A clinical investigator is interested in understanding the relationship between demographic factors and congestive heart failure, measured by a cardiac index (CI). A random sample of 12 patient charts of CHF patients revealed the data below. Using the five-step hypothesis test at the .05 level, determine what demographic factors best explain CHF.
a. Is there evidence of multicullinearity?
b. Are there any variables that do not explain CHF? If so, remove them from the model and rerun it (but show all model tests).
c. What is the overall fit of the model (how much of the variance of CHF is explained)?
d. Which variables best explain CHF?
e. What would be the expected CI for a patient that is 50 years old, controlling for all other variables
adapted from Ott, R. L. & Longnecker, M. (2001). Statistical methods and data analysis (5th ed.). Pacific Grove, CA: Wadsworth Group.
Patient CI age disease duration ht wt gender
A1 1.6 67 5 172 57 1
A2 2.4 45 2 170 87 0
A3 2.2 59 8 172.7 102 0
A4 1.7 63 1 175.3 74 1
A5 2.3 55 1 172.7 92 0
A6 1.6 65 1 178 90 1
A7 1.4 62 2 163 57 0
A8 2.2 60 1 182.5 80 0
A9 1.3 72 2 168 72 1
A10 2.1 44 3 163 75 1
A11 2.1 63 5 172 82 0
A12 1.1 63 1 163 64 1
CI = L/min/m2
age in years
disease duration in years
ht in cm
wt in kg
gender = 0 men; 1 women