Supply Chain Risk Management and Order-Based Costing: A Synthetic Industrial Dataset for Structural Equation Modeling (PLS-SEM)
Description
This dataset provides 450 production order observations from a medical device manufacturing environment (Make-to-Order). It is specifically designed to analyze the impact of Supply Chain Risk (SCR) on cost variability and operational performance. The data includes indicators for risk exposure, SCRM practices, information traceability, and granular cost metrics. It features four distinct order profiles identified via K-means clustering, making it suitable for Multi-Group Analysis (MGA) and robust PLS-SEM path modeling. This dataset supports research in operations management, industrial costing, and supply chain resilience. ======================================================================== STUDY: SCRM Integration and Order-Based Costing in Medical Devices ======================================================================== GENERAL DESCRIPTION: This repository contains 450 production order records (unit of analysis) from a textile medical device industrial plant. It provides empirical- synthetic evidence to study unit cost variability under supply chain risk conditions. REPOSITORY CONTENT: 1. SCRM_Orders_Dataset_450.csv: Primary data (450 obs x 42 vars). 2. Data_Dictionary.csv: Technical definitions, units, and roles. 3. Metadata_Technical_Specs.csv: Simulation parameters and reproducibility. REPRODUCIBILITY METADATA: - Random Seed: 20260424 - Number of Clusters: 4 (K-means algorithm) - Software used for validation: SmartPLS 4 and R (seminr package). - Cluster Distribution: C1 (91), C2 (157), C3 (96), C4 (106). SUGGESTED USE: Optimized for Partial Least Squares Structural Equation Modeling (PLS-SEM). The 'cluster' variable is intended for Multi-Group Analysis (MGA) to explore structural heterogeneity. CITATION: Please cite this dataset as follows: De León Montes, L. (2026). "SCRM and Order-Based Costing Dataset". Mendeley Data, V1. doi: 10.17632/[YOUR_DOI_HERE] ========================================================================
Files
Steps to reproduce
Data Preparation: Download 'SCRM_Orders_Dataset_450.csv' and 'Data_Dictionary.csv'. Ensure the dataset is loaded with UTF-8 encoding to maintain data integrity. Index Construction (PCA): Use the observed indicators for Risk (A1-A6) and Traceability (C1-C3) to replicate the 'RiskIndex' and 'TraceIndex'. Perform a Principal Component Analysis (PCA) using a correlation matrix, retaining the first principal component (PC1). Cluster Analysis: To replicate the order profiles, run a K-means clustering algorithm using 'RiskIndex', 'cost_var_metric', and 'noq_metric' as input variables. Set the number of clusters to k=4 and use the random seed 20260424 to ensure identical group assignments. Structural Model (PLS-SEM): Import the dataset into SmartPLS 4 or R (seminr/plspm). Configure the structural model as defined in the README. Run the PLS-SEM algorithm (Path weighting scheme) and perform Bootstrapping with 5,000 subsamples to validate hypotheses H1-H6. Robustness Check: Run an OLS regression using the same variables to confirm the consistency of the signs and significance levels of the coefficients.