PLS-SEM technique and phases of analysis – implications for information systems’ exploratory design researchers

Abstract

Increasingly, information systems (IS) practitioners and theorists, knowingly or unaware, inappropriately use, and incorrectly apply the partial least squares - structural equation modeling (PLS-SEM) approach in analysis. Associated with the ease of access to statistical packages (RATS and SmartPLS), IS analysts tend to abuse PLS-SEM reporting and interpretations. As a consequence, often insincere, and even meaningless and misleading conclusions, as well as recommendations, are attained, leading to misleading remedial actions. Policymakers without the necessary knowledge of the integrity of PLS-SEM tools, coupled with their lack of awareness of the problems, tend to implement such misleading results in their decision-making, often resulting in disastrous consequences unbeknown to the IS community of practice, and including the general public. Thus, the achievement of valid and objective conclusions drawn from PLS-SEM relate closely with the awareness of the correct PLS-SEM techniques and their consequences. Informed by the misleads, in the present conceptual paper, attention is paid to incorrect practices related to analysis procedures of PLS-SEM as well as the analysis technique in IS related research. We link the discussion with how to predict the relevance of hypothesised models through the selected Technology Acceptance Model (TAM) for illustrative purposes. We end the paper with reflective lessons as applied in IS research.