As a capstone requirement for my graduate program in Applied Quantitative Research, all students were required to complete an original research project analyzing a data set incorporating multiple measures and time points. For my project, I explored associations between reported measures of social position and depressive symptoms among adolescents using data from Waves I and II the National Longitudinal Study on Adolescent to Adult Health. Models I explored included linear ordinary least squares, fixed effects and random effects. I used Stata, version 13 to run my analysis.
Overall, the data indicated that parental educational attainment and subjective social status are each negatively associated with depressive symptoms at a 99.9% confidence level across most model types. This said, given limitations in variable measurements and an inability to eliminate possible violations of model assumptions, most notably temporal ordering threats, we should be cautious about making any further conclusions about these findings beyond identification of association without conducting additional research.
The full study can be accessed through my personal website, sarah-noonan.com, and this repository contains the code files I used to run my analysis. Visit https://addhealth.cpc.unc.edu/data/#public-use for instructions on how to download Waves I and II of the data, then navigate to Thesis/Do-Files/NoSyntax/WorkflowNoSyntax.do and follow the steps as directed to replicate my analysis. I would love to hear your feedback on my work and any ingsights you uncover from the project!