Fig 1.
Study design flow chart.
Table 1.
Latent constructs and indicator variables explored in the SEM analysis.
Fig 2.
Path diagram for the basic conceptual model of Extent of Prosthesis Adoption.
Ovals represent latent constructs modeled through CFA. Rounded white rectangles are indicator measures comprised of >1 self-report items while white rectangles with corners are single item indicator variables. Grey rectangles are covariates. Solid lines represent regression coefficients for direct effects, dotted lines indicate factor loadings of indicator variables onto latent constructs. The double-headed arrow represents a correlation effect.
Table 2.
Demographic and prosthetic characteristics of the analytic sample for structural equation modeling.
Fig 3.
SEM initial model of Extent of Prosthesis Adoption showing significant direct effects only.
Ovals represent latent constructs modeled through CFA. Rounded white rectangles are indicator measures comprised of >1 self-report items while white rectangles with corners are single item indicator variables. Darker grey rectangles are covariates. Note that the sub-categories of amputation level, laterality, and prosthesis type have been included as lighter grey rectangles within their respective covariate. Solid lines represent regression coefficients for direct effects, dotted lines indicate factor loadings of indicator variables onto latent constructs. The double-headed arrow represents a correlation effect. Note that gender, race, and age did not have significant direct effects on Extent of Prosthesis Adoption.
Fig 4.
Final SEM model of Extent of Prosthesis Adoption showing all significant direct, moderating, and mediating effects.
Ovals represent latent constructs modeled through CFA. Rounded white rectangles are indicator measures comprised of >1 self-report items while white rectangles with corners are single item indicator variables. Darker grey rectangles are covariates. Note that sub-categories of amputation level, laterality, and prosthesis type have been included as lighter grey rectangles within their respective covariate. Solid arrows represent regression coefficients for direct effects, and dashed arrows represent regression coefficients for moderating effects. Dotted lines indicate factor loadings of indicator variables onto latent constructs. The double-headed arrow represents a correlation effect.
Table 3.
Distribution of Extent of Prosthesis Adoption and prosthesis usability experience latent construct scores and indicator variable scores by amputation level/laterality.
Table 4.
Prosthesis usability experience and Extent of Prosthesis Adoption: Latent construct and indicator variable scores by prosthesis type (N = 402).