Agree or agree a little? The rating scale of the BFI-2 causes extreme responses

The Big Five Inventory-2 (BFI-2) is one of the most frequently used personality measures in research, but several scales provide extreme scores. The present research shows that extreme scores are caused by the original rating scale, which uses the labels Disagree a little and Agree a little for the intermediate scale points 2 and 4, respectively. In contrast, other major personality inventories (and some BFI-2 versions) use the labels Disagree and Agree , respectively. In Study 1, respondents assessed the agreement expressed by the response labels in each scale. In Study 2, 150 Norwegian students completed the BFI-2 twice, applying both rating scales on the same occasion. In Study 3, 300 U.S. residents completed the English BFI-2 once, with one of the two scales. The original rating scale in the BFI-2 seems to cause extreme scores, and a graded response model reveals that it does not differentiate in the upper range of some personality traits. Moreover, the two different rating scales, which are both used with the BFI-2 (and the BFI), lead to significantly different scores on several personality scales. As researchers usually do not report the scale used, it can be challenging to compare results across studies. • The BFI-2 is used by researchers around the world to measure the Big Five of personality. • Several personality scales provide extreme scores, and do not differentiate well. • Two response scales are used with the BFI-2, and they provide different test scores. • The original rating scale causes the extreme scores, due to inappropriate wording. • It is often impossible to know which response scale has been used in a particular study.