This is data from the Study on the Efficacy of Nosocomial Infection Control (SENIC).The primary objective of SENIC was to determine whether infection surveillance and control programs have reduced the rates of nosocomial (hospital-acquired) infection in hospitals in the United States.The data are from a random sample of 113 such hospitals. For this problem, length of stay (LOS) is the dependent variable. The goal of the analysis is to develop the multiple regression (MLR) model that relates LOS to the "best" subset of the independent variables listed in the following table.

Variable Label Variable Description Age Average age of all patients in the hospital in years infectionrisk Average estimated probability of acquiring infection in the hospital in percent culturingratio Ratio of number of cultures performed to number of patients without signs or symptoms of hospital acquired infection x 100 chestxravratio Ratio of number of x-rays performed to number of patients without signs or symptoms of hospital acquired infection x 100 numbeds Average number of beds in the hospital averagedailycensus Average number of patients in the hospital per day numnurses Average number of full-time equivalent RNs and LPNs facilitiesservices Percent of 35 potential facilities and services that are provided by the hospitalParameter Estimates Parameter Standard Variance Variable Label DF Estimate Error t Value Pr> |t Inflation Intercept Intercept 0.49068 1.66543 0.29 0.7689 age 0.09601 0.02925 3.28 0.0014 1.09359 infectionrisk infectionrisk 0.33555 0.13097 2.56 0.0118 1.98060 culturingratio 0.02713 9060 0 culturingratio 0.01588 1.71 1.69731 cheatxrayratio chestxrayratio 0.01867 0.00751 2.49 0.0145 1.35707 numbeds -0.00958 0.00362 -2.65 0.0094 31.25379 axeragedailycensus averagedailycensus 0.02163 0.00429 5.05 <.0001 numnurses facilitiessery prices facilitiesser collinearity diagnostics condition proportion of variation number eigenvalue index intercept age infectionrisk culturingratio chestxrayratio numbed>