Psychometric properties of the Probability Bias Measure

Abstract

Probability bias is the tendency to think that negative events are relatively likely to happen, and that positive events are relatively unlikely to happen. It is reliably associated with depressive and anxious symptom severity. Here, we describe the initial development and testing of the Probability Bias Measure, which is designed to be relatively brief and general, allowing its convenient use with different populations. In two samples of Turkish-speaking students (Ns = 228 and 170), probability estimates for 12 positive and 12 negative events both displayed adequate one-factor structure, and good internal consistency. More importantly, probability bias - the difference between a participant's mean probability estimate for negative events, and their mean probability estimate for positive events - showed excellent convergent validity, in that it correlated with depressive and anxious symptom severity, positive and negative mood, hopelessness, and dispositional optimism. Furthermore, probability bias accounted for additional variance in depressive symptom severity, over and above that already accounted for by hopelessness and dispositional optimism, which supports its discriminant validity. These findings provide preliminary support for the validity of the Probability Bias Measure for research with Turkish-speaking populations; we also discuss initial evidence for the validity of its English version. The full measure is provided in Turkish and English.