Is there any evidence that multicollinearity is a problem in the analysis done in Problem 12.5? Explain your answer.
Problem 12.5
Adult respiratory distress syndrome (ARDS) is a complication in many critically ill patients. The usual diagnosis of ARDS is based on clinical findings of refractory respiratory failure and x-rays of the lungs showing fluid accumulation. Because it is known that the airways of the lungs are leaky toward plasma proteins in ARDS, accumulation of protein in the fluids in the lung might help diagnose ARDS, particularly because it is possible to noninvasively determine protein accumulation using nuclear imaging of the chest when plasma proteins have been radioactively labeled. Rocker and colleagues † used lung images obtained after labeling the plasma protein transferrin in patients meeting the clinical criteria for ARDS and in patients who did not to calculate a lung protein accumulation index P that is larger as there is more protein in the lungs. They also recorded other characteristics of these patients, including their sex S, age A, x-ray lung fluid score X, and amount of oxygen in their blood O (the data are in Table D-40, Appendix D). A. Is there evidence that this noninvasive index of protein accumulation in the lungs is associated with ARDS? Include the other variables in the analysis to control for potential confounding. B. Interpret any significant regression coefficients.
Table D-40
Already registered? Login
Not Account? Sign up
Enter your email address to reset your password
Back to Login? Click here