An assessment of the use of partial least squares structural equation modeling (PLS-SEM) in hospitality research

Abstract

Purpose -Structural equation modeling (SEM) depicts one of the most salient research methods across a variety of disciplines, including hospitality management. Although for many researchers, SEM is equivalent to carrying out covariance-based SEM, recent research advocates the use of partial least squares structural equation modeling (PLS-SEM) as an attractive alternative. The purpose of this paper is to systematically examine how PLS-SEM has been applied in major hospitality research journals with the aim of providing important guidance and, if necessary, opportunities for realignment in future applications. Because PLS-SEM in hospitality research is still in an early stage of development, critically examining its use holds considerable promise to counteract misapplications which otherwise might reinforce over time.Design/methodology/approach -All PLS-SEM studies published in the six SSCI-indexed hospitality management journals between 2001 and 2015 were reviewed. Tying in with the prior studies in the field, the review covers reasons for using PLS-SEM, data characteristics, model characteristics, the evaluation of the measurement models, the evaluation of the structural model, reporting and use of advanced analyses.Findings -Compared to other fields, the results show that several reporting practices are clearly above standard but still leave room for improvement, particularly regarding the consideration of state-of-the art metrics for measurement and structural model assessment. Furthermore, hospitality researchers seem to be unaware of the recent extensions of the PLS-SEM method, which clearly extend the scope of the analyses and Even though this research does not explicitly refer to the use of the SmartPLS software (www. smartpls.com), Ringle acknowledges a financial interest in SmartPLS.

Indigenous participation in conservation and tourism development: A case of native people of Sabah, Malaysia

“…Using PLS‐SEM to assess a model, we need to conduct a two‐step process that involves assessment of the measurement model, followed by assessment of the structural model (Ali, Rasoolimanesh, Sarstedt, Ringle, & Ryu, in press; Hair et al, ). To assess the measurement model, the validity and reliability of the latent variables (LVs) in the model should be checked, whereas the structural model assessment involves checking the relationships between these LVs (Chin, ; Hair et al, ).…”

“…Table shows that most of item loadings on their associated LVs were higher than 0.7 and that only a few items loaded lower than 0.7. If the loading is between 0.4 and 0.7 and the composite reliability (CR) and average variance extracted (AVE) are higher than 0.7 and 0.5, respectively, we can retain the items (Ali et al, in press; Hair et al, ). Therefore, the indicator reliability for all LVs in this study was acceptable.…”

“…Assessment of the measurement model requires the assessment of convergent and discriminant validity (Hair et al, ). Convergent validity is assessed using AVE (Ali et al, in press; Chin, ), and the value of the AVE of the LVs should be higher than 0.5 to establish convergent validity (Bagozzi, Yi, & Phillips, ; Hair et al, ). Table shows the acceptable convergent validity for all constructs due to the AVEs being higher than 0.5.…”

“…Table shows the acceptable convergent validity for all constructs due to the AVEs being higher than 0.5. The next step in the measurement model assessment involves the determination of discriminant validity, which refers to how distinct each LV is from other LVs in the model (Ali et al, in press; Hair et al, ). Several criteria have been developed to assess discriminant validity; however, for the sake of this study, we prefer the two most conservative approaches: the Fornell‐Larcker criterion and the heterotrait‐monotrait (HTMT) ratio (Henseler, Ringle, & Sarstedt, ; Voorhees, Brady, Calantone, & Ramirez, ).…”

Trend of NFC Technology for Payment Transaction

“…To get satisfactory discriminant validity, the square root of AVE for each construct should be greater than the squared correlation between the constructs [20].…”