Replication package: PLS-SEM, fsQCA, and NCA analysis of consumer responses to the international advertising of a global logistics brand in Vietnam

Version 3 of the replication package for a survey study that combines PLS-SEM, fuzzy-set qualitative comparative analysis (fsQCA), and necessary condition analysis (NCA), with a sample-heterogeneity screen and a Monte Carlo evaluation, applied to consumer responses to the advertising of a global logistics brand (DHL) in Vietnam. The package contains: (1) the anonymized analytic dataset (n = 380 respondents × 34 variables, CSV), including a data-collection wave indicator; (2) a variable-by-variable codebook with the English back-translation of every item; (3) eleven R scripts that reproduce all results in the revised article and its Supplementary Material: the heterogeneity screen (MICOM and multigroup analysis), the main PLS-SEM analysis of wave 2 (measurement and structural model, mediation, moderation, conditional indirect effects, PLSpredict, CVPAT, IPMA), robustness checks, fsQCA with theory-based and percentile calibration, typical and deviant cases of the fsQCA solutions, NCA against a data-calibrated additive null distribution, the Monte Carlo study, and supplementary checks. See README.md for the file list, requirements, run instructions, and the key results to check. Changes from version 2: a base-R script (QQ_R1_fsqca_contradictions.R) was added that reproduces the counts of typical and deviant cases for the fsQCA solutions (Tables S2b, S2c, and S2e of the Supplementary Material), and the README now states that compositional invariance (MICOM step 2) is rejected, so the two survey waves are neither pooled nor formally compared. The data and all other scripts are unchanged. Changes from version 1: two respondents in the under-18 age band were removed (382 → 380); a wave variable derived from the submission timestamps was added (timestamps are not released); construct labels were updated; new scripts were added. The survey was administered in Vietnamese. The data are fully anonymized (no names, contact details, IP addresses, or timestamps). Data and documentation are released under CC BY 4.0; code under the MIT license.