The CILS sampled two metropolitan areas of the United States and while these areas are relevant in terms of understanding the experiences of Latino adolescents, they exclude the subset of the Latino population living elsewhere. In terms of measurement, the items used to measure perceived discrimination did not consider the complexity of the discrimination construct or the experience of discrimination.
Finally, despite our efforts to reduce endogeneity bias (e.g., using longitudinal data, controlling for baseline depressive symptoms and self esteem, controlling for a wide variety of variables to address selection bias), we cannot ensure that our results imply causality between any measure of perceived discrimination and either mental health outcome.