Data Comparative Analysis of Player Satisfaction and Continuance Willingness Between Players of Mobile of Multiplayer Online Battle Arena

Description

This dataset supports a quantitative study examining the effects of Community Interaction (CIN), Gaming Experience (GEX), Social Value (SOV), and Entertainment Value (ENT) on Player Satisfaction (SAT) and Continuance Willingness (CWI) among mobile Multiplayer Online Battle Arena (MOBA) game players in Indonesia. The conceptual model was tested using Partial Least Squares Structural Equation Modeling (PLS-SEM) and extended with Multi-Group Analysis (MGA) to compare two distinct player segments: Mobile Legends: Bang Bang (ML) and League of Legends: Wild Rift (LoLWR). The dataset consists of (1) raw survey responses from 60 respondents collected via online questionnaire, and (2) six SmartPLS 4 output files containing full structural model results, bootstrapping statistics, and MGA comparisons. The data is intended to support replication, secondary analysis, and methodological review of the PLS-SEM and MGA procedures reported in the associated publication.

Steps to reproduce

All analyses were performed using SmartPLS 4 (https://www.smartpls.com). The free trial supports up to 100 cases and is sufficient for this dataset (N = 60). Step 1 — Import Data. Open SmartPLS 4 and create a new project. Import moba_smartpls_n60.csv. Set all 32 indicator columns (CIN1–CWI4) as Metric type and the GAME column as Categorical type. Confirm 60 cases are loaded. Step 2 — Build the Model. In the Model Editor, create six latent constructs: CIN, GEX, SOV, ENT, SAT, and CWI. Assign each indicator to its construct using reflective (Mode A) connections. Draw nine inner model paths: CIN, GEX, SOV, and ENT each pointing to both SAT and CWI, and SAT pointing to CWI. Step 3 — Run PLS Algorithm. Click Calculate → PLS Algorithm. Use factor weighting, 300 maximum iterations, and stop criterion 1×10⁻⁷. Export the Complete output to Excel. This replicates update_Model_PLS-SEM_Permutation.xlsx. Verify measurement quality: all outer loadings > 0.70, AVE > 0.50, HTMT < 0.85, SRMR < 0.08. Step 4 — Run Bootstrapping. Click Calculate → Bootstrapping. Set subsamples to 5,000, confidence interval to BCa, and significance level to 0.05. Export using the Faster option to replicate faster_Model_PLS-SEM_Permutation.xlsx, or the Complete option to replicate slower_Model_PLS-SEM_Permutation.xlsx. Read path significance from the Path coefficients → Mean, STDEV, T values, p values table. Paths are significant at p < 0.05. Step 5 — Run Multi-Group Analysis. Set the grouping variable to GAME to split respondents into Group_ML (n = 30) and Group_LoLWR (n = 30). For Permutation MGA, click Calculate → Permutation, set 5,000 permutations, and export to replicate Permutation_MGA_Model_PLS-SEM_Permutation.xlsx. For Bootstrap MGA, rerun Bootstrapping with the grouping variable active and export to replicate Boostrap_MGA_Model_PLS-SEM_Permutation.xlsx. In both outputs, read the p-value column under Path coefficients to identify significant between-group differences. Note. Bootstrapping and permutation use random resampling, so numerical outputs may vary slightly across runs. Substantive conclusions remain stable at 5,000 resamples. Do not modify moba_smartpls_n60.csv before importing. Reference: Ringle, C. M., Wende, S., & Becker, J.-M. (2024). SmartPLS 4. SmartPLS GmbH. https://www.smartpls.com

Institutions

• Binus University

  Jakarta, Jakarta

Categories

Licence