That mostly younger generation uses Stack Overflow and thus age-over-languages cannot be extracted from the data analysis. We are just learning statistics and probability and i think this happens if the sample data is incorrect or not diverse enough.