Validity and Reliability Testing on Train Driver Performance Model Using a PLS Approach

Azlis-Sani, Jalil; Dawal, Siti Zawiah Md.; Zakuan, Norhayati · 2013 · Crossref

DOI: 10.4028/www.scientific.net/aef.10.361

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the methodological validation of a survey instrument designed to assess the human performance of train drivers in Malaysia. Recognizing that railway operations remain heavily dependent on human operators, the authors emphasize the critical need to systematically evaluate factors influencing driver performance, such as occupational stress, fatigue, and work environment interactions. The primary objective is to demonstrate the validity and reliability of the measurement constructs used in a broader research model, ensuring that the survey items accurately and consistently measure the intended variables. The research employed a survey methodology, collecting data from 229 locomotive and junior drivers across five railway depots in Peninsular Malaysia (Perai, Ipoh, Kuala Lumpur, Gemas, and Kuala Lipis). Participants completed self-administered questionnaires using a 5-point Likert scale. The instrument was adapted from established literature covering fatigue, stress, and safety. To evaluate the "goodness of measures," the authors utilized Partial Least Squares (PLS) structural equation modeling. The analysis followed a two-step approach: first assessing the measurement model for validity and reliability, and second, examining the structural relationships. The constructs evaluated included Activity, Human factors, Performance, Safety, and Work Environment. The results confirmed the psychometric robustness of the survey instrument. Construct validity was established through convergent and discriminant validity tests. All item loadings exceeded the 0.5 threshold, and cross-loadings indicated that items measured their intended constructs more strongly than others. Convergent validity was supported by Composite Reliability (CR) values ranging from 0.750 to 0.879, all above the recommended 0.7 cutoff, and Average Variance Extracted (AVE) values ranging from 0.501 to 0.590, exceeding the 0.5 threshold. Discriminant validity was confirmed as the square root of the AVE for each construct was higher than its correlations with other constructs. Reliability was further verified using Cronbach’s alpha, with values ranging from 0.508 to 0.846. The study concludes that the PLS approach effectively validates the measurement model for assessing train driver performance. By demonstrating adequate validity and reliability, the paper provides a verified instrument for future research into human factors in railway operations. The findings offer a methodological guide for researchers conducting survey-based studies in ergonomics and human performance, illustrating how PLS-SEM can be used to rigorously test measurement models before analyzing structural relationships. This validation ensures that subsequent analyses of how stress, fatigue, and environmental factors impact driver performance are based on sound, reliable data.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• dbq psychometrics
• workload measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Methodological Resource: validation psychometrics