Adaption and validation of the algorithm aversion scale and its relationship with neuroticism and trust

Purpose Previous studies highlighted a dearth of scale to measure algorithm aversion among working professionals. Accordingly, this study tried to adopt and validate an algorithm aversion scale for employees. Design/methodology/approach The scale was developed and validated in three stages. In the first stage, a five-item xenophobic scale was adopted. Its items were rephrased to match the algorithm aversion-related workplace context. In the second stage, the sample was divided into two subsamples using random sampling. The exploratory factor analysis (EFA) applied on the first subsample. The confirmatory factor analysis (CFA) was conducted on the second sample to check the model’s fitness proposed by EFA. In the third stage, a sample of 378 respondents was collected. The Cronbach’s alpha (CA) and composite reliability (CR) values checked internal consistency reliability of the proposed algorithm aversion scale. Findings The EFA yielded a one-factor solution with acceptable factor loading (greater than 0.50) of all (five) statements. All values of model indices in CFA were in the acceptable range, which recommended the factorial structure of the scale. The acceptable (greater than 0.70) values of CA and CR concluded the internal consistency reliability. The convergent validity was suggested by appropriate average variance extracted (AVE) values (greater than 0.50). In addition, as expected, the proposed algorithm aversion scale reported a statistically significant positive correlation with neuroticism and a considerable negative relationship with trust. These findings concluded the criterion validity of the scale. Originality/value This study uses primary data to adopt and validate algorithm aversion scale.