Despite over a century of research on mental health trends in immigrant populations, the relationship between acculturation and mental health remains largely a mystery. The most prominent example of this ambiguity is known as the Hispanic (Mental) Health Paradox, whereby foreign-born immigrants initially have better health/mental health outcomes than native-born ethnics of similar class, income, arid education backgrounds, but overtime these differences disappear. Previous research has attributed this change to """"""""acculturation"""""""" and found varying results. I argue that the variable of interest is not acculturation, but social network ties. This research project will use social networks to bridge stress theory and social capital migration theory to forge an integrated approach to the study of mental health in immigrant populations. After drawing a conceptual distinction between resource ties (e.g., ties that give instrumental, informational, emotional support to an individual) and obligatory ties (e.g., ties that take instrumental, informational, emotional support from an individual), I will begin by testing their effects on mental health outcomes using measures of depressive symptomology, controlling for all relevant variables. I will then add indicators of acculturation (measured by language use/comprehension and media exposure) to the model and examine whether it has any direct effect or mediates any of the social network effects. I will conduct formal multiple-group tests examining differences across country of origin, gender, sending context, and SES. Using the Wave II data from the New Immigrant Survey, in conjunction with the Wave I data, I will test a difference model incorporating acculturation, network ties, and my control variables at time 1 and 2 to add strength to the causal claims being made. All analyses will use structural equation modeling techniques, which allow latent variable measurement models to be specified that directly incorporate (and reduce) measurement error. Using this methodology I will include a one-factor measurement model of acculturation with 12 indicators and a second-order measurement model of depression with eight indicators. Additionally, standard estimation and missing data techniques will be used. ? ? ? ?
