Examining Data Quality in Communication Model: A Partial Least Squares (PLS) Approach to Estimating the Reliability and Validity of the Constructs in the Relationship between Media Literacy Competencies and Use of New Media for Civic Engagement

Abstract

This study applied the Partial Least Squares PLS to test reliability of the instrument used for data collections and to establish the convergent validity and discriminant validity of the constructs specified for study. The aim is to ascertain the performance of PLS in measuring latent constructs (unobserved) and the manifest (observed) indicators in communication and media studies. Communication studies like other fields often involve the modeling of constructs relationships which requires higher level of analysis. The purpose of this study is to demonstrate the potentials of Partial Least Squares PLS approach to estimating the reliability and validity of measurements items and their latent constructs in a communication model. The motivation for the study was the observation that there is inconsistent application of PLS SEM in communication studies leading to low appreciation of its potentials in the field of communication and behavioral areas. PLS offers simple but robust options. In applying PLS SEM, the results enable the researcher to show that the specified model relationship contains measurement items that have strong relationship with their corresponding latent constructs. They also indicate convergent and discriminant validity of the constructs. The implication of the result is that PLS will assist communication and media studies that are designed as exploratory theoretical model that will lead to theory building

Keywords: partial least squares, media literacy, civic engagement, new media