AI-enabled logistics service recovery: anonymous survey data and reproducibility materials

These data and research materials accompany “Prior Authorization in AI-Enabled Logistics Service Recovery: Evidence from Chinese Consumers”, by Wei Lin. The package contains de-identified closed-response data for the direct-validation series (800 records) and expanded-pretest series (480 records, including 409 eligible cases), together with instruments, dictionaries, aggregate summaries, analysis results and reproduction code. The historical series (2,000 records; 1,998 primary-analysis cases) and main-validation series (1,200 records; 1,082 in the quality sensitivity) are represented by aggregate outputs. Their respondent-level files remain outside the public package while the scope of aggregate-use wording in the available participation materials is unresolved. Code for these analyses requires the separately controlled inputs and cannot be fully rerun from the public files alone. The package also includes the exploratory normative/descriptive perceived-authorization content analyses reported in Supplement A50, Tables S84–S86: reliability, correlated-factor MLR and ordinal WLSMV models, and HC3 regressions with a separate Holm correction. These analyses retain the original six-item primary measure and disclose high facet correlations and a reverse-wording confound; they do not establish two independent psychological constructs. Original identifiers, timestamps, recruitment-source fields, free text and third-party raw datasets are excluded. Third-party download links and source-specific licence information are provided. The four survey series are not pooled or linked, and cross-series respondent independence is not established. Results support sample-bounded measurement and associational interpretations. Author-created data, summaries and documentation are provided under CC BY 4.0; author-created code is provided under the MIT licence. Third-party materials retain their own terms.