APPENDIX 3 The following analysis looks at the quality of living conditions among families living in 3 major c in Thaland In this analysis, quality of living conditions (gols ord-o if low standard of living. 1 if medium, 2 high standard of living) was examined as the response with the following predictors: 12 years Father's Education: feduct 8 if <-8 years of education, 11 it9-11 years, 12 Mother's Education: meduc1-8 if c-8 years of education, 11 if 9-11 years, 12 if-12 years Country where mother was born: countryBorn-1 if Thailand; 2 if Burma Read and write in Thai: readwrite-1 if yes, 0 if no Income: in 10,000 Bhat. So, a value of 9 would indicate the family's income was 90,000 Bhat. SAS Output is given below. Data Set The LOGISTIC Procedure Model Information Response Variable WORK QOLS gols ord cumulative logit Number of Response Levels) Model Optimization Technique Fisher's scoring Number of Observations Read320 Number of Observations Used275 Response Profile Valueqols_ord Frequency Ordered 12 21 30 Total 98 87 Probabilities modeled are cumulated over the lower Ordered Values. Note: 45 observations were deleted due to missing values for the response or explanatory variables Class feducl Class Level Information Value Design Variables 8 11 0 12 0 1 meducl 8 0 0 11 1 0 12 0 1 country Bom 1 readwrite Model Convergence Status Convergence criterion (GCONV-11-8) satisfied. Score Test for the Proportional Odds Assumption Chi-Square DF Pr> Chiq 8.6237 7 0.2808 Model Fit Statistics Criterion Intercept Only Intercept and Covariates AIC SC -2 Log L 607.537 614.770 603.537 579.436 611.987 561.436 Testing Global Null Hypothesis: BETA-0 Test Chi-Square DF Pr> ChiSq Likelihood Ratio 42.1006 7 <.0001 Score 36.8569 7 <.0001 Wald 37.1203 7 <.0001 Effect feduel Type 3 Analysis of Effects Wald DF Chi-Square Pr> ChiSq 2 4.9680 0.0834 meduc1 2 0.0485 0.9760 country Born 1 10.3629 0.0013 readwrite Income 1 7.6497 0.0057 17.6500 <.0001 Analysis of Maximum Likelihood Estimates Standard Parameter DFEstimate Error Chi-Square Pr> ChiSq Intercept 2 1-2.9119 0.4847 36.0875 <.0001 Intercept 1 1-1.4323 0.4618 9.6213 0.0019 feduc1 0.0593 0.8077 3.8229 0.0506 0.8456 111 0.0807 0.3316 feduc1 12 1 0.6796 0.3476 meduc 11 1-0.0734 0.3771 meducl 12 1 -0.0227 0.3863 country Born2 readwrite 1 Income 0.0379 0.0035 0.9531 10.3629 0.0013 1 1.6156 0.5019 1 1.3890 0.5022 7.6497 0.0057 1 0.6291 0.1497 17.6500 <.0001 Odds Ratio Estimates 95% Wald Effect feduc1 11 vs 8 feducl 12 vs 8 meducl 11 vs 8 meduc1 12 vs 8 country Born 2 vs 1 readwrite 1 vs 2 Income Point Estimate Confidence Limits 1.084 0.566 2.077 1.973 0.998 3.899 0.929 0.444 0.978 0.458 2.084 5.031 1.881 13.454 4.011 1.499 10.733 2.516 1.876 1.399 Association of Predicted Probabilities and Observed Responses Percent Concordant 68.6Somers' D0.383 Percent Discordant 30.3Gamma 0.387 Percent Tied Pairs 1.1 Tau-a 0.256 25176c 0.692 812 9