Influence of Learners’ Welfare TQM practices on learners’ academic achievement in Integrated Science in junior schools in Bomet County, Kenya

Total Quality Management (TQM) offers strong potential to transform science education through strategic stakeholder engagement, systematic process management, and continuous improvement practices. However, academic achievement in Integrated Science among Junior School learners in Bomet County, Kenya, remains below national benchmarks, threatening the Vision 2030 target of transitioning 60% of learners into Senior School STEM pathways. Baseline data from the 2023 Kenya Primary School Education Assessment (KPSEA) revealed that 59.92% of learners scored at the “Approaching Expectations” or “Below Expectations” levels. This trend persisted across subsequent School-Based Assessments (SBAs), where 67.3% of Grade Seven learners and 62.3% of Grade Eight learners failed to attain mastery. Furthermore, results from the 2025 Kenya Junior School Education Assessment (KJSEA) recorded a county mean score of 31.63%, reflecting only 30–40% core competency mastery. Guided by Deming’s Total Quality Management Theory and Hattie’s Visible Learning Theory, this study investigated the influence of Learner Welfare (LW) TQM practices on academic achievement in Integrated Science in Junior Schools within Bomet County. A pragmatist philosophy was adopted alongside a mixed-methods concurrent triangulation design. The target population comprised 21,653 participants across Bomet County. Using Yamane's formula, a quantitative sample of 230 Heads of Institutions (HOIs) was selected through proportionate stratified random sampling. The qualitative phase involved five Sub-County Directors of Education, 23 Grade Nine parents' representatives, and 36 classroom observations involving 36 Integrated Science Facilitators and 2,756 Grade Nine learners (overall N = 3,050 active participants). Data were collected using questionnaires, interview guides, observation checklists, and document analysis guides. Reliability was confirmed through Cronbach’s alpha (alpha = .853) for the HOI instrument. Quantitative data were analyzed using descriptive statistics and regression analysis (alpha = .05) in SPSS version 27, while qualitative data were analyzed thematically in NVivo version 14. Simple linear regression revealed that Learner Welfare emerged as a statistically significant predictor (R2 = .061, beta = .246, p = .001), Qualitative findings and document audits revealed that LW is currently constrained by an operational disconnect, being treated as an informal administrative afterthought with severely underutilized feedback and inadequate career guidance . Furthermore, severe classroom overcrowding exceeding 60–80 students per class severely disrupted personalized psychosocial support and safe practical instruction. The study concludes that learner welfare management provides an essential affective foundation for science mastery. To realize national STEM targets, institutional leadership and policy must transition learner welfare into a systematic support framework by enforcing routine learner feedback mechanisms, integrating structured STEM career counseling, and managing classroom congestion to support safe, learner-centered practical instruction.