A Partial Least Squares Structural Equation Model of Parental Support As a Driver of Hope and Life Satisfaction Among Adolescent Females

   Supportive parental/guardian (PG) relationships are thought to be important contributors to an individual’s psychological wellbeing. In fact, individual attachment style is thought to be heavily influenced by the quality of early PG relationships.

   One psychological well-being variable thought to be heavily influenced by the quality of PG support is hope. C.R. Snyder theorizes that hope is a cognitive set of goal-directed expectations that drive resilience. Psychometrically suitable measures of hope have shown that those with higher hope report greater life satisfaction, self-worth, and meaning to life.

   Because of the theoretical importance of PG support as a causal driver of psychological well-being, and hope’s importance to positive mental health, we sought to test a path model with PG support as a driver of hope that results in global life satisfaction.

Methods

   Data was collected from a cross sectional sample of adolescent females (µ = 15.6, SD ± 1.21) collected within high schools (N = 484). The survey was paper and pencil, including scales capturing perceptions of PG support, hope, and life satisfaction.  Partial Least Squares structural equation modeling (PLS-SEM) was used to test the path model with perceptions of PG support as the causal driver of hope,  with hope serving as a mediator between PG support and the endogenous variable of life satisfaction. PLS-SEM is a non-parametric technique that provides greater power with smaller samples than covariance based SEM. Measurement models with PLS-SEM are evaluated by factor loadings of items on latent variables along with composite reliability and average variance extracted scores (AVE).  Structural models are evaluated according to the predicted variance in the endogenous variables and variance accounted for scores (VAF) of proposed mediators in the model. To test for statistical significance of the theorized paths, PLS-SEM incorporates bootstrapping techniques to developed point estimates of model parameters to develop confidence intervals [CI] for those parameters within the population.

Results                                                

   For the measurement model, the composite reliabilities for all the measures exceeded the recommended threshold value of 0.70.   All the item loadings for the constructs were also greater than .40 cut-offs, and all AVE scores met Fornell-Larcker criteria for acceptability. Results suggest the measurement of the latent variables of perceptions of PG support, hope, and life satisfaction.  

   The evaluation of the structural model indicated the model accounted for an R²_adj  = .639 in variance in the endogenous variable of life satisfaction, considered a large effect according to the accepted heuristics of social sciences.  Bootstrapping (N = 5000) indicated all proposed path scores were significant, with the bootstrapped VAF score (0.574) indicating hope was a robust mediator of the relationship between perceptions of PG support and life satisfaction.

   Results suggest the importance of strong PG relationships as a driver of hope and overall life satisfaction among adolescent females. Results are consistent with hope theory that suggests caregivers causally build hope among adolescents, and hope’s importance to overall wellbeing as measured by life satisfaction. Results provide further support for the importance of interventions designed to increase the quality of PG support among adolescents.