Answers are grounded in this paper and cite their sources.
This OSF project archives the analysis plan, statistical code, and supplementary outputs for a randomized factorial survey experiment fielded within the 2024 wave of the Chinese National Sports and Society Survey (CNSSS). Design: 5,656 adults each evaluated one hypothetical exercise scenario combining six randomized supply-side attributes (facility distance 5/15/30 minutes; fee level free/low/high; presence of fitness organization, social sports activities, skill instruction, and chronic-disease exercise prescription). Attribute assignment was fully orthogonal (15 pairwise tests) with covariate balance across arms. Primary analysis: OLS with heteroskedasticity-robust standard errors, unweighted. Robustness: ordered logit, average marginal effects on a binary threshold, randomization (permutation) inference, Holm family-wise correction, Bayesian joint-posterior simulation, sampling-weighted replication, and a fatigue-restricted subsample. Heterogeneity: 48 prespecified interactions with false-discovery-rate correction, TOST-style equivalence bounds, cross-fitted DR-learner causal forests, and horseshoe shrinkage regressions. Contents: (1) frozen analysis specification (pre-dated before outcome analysis), (2) replication code (Stata and Python), (3) data access instructions. The CNSSS public-use data are not redistributed here; they are freely available to registered researchers from Wuhan Sports University.
No takes yet. Share an insight, caveat, or question.
baoli yin (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: