Felt depressed – control v
Research question: What kind of factors have an effect on whether one feels depressed?
weight by pspwght.
fre fltdpr.
RECODE fltdpr (1 2=0)(3 4=1) into fltdpr_2cat.
VARIABLE LABELS fltdpr_2cat ‘Felt depressed or not last week? 2cat=rather felt depressed’.
VALUE LABELS fltdpr_2cat 0’rather not felt depressed’ 1’rather felt depressed’.
fre fltdpr fltdpr_2cat.
fre pdwrk.
fre rlgblg.
RECODE rlgblg (1=1)(2=0) into rlgblg_2cat.
VARIABLE LABELS rlgblg_2cat ‘Belonging or not to particular religion or denomination?=yes’.
VALUE LABELS rlgblg_2cat 1’yes’ 0’no’.
fre rlgblg rlgblg_2cat.
fre agea.
LOGISTIC REGRESSION fltdpr_2cat WITH pdwrk.
LOGISTIC REGRESSION fltdpr_2cat WITH pdwrk rlgblg_2cat.
LOGISTIC REGRESSION fltdpr_2cat WITH pdwrk rlgblg_2cat agea.
Variables: pdwrk: Doing last 7 days: paid work: 0-no,1-yes
Step 1
b0: The log odds of feeling depressed among those who did not do any paid work in the last 7 days is -1,340.
Exp(b0): The odds of feeling depressed among those who did not do any paid work in the last 7 days are 0,262.
b1: The log odds of feeling depressed among those who did paid work in the last 7 days is by 1,088 lower than among those who did not.
Exp(b1): The odds of feeling depressed among those who did paid work in the last 7 days are 0,337 times as low as among those who did not.
b0+b1: The log odds of feeling depressed among those who did paid work in the last 7 days is (-1.340+(-1,088)).
Exp(b0)*Exp(b1): The odds of feeling depressed among those who did paid work in the last 7 days are (0,262*0,337).
Conclusion: Having a paid job has a significant negative effect on feeling depressed.
Step 2
b0: The log odds of feeling depressed among those who do not have paid work and do not belong to any particular religion or denomination is -1.476.
Exp(b0): The odds of feeling depressed among those who do not have paid work and do not belong to any particular religion or denomination are 0,228.
b1: The log odds of feeling depressed among those who have a paid job everything else hold constant is by 1,066 lower than among those who have paid work.
Exp(b1): The odds of feeling depressed among those who have a paid job everything else held constant are 0,344 times lower than among those who have paid work.
b2: The log odds of feeling depressed among those who belong to a particular religion everything else held constant is by 0,265 higher among those who do not.
Exp(b2): The odds of feeling depressed among those who belong to a particular religion everything else held constant is 1,303 times as high as among those who do not.
Conclusion: Having a paid job has a significant negative effect on feeling depressed.
Step 3
b0: The log odds of feeling depressed among those who do not have paid work, do not belong to any particular religion, and are 0 years old is -2,860.
Exp(b0): The odds of feeling depressed among those who do not have paid work, do not belong to any particular religion, and are 0 years old are 0,057.
b1: The log odds of feeling depressed among those who had paid work in the past 7 days everything else held constant are by 0,687 lower than among those who did not.
Exp(b1): The odds of feeling depressed among those who had a paid work in the past 7 days everything else held constant are 0,503 times as low as among those who did not.
b2: The log odds of feeling depressed among those who belong to a particular religion everything else held constant is by 0,016 higher among those who do not.
Exp(b2): The odds of feeling depressed among those who belong to a particular religion everything else held constant are 1,016 times as high as among those who do not.
b3: The log odds of feeling depressed for the respondent increases by 0,026 every life year, everything else held constant.
Conclusion: At the beginning the b1 was significant and it remained significant even after controlling for religion and age of the respondent. It is interesting to see that in step 2 the effect of belonging to a religion was just out of the significance tolerance level (p=0,066) but in step 3 the effect was eliminated by age (p=0,916). It may indicate that it is not the religion but the person’s age has an effect on feeling depressed.