Small area estimation (SAE) is a well-known method to produce reliable estimates for target variables associated with small areas and small sample sizes, and it is remarkably growing for public health applications. Direct survey estimates (DESvy) typically produce imprecise and unreliable estimates since they are ob tained from the target variable of interest under the sampling model. To address this problem, this paper proposes generalized direct survey estimates (GDESvy) by incorporating the survey independent variables into the current DESvy. This approach improves the SAE model estimates by including survey independent vari- ables in the sampling model through the proposed generalized small area estimation (GSAE). To validate and assess the performance of the proposed GSAE model, we first utilized independent variables from the Ethiopian Demographic and Health Survey (EDHS) to produce GDESvy estimates. Subsequently, we employed auxiliary variables from the population and housing census at the local level of Ethiopian administrative zones to provide precise GSAE estimates under the Fay-Herriot model. The results demonstrate that the GDESvy and GSAE estimates outperform the corresponding DESvy and SAE estimates, respectively, for anemia status among children aged 6–59 months by producing lower standard errors. These findings are crucial for informing policy formulation and budget allocation at lower levels of government administration.