IBaby Month One: Maternal Depression, Stress, and Technology Use
Overview
This project examined whether parenting stress and tech-related distraction are associated with depression risk one month postpartum.
Data
Sample: 65 mothers assessed one month after birth.
Outcome: CES-D–based depression indicator (score ≥ 13).
Predictors: parenting stress, perceived distraction, age, income, and marital status.
Given sample size, we used summarized scale scores (instead of item-level inputs) and removed cases with substantial missingness.
Methodology
We fit logistic regression for depression risk, used backward stepwise selection to keep the model interpretable, and standardized continuous predictors so odds ratios were easier to compare.
Key Findings
Higher parenting stress and higher perceived distraction were both associated with higher odds of depression. Maternal age was negatively associated with depression risk. Income and marital status were retained, suggesting possible socioeconomic/context effects.
Model Performance and Diagnostics
The final model showed strong in-sample discrimination (ROC AUC ≈ 0.91). Diagnostics did not suggest major assumption violations or severe multicollinearity.
Interpretation and Limitations
Findings support a meaningful association between stress/distraction and early postpartum depression, but this is observational work with a modest sample. Effect sizes should be interpreted cautiously, especially for demographic terms that may be sensitive to sample imbalance and unmeasured confounding.
Tools
SAS (logistic regression, model selection, diagnostics).