Abstract Student dropout is a multi-causal process. Different theoretical models on student dropout consider dysfunctional study behavior (e.g., academic procrastination) and low study satisfaction as possible determinants of students’ dropout intentions during their university studies. However, these models neglect contemporary conceptualizations that assume reverse relationships between dropout intentions and other determinants of the dropout process. Until now, empirical evidence on these assumptions is scant. The present three-wave longitudinal study explored the reciprocal relationships between academic procrastination, study satisfaction, and dropout intentions over one semester. To this end, we used data of N = 326 undergraduate students enrolled in mathematics and law. Our latent cross-lagged panel model replicated existing empirical cross-sectional findings between the variables (i.e., academic procrastination, study satisfaction, and dropout intentions). Regarding the longitudinal relations, as expected, the cross-lagged effects showed that higher dropout intentions significantly related to subsequent higher academic procrastination and lower study satisfaction. Unexpectedly, academic procrastination did not significantly relate to subsequent dropout intentions. Additionally, higher study satisfaction significantly associated with subsequent higher dropout intentions—possibly due to unfulfilled expectations. Further, higher study satisfaction significantly related to subsequent higher procrastination—possibly due to more confidence among satisfied students. Our results broaden the view on dropout intentions as part of the dynamic interplay of student dropout determinants and the need to refine dropout models’ assumptions accordingly. Practically, realistic expectations seem important to reduce dropout intentions. Further, student counselors should have a closer look at the reasons for academic procrastination to develop individual solutions for this dysfunctional behavior.
Stored only when both papers are in the corpus — 91 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.