The data are from a clinical trial of patients with respiratory
illness, where 111 patients from two different clinics were randomized
to receive either placebo or an active treatment. Patients were
examined at baseline and at four visits during treatment. At each
examination, respiratory status (categorized as 1 = good, 0 = poor)
was determined.