Examining the Wording Effect: What Are We Measuring?

ABSTRACT

In this study, the extent to wording effects impact structure and factor loadings, internal consistency and measurement invariance was outlined. The modified form, which includes items that semantically reversed, explains %21.5 more variance than the original form. Also, reversed items’ factor loadings are higher. As a result of CFA, indexes obtained from modified form are better. Measurement invariance is not ensured at any step, which is a significant finding. These findings point to two forms measure different structures. Despite the quite high internal consistency coefficients obtained from both forms, the reliabilities in the modified form are statistically significantly higher than the original form. Although there are many opinions in the literature on the use of reversed items, this study shed light on the practical implications and psychometric findings.

PUBLIC SIGNIFICANCE STATEMENT

This study suggests that to use semantically reversed items may increase explained variance near to %21.5 and internal consistency but using the reversed items considerate that it may measure a different construct. Additionally, it is clearly seen that when reversing the items, caution should be taken to ensure that they are similar to the other items in the scale.

KEYWORDS:

• Wording effect
• measurement invariance
• reversed items

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15366367.2024.2329505

Disclosure statement

No potential conflict of interest was reported by the author(s).