Background info:

I am doing a correlational study design and my hypothesis is that there is a correlational relationship between marital status and health, with people who are married scoring substantially healthier than those never married, divorced, separated, or widowed.

My independent variable is marital status.

My dependent variable is health.

I gathered data from GSS. I coded data as follows: marital status 5=married, 4=widowed, 3=divorced, 2= separated, 1= never married. Health is coded as 4=excellent, 3= good, 2= fair, 1= poor.

Sample size is 18,142 participants

I have SPSS.

*** I already did a correlation analysis and tested for equality of variances (homoscedasticity)

Correlation analysis shows significance, p-value is approx 0

Test for homogeneity of variances indicates that the variances are not the same.

Tests of Homogeneity of Variances Levene Statistic dfl diz Sig. Condition of health Based on Mean 57.675 4 18120 .000 Based on Median 19.800 4 18120 .000 Based on Median and 19.800 4 17863.256 .000 with adjusted df Based on trimmed 39.555 18120 .000 mean