Dataset Information

Identifying the top predictors of student well-being across cultures using machine learning and conventional statistics.

ABSTRACT: Alongside academic learning, there is increasing recognition that educational systems must also cater to students' well-being. This study examines the key factors that predict adolescent students' subjective well-being, indexed by life satisfaction, positive affect, and negative affect. Data from 522,836 secondary school students from 71 countries/regions across eight different cultural contexts were analyzed. Underpinned by Bronfenbrenner's bioecological theory, both machine learning (i.e., light gradient-boosting machine) and conventional statistics (i.e., hierarchical linear modeling) were used to examine the roles of person, process, and context factors. Among the multiple predictors examined, school belonging and sense of meaning emerged as the common predictors of the various well-being dimensions. Different well-being dimensions also had distinct predictors. Life satisfaction was best predicted by a sense of meaning, school belonging, parental support, fear of failure, and GDP per capita. Positive affect was most strongly predicted by resilience, sense of meaning, school belonging, parental support, and GDP per capita. Negative affect was most strongly predicted by fear of failure, gender, being bullied, school belonging, and sense of meaning. There was a remarkable level of cross-cultural similarity in terms of the top predictors of well-being across the globe. Theoretical and practical implications are discussed.

SUBMITTER: King RB

PROVIDER: S-EPMC11006657 | biostudies-literature | 2024 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Identifying the top predictors of student well-being across cultures using machine learning and conventional statistics.

King Ronnel B RB Wang Yi Y Fu Lingyi L Leung Shing On SO

Scientific reports 20240410 1

Alongside academic learning, there is increasing recognition that educational systems must also cater to students' well-being. This study examines the key factors that predict adolescent students' subjective well-being, indexed by life satisfaction, positive affect, and negative affect. Data from 522,836 secondary school students from 71 countries/regions across eight different cultural contexts were analyzed. Underpinned by Bronfenbrenner's bioecological theory, both machine learning (i.e., lig ...[more]

PMID: 38600124

Similar Datasets

Project description:University represents a time of both great uncertainty and change as well as a time of opportunity and learning. University students represent a population both at a higher risk of experiencing poor mental health and diagnosis and a population with relatively greater access and communication of both mental health literacy resources and mental health support services. Despite this, we consistently see low intention of help-seeking for mental health services or health services, with a clear preference shown for personal contacts such as friends or parents. To understand help-seeking intentions and their relationship with well-being, the current study explored two core hypotheses, when assessing a broad range of help-seeking options, the likelihood of seeking support from a range of sources will cluster together to create help-seeking groupings that can be further explored (1) and that across these help-seeking factors, there would be difference in intention score across students who fall within "at-risk," "low," or "normal to high" well-being groupings (2). Through a series of exploratory factor analysis (EFA) on a subsample (N = 178) and final confirmatory factor analysis (N = 1597) identified five help-seeking factors: Intimate Partner (single item), Personal Relations (friends, parents, and other relatives), External Health Service (external mental health provider and health provider), University Health Service (university mental health and health provider), and Digital and Distal Professional (digital apps, websites or forums, telehealth, religious leaders, and phone or online emergency services). To address hypothesis 2, a multivariate analysis of covariance was run to assess help-seeking intentions across factors between students with "at-risk" (N = 453), "low" (N = 484), or "normal to high" (N = 563) well-being scores. Although significant differences were found between groups on almost all help-seeking factors (except External Health Service), the differences between groups were small. However, consistently those with "normal to high" well-being demonstrated higher intention to seek help compared to "low" and "at-risk" groups. Across all groups, Personal Relations demonstrated the highest average help-seeking intention score and, in addition to supporting findings in previous literature, represents a potential "lowest hanging fruit" of help-seeking source for university students.

Project description:IntroductionUK higher education (HE) student numbers are increasing and students report higher levels of mental health and well-being issues. Social prescribing links individuals to community-based, non-medical support. It is widely implemented throughout the UK, and is supported by the Welsh Government. This protocol presents an evaluation of a new social prescribing service to enhance student well-being, a first for UK HE students.Methods and analysisA realist evaluation to articulate why, how and to what extent and circumstances social prescribing works for students, using a mixed-methods sequential design of four cycles. Cycle 1 informs the model and programme theory development of how the model works; activities include a Realist Review, Group Concept Mapping and producing bilingual short films about the evaluation and model. Cycle 2 involves secondary analysis of routine service data, and outcome measurements from students receiving a social prescription. Cycle 3 uses reflective diaries and qualitative realist interviews with stakeholders to understand the process and outcome of the model. Cycle 4 concludes with a world café workshop with stakeholders to agree and finalise the framework specification of 'how, why, when and to what extent' the model works. A meta-matrix construction will determine convergence, complementarity or discrepancy across the cycles. An advisory group of key stakeholders informs each cycle.Ethics and disseminationUniversity of South Wales Life Sciences and Education Ethics Committee and Wrexham Glyndwr University (WGU) Research Ethics Sub-Committee approved secondary data analysis of participant demographics (200 805LRL:USW, id441:WGU), outcome measurement tools (200 902LR:USW, id441:WGU) and qualitative data collection (200 804LR:USW, id449:WGU). The authors will publish findings in peer-reviewed journals, produce an evaluation report to the funder and a short film for dissemination via stakeholders, university networks, United Nations Regional Centre of Expertise in Wales, PRIME Centre Wales, Wales School for Social Prescribing Research, conferences and social media.

Dataset Information

Identifying the top predictors of student well-being across cultures using machine learning and conventional statistics.

Publications

Identifying the top predictors of student well-being across cultures using machine learning and conventional statistics.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets