Negative Binomial Regression Analysis of Factors Influencing Stunting Cases in Central Lombok Regency

Poisson regression is commonly used to model count data, relying on the crucial assumption of equidispersion, where the mean and variance are equal. However, this assumption is often violated in real-world data, which can exhibit overdispersion or underdispersion. When this occurs, the standard Poisson model becomes unsuitable, leading to biased and inaccurate parameter estimates. To address overdispersion in count data, Negative Binomial Regression (NBR) is a viable alternative, as it incorporates an additional parameter to account for variability greater than the mean. Stunting, a condition characterized by significantly impaired growth in infants, has been a primary concern for the Indonesian government during the 2019-2024 period, particularly in Central Lombok district. Reducing stunting rates is critical to ensuring an optimal quality of life for future generations. Despite extensive research on stunting, the application of NBR to analyze factors influencing stunting cases in Central Lombok Regency has not yet been explored. This study aims to implement the NBR model to identify the determinants of stunting in Central Lombok. Data were collected from 29 community health centers (PUSKESMAS) in Central Lombok. The findings indicate that an increase in the number of malnourished toddlers is associated with a corresponding rise in stunting cases. Similarly, a higher prevalence of low-birth-weight infants is linked to an elevated incidence of stunting.