Advancing mixed-methods research through PLS-SEM and NVivo: a methodological integration in AI literacy studies

This study examines Partial Least Squares Structural Equation Modelling (PLS-SEM) and NVivo as tools for analysing non-normal data and small samples. PLS-SEM is a robust tool for analysing small samples and offering predictive modelling advantages, while NVivo is a leading qualitative data analysis (QDA) tool. The study uses mixed-methods research, using empirical data to assess PLS-SEM’s performance and surveys, interviews, usability testing, and case studies to evaluate NVivo’s capabilities. PLS-SEM can model latent concepts like customer engagement and efficacy, while NVivo can analyse qualitative data like interview transcripts, reflective diaries, and business discourse. NVivo outperforms competing QDA tools in advanced coding, data visualisation, and integration features, with 72% of surveyed researchers preferring it for its effectiveness and usability. Usability testing revealed NVivo had a 30% higher task efficiency and a high user satisfaction score (8.5/10), despite a moderate learning curve. PLS-SEM is a robust and adaptable statistical method for complex quantitative research, especially when data quality or sample size is constrained; as this research is based on 350 samples from AI literacy in university students. NVivo is a versatile and user-friendly QDA tool, enhancing the rigour and efficiency of qualitative analysis. Together, these tools offer a methodological advancement for researchers undertaking mixed-methods studies, promoting more accurate, predictive, and interpretable research outcomes across disciplines.