1.A study was conducted to assess the association between certain pollutants/behaviors and respiratory function. The response measured was chronic respiratory disease status where subjects were assigned to one of four possible categories (Level 1 = no symptoms, Level 2 = cough or phlegm 3 months per year, Level 3 = cough or phlegm > 3 months per year, Level 4 = cough or phlegm and shortness of breath > 3 months per year). The explanatory variables of interest are Air Pollution (air = low, high), Job Exposure to Pollution (exposure = no, yes), and Smoking Status (smoking = non, ex, current).The SAS data step for this data set is given below:

data respire;

input air $ exposure $ smoking $ level count @@;

datalines;

low no non1 158 lownonon 29

low no ex1 167 lownoex219

low no cur1 307 lownocur 2 102

low yes non126 lowyes non 25

low yes ex138 lowyes ex212

low yes cur194 lowyes cur 248

high no non194 high nonon27

high no ex167 high noex28

high no cur1 184 high nocur2 65

high yes non 132 high yes non 23

high yes ex139 high yes ex211

high yes cur 177 high yes cur 248

lownonon 35lowno non40

lownoex35lowno ex43

lownocur 3 83lowno cur4 68

lowyes non 35lowyes non 41

lowyes ex34lowyes ex44

lowyes cur 3 46lowyes cur 4 60

high nonon 35high nonon 41

high noex34high noex43

high nocur 3 33high nocur 4 36

high yes non 36high yes non 41

high yes ex34high yes ex42

high yes cur 3 39high yes cur 4 51

;

run;

Construct a proportional odds model using level as the response variable.Select only the appropriate predictors in the final model.After the final model is constructed, compare each level of response against the level 1 symptom group.Based on the final model, report your summary findings - include hypotheses, descriptive statistics, interpret the parameters, discuss your findings and provide conclusion.Two-sided alpha is set at 0.05