Would it be useful to know that kids that watch more TV are more disobedient in general? Let's say we found this outcome. Now this may be result of not surveying all groups proportionately. For example, maybe boys are less obedient overall so we wouldn't want more boys in our sample to sway our results. Or maybe kids from one ethnic group are more disobedient than other groups; we wouldn't want more of that group disproportionately. I read some info on the school and makeup of students. Here is the link: https://www.usnews.com/best-colleges/university-of-houston-downtown-12826Links to an external site.

There are vast differences in student population in terms of gender and ethnicity. Should we account for this in our sample? Remember a sample is used because we can't get every student to participate so our sample has to represent the entire population; we can do this with only 30.