Item-Level Survey Data and PLS-SEM Reproducibility Files on Local Champion and Global Player On-Demand Platform Driver-Partners in Indonesia
Description
This dataset contains anonymised survey responses from 400 driver-partners in Indonesia’s on-demand platform sector, comprising 167 Local Champion and 233 Global Player platform partners. The data were collected in the Jabodetabek region between 12 July and 20 October 2025 using a structured self-administered questionnaire in Bahasa Indonesia with a seven-point Likert scale. The dataset includes item-level responses across Market Orientation, Marketing Innovation, Dynamic Marketing Capabilities, and Sustainable Competitive Advantage constructs, together with group identifiers for multigroup comparison. The repository also includes complete PLS-SEM reproducibility files generated in SmartPLS, covering first-stage measurement model assessment, indicator purification, second-stage higher-order construct modelling, bootstrapping, blindfolding, MICOM measurement invariance testing, and multigroup analysis. These files enable replication of the hierarchical component model and support secondary analyses of platform strategy, gig work, market orientation, marketing innovation, dynamic marketing capabilities, sustainable competitive advantage, and cross-archetype comparisons in sharing economy contexts.
Files
Steps to reproduce
1. Open the folder “01. Dataset” and use “01. Data Item Only.xlsx” as the anonymised item-level dataset. The dataset contains 400 respondents, consisting of 167 Local Champion and 233 Global Player driver-partners, with item-level indicators and a grouping variable. 2. Import the dataset into SmartPLS. 3. Specify the first-stage reflective measurement model using the lower-order constructs: Cultural Market Orientation, Behavioral Market Orientation, Product/Service Innovation, Communication Innovation, Pricing, Channel Management, Marketing Research, Marketing Implementation, Valuable Resource, Rare Resource, Inimitable Resource, and Non-substitutability Resource. 4. Run the first-stage PLS Algorithm to assess outer loadings, construct reliability, convergent validity, and AVE for the initial 66 indicators. 5. Examine discriminant validity and cross-loading results. Remove indicators DMC4, DMC8, DMC9, DMC14, and SCA5 due to cross-loading issues, resulting in the purified 61-indicator measurement model. 6. Re-run the first-stage PLS Algorithm and bootstrapping procedure on the purified measurement model to confirm retained indicator validity and significance. 7. Extract the latent variable scores from the validated first-stage model and use “01. LV Score Extracted.csv” as the input for the second-stage higher-order construct model. 8. Specify the second-stage model using four higher-order constructs: Market Orientation, Marketing Innovation, Dynamic Marketing Capabilities, and Sustainable Competitive Advantage. 9. Run the second-stage PLS Algorithm to assess the higher-order construct measurement model and structural model, including path coefficients, R-square, f-square, SRMR, and NFI. 10. Run bootstrapping on the second-stage model to assess direct effects, indirect effects, and hypothesis testing results. 11. Run blindfolding to assess predictive relevance through Q-square values for endogenous constructs. 12. Conduct the MICOM procedure to assess measurement invariance between the Local Champion and Global Player groups. 13. After partial or full measurement invariance is established, conduct multigroup analysis using permutation-based MGA and bootstrapping MGA to compare group-specific path coefficients between Local Champion and Global Player respondents. 14. Compare the reproduced outputs with the Excel and PNG files provided in the repository folders “02. First-Stage Measurement Model”, “03. Second-Stage of Higher-Order Construct Model”, and “04. MICOM and Multi Group Analysis”.
Institutions
- LSPR Institute of Communication and BusinessJakarta, Jakarta