cognitive resilience

Final-year engineering students frequently face severe psychological stress driven by demanding capstone projects and rigorous accreditation standards. This research proposes a novel decision-support framework to objectively evaluate, categorize, and systematically manage academic stress among engineering undergraduates. By integrating automated non-parametric machine learning with a multi-criteria network decision model, the research bridges diagnostic psychological profiling and strategic institutional intervention. Data were gathered from six hundred eleven final-year students across four engineering departments using a validated psychometric questionnaire. The automated classification engine utilizes a nearest-neighbor algorithm based on spatial Euclidean proximity to categorize individual distress severity into depression, anxiety, and stress levels without relying on rigid parametric assumptions. Subsequently, a strategic situational assessment and a multi-perspective personal performance framework were established to translate abstract psychological findings into measurable indicators. Clinical experts and academic domain specialists evaluated these indicators using an analytical network process, incorporating complex interdependencies and feedback loops to compute steady-state priority weights. Empirical results revealed that over ninety percent of respondents experienced at least moderate psychological distress. Crucially, the mathematical prioritization demonstrated that the financial perspective holds the paramount priority weight, followed by internal cognitive resilience, continuous knowledge acquisition, and external social relationships. This proves that financial stability serves as a foundational prerequisite for academic recovery and intellectual execution. The primary contribution of this research lies in converting static psychometric evaluations into a dynamic, clinically validated decision-support tool. Higher education institutions and academic counselors can implement this prioritized framework to allocate institutional resources effectively, design targeted financial relief initiatives, and optimize student well-being alongside academic excellence.

Whether breadth of leisure activity engagement (LAE) is associated with cognitive resilience (CR) in older Chinese adults, and whether this association differs by gender, remain unclear. Using five waves of the China Health and Retirement Longitudinal Study (2011–2020), we estimated CR as the residual of global cognition after accounting for age, education, gender, residence, and chronic disease burden, and examined activity × gender interactions in generalized estimating equations ( n = 7130 observations from 3293 adults aged ≥60). Cognitive and social activity breadth were each independently associated with higher CR, but with contrasting domain patterns. Cognitive activities showed a linear dose-response with comparable associations in both genders (males β = 0.164, females β = 0.171, both p < 0.001; Activity × Gender interaction p = 0.435). Social activities followed a quadratic dose-response peaking at approximately three activity types. In gender-stratified models, the social-activity association was concentrated in males ( β = 0.164, p < 0.001) and was not significant among females ( β = 0.032, p = 0.323; activity × gender interaction β = −0.111, p = 0.020), despite higher social participation rates among women. Lagged analyses supported temporal precedence. These findings indicate that the resilience associations of LAE breadth in older Chinese adults are domain- and gender-dependent: cognitive activity breadth was positively associated with CR in both genders, whereas this association was concentrated in men for social activities. These findings warrant further investigation of whether the content, context, or measurement of social engagement contributes to the observed gender difference.