Health and scale variable: number of people
weight by pweight.
fre health.
RECODE health (1 2=0)(3 thru 4=1) into health_2cat.
VARIABLE LABELS health_2cat ‘Subjective general health’.
VALUE LABELS health_2cat 1’fair or bad’ 0’good’.
fre health health_2cat.
fre hhmmb.
LOGISTIC REGRESSION health_2cat WITH hhmmb.
b0: The logarithm of the odds for feeling bad or fair when the independent variable is 0 is 0.659. (Note: We cannot interpret the 0 value of the independent variable, just as in the previous example.)
b1: The logarithm of the odds of feeling bad or fair is lower by 0.411 for each additional person living in the same household. This is the difference in the logarithm of the odds between those who live alone and those who live with 1 person and also between living with 1 or 2 persons and also between living with 2 or 3 persons, etc.
Exp(b1): The odds of feeling bad or fairly good is multiplicated by 0,663 for each additional person living in the same household.
Exp(b0): The odds of feeling bad or fairly good when the independent variable is 0. Note: we cannot interpret this.
Questions:
- What are the odds of feeling bad or fairly good among those who live alone in the household? (So, there is only 1 person living in the household)
Exp(b1)*Exp(b0)=-0,663*1,932=-1,28 - What are the odds of feeling bad or fairly good among those who live with 1 persons together in the same household?
Exp(b1)*Exp(b1)*Exp(b0)=-0,663 *(-0,663)*1,932 - What are the odds of feeling bad or fairly good among those who live with 2 persons together in the same household? (So, there are 3 people living in the same household.)
Exp(b1)*Exp(b1)*Exp(b1)*Exp(b0)=-0,663 *(-0,663)*(-0,663)*1,932 - What are the odds of feeling bad or fairly good among those who live with 3 persons together in the same household? (So, there are 4 people living in this household.)
Exp(b1)*Exp(b1)*Exp(b1)*Exp(b1)*Exp(b0)=-0,663*(-0,663)*(-0,663)*(-0,663)*1,932