Acquiescent response style refers to the tendency of an individual to systematically agree (yea saying) or disagree (nay saying) with questionnaire items, regardless of their content. For example, consider the two following items, with answering scales from 1 (strongly agree) to 5 (strongly disagree):

- “Are you relaxed during stressful situations?”

- “Do you get nervous easily?”

They both aim at measuring the same PT (Emotional Stability), but only the second one is reverse-coded in the sense that a higher degree of agreement in the response is associated with a lower Emotional Stability. Hence, a respondent that strongly agrees with both statements shows a form of contradiction, indicative of a positive acquiescence bias. This response pattern ended up being a strong driver of the variation in the data. McCrae et al. (48) found that, when not corrected for, acquiescence bias came out as the first factor, highlighting the importance to correct it, which is now standard in psychometric analysis (33, 58).

The calculation of the acquiescence bias (AB) requires questions that aim at measuring the same PT but of which at least one is reversed and at least one is not reversed. The AB is calculated at the individual level. In the case of the 15 STEP items, because Agreeableness and Openness do not include any reverse question, AB can only be calculated using items of Conscientiousness, Extraversion, and Emotional Stability, but the correction is then applied to all items.

We calculate the acquiescence bias and apply the correction using the following steps:

1) Reverse the reverse-coded items. For example, if the possible answers range from 1 to 5, then answer 1 (fully disagreeing with a reverse-coded statement) is assigned a value of 5, answer 2 is assigned a value of 4, and so on.

2) For each PT that has at least one reverse-coded item and one nonreverse-coded item, calculate the average answer of reverse-coded items and the average answer of nonreverse-coded items.

3) For each PT, take the difference between the average of nonreverse-coded items and the average of reverse-coded items and divide it by 2.

4) The AB is the average of the differences obtained in (3) across all PTs

5) To correct for AB, add the AB obtained in (4) to every reverse-coded item and subtract the AB from every nonreverse-coded item.

The intuition of the AB correction is that in the absence of contradiction, in a scale from 1 to 5, the average raw answer between reversed items and nonreversed items should be 3 (the more one agrees with a statement, the more one should disagree with the opposite statement). Any average deviation from 3 is attributed to the AB and corrected by making the adjustment that will bring this average back to 3 after the correction.

In table S5, we show Cronbach’s alpha, calculated by dataset and PT, using the items that were corrected for AB and without the correction for AB. We find that Cronbach’s alphas are substantially higher after the AB correction, which indicates that the correction increased internal consistency. The within correlations, as well as the difference between the between and within correlations, are also lower without the correction. Furthermore, while the calculation of Cronbach’s alpha uses the absolute correlation between the items, the dagger (†) in table S5 indicates that at least one of the correlations between two items belonging to the same PT is negative. Without acquiescence bias correction, we found a large number of cases of these negative correlations. This is driven by negative correlations between reverse and nonreverse items, suggesting that the AB response pattern is more influential than the PT that the items aim to measure.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.