Résumé (EN)

The unfolding model of voluntary turnover, as tested by Niederman et al. [1], surprisingly appears to only capture 12% of the job leavers in IT, whereas when applied to other occupational groups, such as accountant or nurses, this ratio is much higher (respectively 92% and 77%). To address this issue, we take a closer look at the classification rules of the original model and we clarify them, highlighting the discrepancy between figure 1 of Lee et al. [2] and their written explanations, [3], [2]. Acknowledging a lack of clarity of those rules, we clarify the meaning of the logical impossibility of the joint existence of an engaged script with a search/evaluation of alternatives. In addition, we discuss the path-switching approach, [3]. Considering this clarification of the classification rules, we recalculate the results of two previous studies ([1] [4]) and we compare and assess those results with the other replications of the model. Those results suggest that a more parsimonious model (i.e. only path 1, 3 and 4b, with engaged script only present in path 1) may be sufficient to capture the majority of those leaving jobs in IT (74% and above).