Publication details
Efekt polarity: Jsme na cestě k formální teorii nesouhlasných položek?
| Title in English | The Polarity Effect: Toward a Formal Theory of Reverse-Keyed Items |
|---|---|
| Authors | |
| Year of publication | 2026 |
| Type | Appeared in Conference without Proceedings |
| MU Faculty or unit | |
| Citation | |
| Attached files | |
| Description | Reverse-keyed items are an integral part of psychological questionnaires; already Likert (1932) recommended that at least half of the items in a scale should be reverse-keyed. The rationale was to minimize response biases, particularly acquiescence, and this argument is still commonly invoked today. However, including reverse-keyed items often results in suboptimal psychometric properties: reliability decreases, while the apparent dimensionality of the questionnaire increases. Separate factors are often observed for positively and reverse-keyed items, as if the two types measured different constructs. Some sources therefore recommend using positively keyed items exclusively. This phenomenon is most commonly referred to as the polarity effect. Although it has been known for more than 100 years, its causes remain poorly understood. Moreover, some commonly cited explanations in the literature, such as social desirability (Weijters et al., 2013), are difficult to reconcile with the phenomenon without introducing complex and highly restrictive assumptions. Existing research on the polarity effect also largely ignores linguistic aspects, particularly the existence of several fundamentally different types of reverse-keyed items (Elek et al., 2025; Kam et al., 2021). The most important limitation, however, is the lack of a satisfactory definition of the phenomenon itself, without which its specific causes are difficult to investigate systematically. Our aim is to develop a formalized theory of the polarity effect using the framework of productive explanations (van Dongen et al., 2025). In this presentation, we will introduce an existing, though as yet unpublished, formal definition of the phenomenon. We will also present a preliminary categorization of reverse-keyed items developed through linguistic analyses. Finally, we will report initial findings from a systematic review mapping existing explanations of the phenomenon and critically evaluate them in terms of their plausibility and empirical support. Although our research is primarily theoretical, it has the potential to address several key practical questions: Should reverse-keyed items be included in questionnaires or not? What are the risks associated with including or excluding them? And how do different types of reverse-keyed items function? |
| Related projects: |