a mathematical equation used to approximate values and properties (e.g., bias, consistency, efficiency) for a specific parameter (e.g., regression coefficient) by using information from the sample data and the parameter of interest. This approach is particularly suitable when a likelihood function is not easily known.