What is the response (Y) of interest and how is it measured? Total costs is the response variable found in the problem statement. It is a continuous variable.

Are predictor variables mentioned in the problem statement? If so, how many and what are their measurement levels? The problem statement requests a simultaneous prediction of total costs from length of stay and birthweight. Both of these predictor variables are continuous.

What are you being asked to deliver? A data description, an interval estimate, an answer to a question, or a predictive model? You are being asked to quantify the relationship between total costs and length of stay and birthweight. A predictive equation will quantify this relationship. A multiple linear regression is a good starting point for developing a predictive equation. Multiple regression is a flexible modeling method to predict a continuous variable from nominal, ordinal, or continuous independent variables.