Validating Measurements of Veteran Community Reintegration (VCR) Risk & Resilience Factors to Empower Veterans, Their Children, Families and Our Communities

 

Background and Purpose: As the wars in Iraq and Afghanistan wind down, veterans returning to their communities face increased risk factors for Veteran Community Reintegration (VCR) difficulties due to multiple deployments. This study aims to validate measurements commonly used in practice settings to assess VCR difficulties among a sample of participants (N=131) outside the VA setting. The following research questions are addressed: (1) What is the reliability coefficient for the following scales: Military to Civilian Questionnaire (M2CQ); Primary Care PTSD scale (PC-PTSD); Center for Epidemiological Studies Depression Scale-Short Version (CESD-10); Suicide Behavioral Questionnaire-Revised (SBQ-R) & Ultra-Short Form Suicide Measure (USSTM); Brief Sense of Community Scale (BSCS); (2) What is the construct validity of the scales among the sample. Methods: Primary data were collected from a purposive sample at a career fair, using a 77-item Veteran Community Reintegration Survey (VCRS). Data screening revealed a large amount of missing data (46.6%) and violated multivariate normality assumptions. Reliability was established (Cronbach’s alpha) and construct validity tested using CFA. Maximum Likelihood (ML) estimation was used, and the following model fit indices were used to analyze data in AMOS: Root Mean Square Error of Approximation (RMSEA); Tucker Lewis Index (TLI); Comparative Fit Index (CFI). Results: Reliability analyses indicated good internal consistency of scores on the following scales among the sample: M2C-Q (a = .90); PC-PTSD (a = .88); USSTM & SBQ-R (modified to 1 item each per IRB) (a = .75); CESD -10 (a = .85); BSCS (a = .91). Results of the CFA indicate that the model does not fit the data very well χ² (761 df N=131) = 1480.072, p =.000 RMSEA = .085, PCLOSE =.000 TLI= .721, and CFI =.754. However, all the loadings of the observed variables on the latent variables were statistically significant (p < .001). The factor loading between the construct of depression symptoms and the construct of SOC was statistically significant (r = .215, p < .001) as was the factor loading between the construct of depression symptoms and the construct of VCR difficulties (r = .315, p < .001). Additionally, the factor loading between the construct of PTSD symptoms and the construct of SOC was statistically significant (r = -.100, p < .001) and the factor correlation between the constructs of PTSD symptoms and VCR difficulties (r = -.129, p < .001). The factor loading between suicidal ideation and VCR difficulties was also statistically significant (r = .064, p < .001). Lastly, a zero-order factor correlation between SOC and VCR difficulties was statistically significant (r = .296, p < .001). Results from the CFA indicate good convergent and divergent validity. Conclusions and Implications: Good internal consistency was found on scales used in the VCRS scales. Regarding construct validity, results indicated good convergent and divergent validity, and Model fit indices met the cut off scores. Results suggest the continued use of scales on VCRS to measure risk and resilience factors among veterans.  Future research should expand validation efforts of practice measures used with veterans and other diverse populations.