Class A: Application of Structural Equation Modelling (SEM) in mHelath Studies
Dr. Hashem Salarzadeh Jenatabadi
Department of Science and Technology Studies, Faculty of Science, Universiti Malaya, Kuala Lumpur; 50603; Malaysia
Abstract
mHealth has been identified as a health application and public medical which is to be paired with mobile devices (Chandler et al., 2019). Some literatures in this area concentrated on mHealth apps for health purposes. The potential benefits in precautionary health, containing healthcare and the challenges of using the apps (Bol, Helberger, & Weert, 2018; Furlong, Morris, Serry, & Erickson, 2018). Most statistical analyses of fitness apps interconnect other aspects (e.g. users’ perceptions, health behavior, trajectories of app use, quality assessment, app evaluation, and purpose of downloading the apps) that mostly include descriptive statistics (Bender et al., 2014; Chen, Cade, & Allman-Farinelli, 2015; Payne, Moxley, & MacDonald, 2015; Sama, Eapen, Weinfurt, Shah, & Schulman, 2014) and regression (Goh et al., 2015). Along with prior studies, a few types of research on mHealth apps with Structural Equation Modelling (SEM) technique have been explored (Hoque, Karim, & Amin, 2015; Lee, Han, & Jo, 2017; Mburu, 2017). This study is about some lessons of SEM in mHealth Studies. We introduce a new framework of fitness apps associated with obesity modeling by using SEM and examining data of postpartum women through moderation and mediation analysis. Fitness mHealth apps are considered the moderator in the research framework. Online questionnaires were sent to Malaysian postpartum women within one year after pregnancy and 468 completed questionnaires were returned. The frequency of mHealth fitness app use was categorized into four groups: daily, weekly, rarely and never. Therefore, four models were considered for the final analysis. This study will help data scientists to interpret their modeling with application of SEM.