Dataset for pretest posttest serious game evaluation through Structural Equation Model

This dataset contains a curated pre–post dataset from a workshop with informal small vendors (bodegueros) in San Juan de Lurigancho, Lima (Peru), conducted in November 2023 using the serious game Bodegus as an ethics-oriented intervention. The dataset captures two constructs—Perceived Usefulness of Formal Business Practices (TAM-inspired) and Intention to Formalize (TPB-inspired)—modeled as second-order formative composites across three domains: governance/leadership, legal/tax, and accounting/finance. Materials include: (i) item-level pre/post responses and descriptive summaries; (ii) an Excel workbook mirroring the analysis workflow; (iii) SmartPLS 4 artifacts for PLS-SEM (formative) with bootstrapping, MICOM invariance inputs, multigroup tests (permutation, Henseler’s MGA), and PLSpredict outputs; (iv) the English questionnaire and a codebook detailing variables, levels, and coding rules; and (v) a brief sensitivity note (G*Power) regarding small-sample constraints. The sample (n = 38) consists of hard-to-reach informal entrepreneurs; the HR domain is retained in the instrument but not analyzed due to limited valid cases. Files are organized for replication, classroom use, and aggregation with future cohorts, and can be adapted to other serious-game or active-learning interventions.