Were the neighborhoods specified correctly?
Was personal socio-economic position adequately controlled for?
What about the interpretation of between-neighborhood variation, σ2u0?
Were other individual-level confounders adequately controlled for?
Do I need to use multilevel statistical methods?
What about endogeneity?
What about the sample of neighborhoods?
Rice, N., Jones, A., Goldstein, H. (1998) Multilevel models where the random effects are correlated with the fixed predictors: A conditioned iterative generalised least squares estimator (CIGLS). Multilevel Modelling Newsletter 10: 10-14.
Duncan G., Raudenbush, S. (1999) Assessing the effects of context in studies of child and youth development. Educational Psychologist 34: 29-41.
Spencer, N. (1998) Consistent parameter estimation for lagged multilevel models. Hertfordshire: University of Hertfordshire Business School. UHBS 19. 4-7 p.
Spencer, N., Fielding, A. (2000) An instrumental variable consistent estimation procedure to overcome the problem of endogenous variables in multilevel models. Multilevel Modelling Newsletter 12: 4-7.
Heagerty, P., Zeger, S. (2000) Marginalized multilevel models and likelihood inference (with discussion). Statistical Science 15: 1-26.
Rasbash, J., Goldstein, H. (1994) Efficient analysis of mixed hierarchical and cross-classified random structures using a multilevel model. Journal of Educational and Behavioural Statistics 19(4): 337-350.
Morris, C. (1995) Hierarchical models for educational data: An overview. Journal of Educational and Behavioral Statistics 20: 190-199.
Draper, D. (1995) Inference and hierarchical modeling in the social sciences. Journal of Educational and Behavioral Statistics 20: 115-147.