An expert-informed framework of school conditions potentially supporting adolescent physical activity habits: a three-stage consultation in China

BackgroundSchools can provide conditions that may support repeated physical activity, but expert endorsement of those conditions does not demonstrate habit formation. This study developed an expert-informed content framework for Chinese adolescents aged 12–18 years.MethodsA three-stage, Delphi-informed consultation comprised qualitative elicitation and item development, one full quantitative rating stage, and targeted review of proposed final decisions. Of 25 experts approached, 23, 18, and 16 were reported to complete Stages 1, 2, and 3, respectively. The reported analyses used 16 complete cases. Seven earlier-stage contributors withdrew consent; their data are excluded and recorded as no longer retained, preventing full-stage reanalysis. Five theory-informed domains guided the interviews; a predominantly deductive qualitative analysis using a codebook specified before the interviews supported indicator refinement rather than independently establishing their structure. Forty-five literature-derived provisional indicators and five interview-derived additions were consolidated into 34 candidates. Importance, feasibility, and clarity were rated on five-point scales. The importance-based retention rule specified before rating combined a mean ≥4.00, coefficient of variation ≤0.20, and ≥80% ratings of 4 or 5.ResultsThirty-three candidates met the retention rule; SPC9 did not. The retained indicators received uniformly high importance endorsement, but concordance in relative rankings was limited (Kendall’s W: importance 0.115, feasibility 0.160, clarity 0.163). Individual-level Stage 3 confirmation distributions were not available in the submitted analytic record, and quantitative convergence across stages was not assessed. The exploratory, sample-relative importance–feasibility matrix assigned 10, 5, 10, and 8 indicators to its four quadrants; 11 of 33 indicators changed quadrant in at least one leave-one-expert-out analysis.ConclusionThe resulting curricular, extracurricular, peer, cultural, and institutional indicators specify school conditions that experts considered potentially relevant to habit support. The findings do not establish strong inter-expert consensus, a validated measurement instrument, actual habit formation, or causal effects. Complete-case selection, limited stakeholder representation, and incomplete process records restrict interpretation. More complete source documentation, stakeholder co-design, and prospective measurement and implementation studies are needed.