findings contradict common beliefs; for example, that binary items might have higher criterion validity than longer response scales. Our talk briefly outlines the Height Inventory and its psychometric properties. Then, we describe our methodological approach in detail, focusing on its advantages and limits. |
| Related projects: |