Description
The data in the table below are from an investigation into satisfaction with housing
conditions in Copenhagen (Madsen 1971). Residents in selected areas living in rented
homes built between 1960 and 1968 were questioned about their satisfaction and the
degree of contact with other residents.
Low satisfaction Medium satisfaction High satisfaction
Contact with others: Low High Low High Low High
Tower block 65 34 54 47 100 100
Apartment 130 141 76 116 111 191
House 67 130 48 105 62 104
1. Summarize the data using appropriate tables of percentages to show the pair-wise
associations between the levels of satisfaction and 1) contact with other residents
and 2) type of housing. Comment on patterns in the associations.
2. Use nominal logistic regression model for the associations between response variable, the levels of satisfaction, and the other two variables. Obtain a model that
summarizes the patterns in the data. Describe your findings (the pattern in the
associations, odds ratios with 95% confidence intervals, goodness-of-fit). (Hint: use
dummy variable for house types.) Is there interaction of contact level by house
type?
3. As the response has ordinal categories, fit proportional odds model to the data that
include the same variables as used in the nominal logistic model obtained in (ii).
What does the fitted model tell?
4. Calculate Pearson residuals from the proportional odds model for ordinal response
to find where the largest discrepancies are between the observed frequencies and
expected frequencies estimated from the model.
1 P8131 Homework #4
