Birthweight Multivariate Analysis

This project explores relationships between various biological and behavioral factors (e.g., maternal smoking, parental height, baby's length) and newborn birthweight using statistical methods and data visualization in Python.

---

Objectives

• Investigate whether variables like father’s age, maternal smoking, and parental height are associated with birthweight
• Compare subgroups using statistical tests
• Build regression models to predict baby’s length and birthweight
• Communicate findings using clear plots and interpretations

---

Techniques Used

• Shapiro-Wilk Test (Normality)
• Spearman & Pearson Correlation
• Boxplots & Scatterplots with Regression
• Multiple Linear Regression using statsmodels
• Hypothesis Testing: Levene’s test, Independent T-Test

---

Sample Insights

• Babies of non-smoking mothers tend to have higher birthweight
• Baby length is strongly correlated with birthweight (r ≈ 0.74)
• Father’s height is not a significant predictor of baby length
• Mother’s height is positively and significantly related to baby’s length

Author

Lohith Basavanahalli Anjinappa