Dataset : Generative AI in HRM: Enhancing Performance Using TPC Model with ChatGPT Literacy as Moderating Variable
Description
This dataset supports the study “Generative AI in HRM: Enhancing Performance Using TPC Model with ChatGPT Literacy as a Moderating Variable.” It contains survey data from 381 HR professionals in Indonesia examining how ChatGPT literacy influences performance through the Technology-to-Performance Chain (TPC) framework. The dataset includes demographic characteristics of respondents and full SmartPLS 4 outputs: lower-order and higher-order construct analyses, factor loadings, reliability and validity tests (Cronbach’s Alpha, Composite Reliability, AVE), and Heterotrait-Monotrait (HTMT) ratios. This dataset provides comprehensive empirical evidence for the moderating role of ChatGPT literacy in enhancing performance outcomes through task–technology alignment in HRM practices. It serves as a valuable reference for researchers and practitioners exploring the application of generative AI and digital literacy in organizational performance contexts.
Files
Steps to reproduce
The data was collected through an online survey distributed to HR professionals across Indonesia after receiving permission to conduct the study. Respondents were selected using purposive sampling, focusing on individuals actively working in Human Resource Management (HRM) functions and with experience using ChatGPT in their professional tasks. A total of 381 valid responses were obtained, consisting primarily of HR specialists and generalists. The main objective of the study was to examine how generative AI—specifically ChatGPT—enhances performance outcomes in HRM through the Technology-to-Performance Chain (TPC) model, with ChatGPT literacy serving as a moderating variable. The constructs measured in the dataset include Task Characteristics (TC), Technology Characteristics (TE), Task-Technology Fit (TTF), ChatGPT Use (CGU), ChatGPT Literacy (GPTL), and Performance Impact (PI). Data were analyzed using the Partial Least Squares Structural Equation Modeling (PLS-SEM) method via SmartPLS 4 software, following Hair et al. (2019). The analysis covered both measurement and structural model assessments, including factor loadings, reliability, validity, and hypothesis testing.
Institutions
- Binus University - JWC Campus