Online Program Home
My Program

Abstract Details

Activity Number: 86
Type: Contributed
Date/Time: Sunday, July 31, 2016 : 4:00 PM to 5:50 PM
Sponsor: Section on Nonparametric Statistics
Abstract #321162
Title: A Structural equation model for a dental health related quality of life framework
Author(s): Ana Nora Donaldson* and Nairn Wilson and James Wallace and Nelarine Cornelius and Angelo Passalacqua
Companies: SUNY Stony Brook and King's College London and Bradford University and Bradford University School of Management and King's College London
Keywords: SEM ; Partial Least Squares ; Quality of life outcomes ; Public Health

Background: Covariance-based (CB) structural equation modelling (SEM), implemented in Lisrel, EQS and AMOS, has been the default SEM approach but partial least square (PLS-SEM) is a new approach which offers, relatively, more flexibility.

Aim: We present the use of PLS-SEM in the context of assessing the factors that drive oral health related quality of life (QOL).

Methods: PLS-SEM is used on baseline data of 149 patients going for tooth extraction at a London dental clinic. QOL was measured using a well-established survey, the OHIP-14. Factors considered are: socio-demographics, oral health, health service availability, dental anxiety, locus of control (self-care and distrust in dentists) and dental-related knowledge, attitudes and behaviours.

Results: Oral health is the most significant direct driver of QOL, followed by dental anxiety. Ignorance (of harmful/beneficial practices) showed a marginal effect. Indirect effects are identified for health service (via oral health), dental anxiety (via oral health) and Ignorance (via dental fear). Mediating relationships are found between age and oral health, and between ignorance and dental anxiety. No significant effect of distrust-dentists, self-care or any other socio-demographic is found.

Conclusion: We verified hypothesized relationships. PLS-SEM is. Under multivariate normality PLS-SEM produces similar results to CB_SEM but PLS-SEM offers more flexibility: it is distribution-free and able to model complex relationships with relatively smaller sample sizes. PLS-SEM is an excellent complement to CB-SEM.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association