From Structural Equation Modelling to Double Machine Learning: Robustness Analysis for Survey-Based Research

Jul 16, 2026

Upload by Himbo

Structural equation modelling (SEM) is widely used in survey-based business and information systems research to evaluate latent constructs and theory-driven structural relationships. Yet SEM path significance depends on a single model specification. The study develops a staged robustness framework that links SEM, ordinary least squares (OLS) regression, and Double Machine Learning (DML). SEM refines measurement and estimates a robustness-baseline SEM model, OLS uses SEM-derived construct scores as a transparent benchmark, and DML-style residualisation tests stability of focal relationships after ML-based adjustment for observed controls. Learner-sensitivity and reverse-direction diagnostics support directional robustness. Demonstrated with a FinTech Digital Customer Intimacy survey model, the framework identifies stable relationships across methods and flags those requiring more cautious interpretation, with reusable Google Colab materials.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages21

Reading Time53 min

Introduction

Research questions

Contributions

arXiv :2607 .005 12v 1 [ cs .LG] 1 Jul 2026 From Structural Equation Modelling to Double Machine Learning: Robustness Analysis for Survey-Based Research Ka Ching Chana , Qiana Liua , Sanjib Tiwaria , Ranga Chimhundua a School of Business, Law, Humanities and Pathways, University of Southern Queensland, Springfield, 4300, Queensland, Australia Abstract Structural equation modelling (SEM) is widely used in survey-based business and information systems research to assess latent constructs and theory-driven structural relationships. However, SEM path significance is obtained within a particular model specification and may not show whether findings remain stable under alternative estimation frameworks. This study develops and demonstrates a staged robustness analysis framework that connects SEM, ordinary least squares (OLS) regression, and Double Machine Learning (DML) . SEM is first used to refine the measurement structure and estimate the robustness-baseline SEM model, in which the full theory-specified structural path system is retained for downstream robustness analysis before final structural path evaluation. OLS regression is then applied to SEM-derived construct scores as a transparent regression benchmark. Finally, DML-style residualisation is used to examine whether each tested focal relationship remains stable after flexible machine-learning-based adjustment for observed controls. Learner-sensitivity checks compare Random Forest, Gradient Boosting, and Support Vector Machine learners, and selected reverse-direction diagnostics are used to examine directional sensitivity. The framework is demonstrated using a FinTech Digital Customer Intimacy survey model. The findings identify which relationships are stable across SEM, OLS, and DML-style checks, and which require more cautious interpretation. A reproducible Google Colab workbook and generated result files are publicly available, providing a reusable template that researchers and students can adapt to other survey-based latent-construct studies. The paper contributes a practical robustness workflow and interpretation guide for survey-based researchers seeking to complement SEM with conventional and machine-learning-based robustness checks. Keywords: Structural Equation Modelling, Double Machine Learning, OLS, Robustness Analysis, Survey Research, Information Systems, FinTech 1. Introduction Survey-based research frequently relies on latent constructs such as trust, satisfaction, attitude, intention, perceived quality, and customer intimacy. These constructs are usually measured through multiple survey items rather than observed directly. Structural equation modelling (SEM) is therefore well-suited to this type of research because it enables researchers to assess the measurement model and estimate theory-driven structural relationships within a single framework. However, a significant SEM path coefficient remains a result within a particular model specification. Researchers may therefore benefit from supplementary robustness analyses that examine whether key relationships remain stable under alternative score representations and estimation assumptions before final model decisions are made. This study develops a staged SEM–OLS–DML robustness framework for survey-based latent-construct research. The framework is organised around two analytical layers. The first layer generates the empirical evidence: the measurement model is refined, the full theory-specified structural model is re-estimated as the robustness-baseline SEM model, construct scores are constructed, and OLS and DML-style robustness checks are run for all SEM paths before final structural path evaluation; SEM factor scores and mean composite scores are constructed, and OLS and DML-style robustness checks are run for all tested SEM paths. The second layer compares and interprets the generated outputs across methods, score representations, nuisance learners, and, where theoretically useful, selec

FAQs

What does the FinTech Digital Customer Intimacy demonstration contribute to the framework’s results?

The FinTech case identifies which initially tested structural paths remain robust across SEM, OLS, and DML-style checks and which paths need more cautious interpretation, theory discussion, measurement refinement, or directional-diagnostic consideration.

Jul 16, 2026

Upload by Himbo