In the code demo video we make a brief comparison to UC Davis, saying that a rate of 1 in 10,000 famous people would correspond to 3 famous students at any time at a 30,000 student university like Davis. However, the Wikipedia category UC Davis alumni shows about 650 entries covering its whole history which, over about 110 years and ballpark 300,000 students since its founding, comes closer to 1–2 per 1000, or maybe 30 famous students at any time. Why would alumni of a university show a higher rate of being famous? Generate at least two hypotheses for why, predictions relating to each hypothesis, and datasets that make each prediction testable. Then relate the content of a peer’s post to yours.
-
Restate phenomenon
-
Generate competing hypotheses
- Prediction about effects in world
- Propose a dataset
-
Respond with how data might test or inform their prediction, or vice versa
- Alumni of university are more driven and are biased to becoming famous
- A dataset of random sampling of the population and university students,
- Being at a university with other motivated students creates larger networks that propel them towards success and fame
- Find a dataset of
- Universities provide resources that can help them to become famous
- Measure resources at a university (maybe total spending on student success per student) vs rate of
- University students are often affluent enough to afford college
The phenomenon we are interested in is our observation that alumni of universities show a higher rate of being famous.
A possible hypothesis for this effect could be that through a selection bias, alumni of a university are more driven and are biased to becoming famous. A possible cause for this could be the admissions process of universities that seeks to maximize the quality of students attending the university. We could test this by having a dataset that matches the selectivity of universities with the relative rate of fame at the university. If our hypothesis were true, we would expect to see more selective universities have higher rates of fame than universities with higher admission rates.
Another possible hypothesis for this occurring could be the larger networks that of motivated students that college allows students to build. Having a larger network could allow students to become more aware of opportunities and propel them towards success and fame. We could test this by creating a dataset that compares the social networks of university alumnis/students with people who did not attend university. If our hypothesis was correct, we would expect to see university students having larger networks than those who didn’t attend university.