A psychometric evaluation of the 12-item EPQ-R neuroticism scale in 384,183 UK Biobank participants using item response theory (IRT)

Abstract

Neuroticism has been described as a broad and pervasive personality dimension or heterogeneous trait measuring components of mood instability, such as worry; anxiety; irritability; moodiness; self-consciousness and sadness. Consistent with depression and anxiety-related disorders, increased neuroticism places an individual vulnerable for other unipolar and bipolar mood disorders and therefore highly relevant in epidemiologic research. However, the measurement of neuroticism remains a challenge. We aimed to adapt the 12-item Eysenck Personality Questionnaire-Revised Neuroticism (EPQ-RN) scale for use in epidemiologic studies by identifying psychometrically efficient items using item response theory. The 12-item EPQ-RN scale was evaluated by estimating an IRT model on data from 401,527 UK Biobank participants aged 39 to 73 years (M = 56.41 years; SD = 8.06), 53.68% female. The IRT model yielded two item characteristics: item discrimination, an indicator of how well an item differentiates between respondents, and item difficulty, an indicator of the amount of the latent construct (neuroticism) needed to endorse an item. The EPQ-RN exhibited psychometric inefficiency with poor discrimination at extremes of the scale range. High and low scores are relatively poorly represented and uninformative suggesting that high neuroticism scores derived from the scale are a function of cumulative mid-range values. Following systematic item deletion, a 3-item scale was found to have high levels of discrimination, but offered a narrow range of difficulty i.e. was not sensitive to low levels of neuroticism. A 7-item scale was found to be most informative; providing high levels of discrimination across the range of neuroticism scores.