Analysis
I used statsmodels in python to perform linear regression on respondents' age, household income and 'Disscore' to explore how well does discrimination relate to age and financial status. I also set my null hypothesis that the ratio of female experienced discrimination is the same or lower than male in this survey, I then performed z tests at significance level p=0.05 to test it.
Results
In the descriptive analysis, among 3545 valid respondents, 72% of them think they perceived discrimination in life, and most people perceived discrimination think the cause is their age, and the second and third causation is gender and financial status respectively. In the linear regression model, the relationship between age and perceived discrimination or household income and perceived discrimination is not significant indicated by the R2(Fig3-4).