¹Hokkaido University, Psychiatry, Sapporo, Japan

Background: Error monitoring, cognitive flexibility, and rule maintenance are key parts of executive control, and these functions are often impaired in schizophrenia. However, it remains unclear whether these impairments arise from a shared mechanism or reflect separate processes. Post-error slowing (PES) is a common marker of error monitoring. Lose-shift behavior reflects flexibility in responding to feedback, while perseverative errors indicate difficulty in maintaining or updating rules. The Wisconsin Card Sorting Test (WCST) is widely used to assess these functions, but standard summary scores cannot capture the trial-level processes that shape behavior. Although previous studies have examined certain aspects of WCST performance, few have analyzed PES, lose-shift tendencies, and perseverative errors together at the trial level in schizophrenia.

Aims & Objectives: This study aimed to quantify PES, lose-shift behavior, and perseverative errors using trial-level WCST data, and to examine how these three executive components are related.

Method: We analyzed WCST data collected between April 2019 and October 2025 from 41 patients with schizophrenia (17 females, 24 males; mean age 35.6 ± 12.2 years). Reaction times were converted into within-subject z-scores, and PES was defined as the difference between the mean z-scored reaction time after incorrect versus correct trials. Lose-shift behavior was measured as the probability of switching the sorting rule following negative feedback. Perseverative errors were identified using Milner’s criteria. Because lose-shift behavior and perseveration were strongly correlated (r = −0.77), raising concerns about multicollinearity, we did not include both variables simultaneously in the regression model. To maintain model interpretability and stability, we focused on lose-shift rate as the primary predictor of PES. The study was approved by the Hokkaido University Ethics Committee (024-0326).

Results: Patients exhibited robust post-error slowing (PES; Cohen’s d = 1.41). In the regression model, lose-shift rate was a significant negative predictor of PES (β = 1.40, P = 0.002), indicating that patients who were less likely to shift their sorting strategy after negative feedback showed reduced slowing following errors. Age and sex were not significant predictors (both P > 0.11). The overall model was significant (F (3,37) = 3.78, P = 0.019) and explained 23.0% of the variance in PES (R² = 0.23).

Discussion & Conclusions: In patients with schizophrenia, reduced post-error slowing, reflecting impaired error monitoring, was associated with lower cognitive flexibility, as shown by a reduced tendency to shift behavior after negative feedback. One possible interpretation is that weak error monitoring makes it harder to adjust behavior after errors, leading to fewer adaptive strategy shifts. However, because the data are cross-sectional, the direction of this relationship cannot be determined, and both processes may depend on shared executive-control mechanisms. These findings suggest that error monitoring does not operate in isolation but is part of broader executive control processes. Trial-level WCST analysis may help clarify interactions among executive functions and provide sensitive markers of executive dysfunction in schizophrenia.