IEEE Access (Jan 2020)
Analysis of Operational Communication Through Structural Equation Modeling
Abstract
In this study a model was developed for analyzing communication on the shop floor using a Structural Equation Modeling with Partial Least Squares estimation (Partial Least Square, PLS). The model included as an exogenous variable (independent) Knowledge Management (KM), and four endogenous variables (dependent): Written Communication (WC), Oral Communication (OC), Man interaction with Information and Communication Technologies (M-ICT), and Operational Communication (OpC). The data for the exploratory study were collected in a land-based oil production region in the Northeast of Brazil. The data analysis was performed using the SmartPLSR software. The structural model supported six of the nine hypotheses proposed, presenting statistical significance and predictive relevance. The evaluation of Pearson's determination coefficients (R2) demonstrated a satisfactory degree of adjustment and adherence in explaining the variance of endogenous variables. The effect size (f 2) of the variance explained by the exogenous variable in the four endogenous variables was classified as “very strong”, as shown: in WC f 2 = 0.802; OCf2 = 0.825; M-ICT f 2 = 2, 534; and OpCf2 = 2.003. The model presented predictive validity, KM showed effect (q2) classified as “very high”in the M-ICT and OpC. The effect (q2) of KM on WC and OC presented predictive relevance between medium and high, with q2 = 0.337 and q2 = 0.267, respectively. The causal relationships between the observed variables and the percentage of participation of each indicator in its construct are contributions to the management of information about production.
Keywords