Open LisaGee opened 9 years ago
For the predominant two-/four-year school comparisons: Associate and bachelor’s degree have very different costs and payoffs, and in general, it’s not quite fair to expect them to live up to the same standard. Students may also have different goals for each program. For instance, a community college will essentially always be cheaper than even a public four-year, and I doubt there’s any community college where the median debt of graduating students approaches that of a public four-year, given that it’s meant to take half the time. Moreover, we aren’t actually using the same completion rates for two- and four-year schools (200% for two-year, 150% for four-year), so comparing an average of all schools would be confusing at best. Bottom line, benchmarking requires defining a peer set of institutions, and this is a broad enough grouping to help students recognize the choice they have, while being fair to the mission of the schools and expectations of their students.
Moving @saintsoup52 https://github.com/saintsoup52 comment from another ticket #493, so it doesn't get lost.
Of all the emails I read over the weekend… I think this one has the most policy implications… because this is definitely a new yard stick schools are being measured against. And not that you asked… but I am going to offer my opinion on this anyways…
1) Median Earnings (use - PREDDEG): I think it is probably unfair to compare the earning capacity of a completer of an associates degree against a bachelor’s degree here. Especially when this will just strongly discourage against going to community colleges.
2) Threshold Earnings (use - All schools): Since this a more universal measure, I think it is ok to compare all schools against a national average.
3) Completion/Graduation Rates (use - PREDDEG): Tool should have separate national averages by the predominate degree awarded (same as scorecard)
4) Net Price (use - PREDDEG): Should definitely NOT be separated by public/private. That will cause all sorts of data anomalies and mask expensive private schools. The scorecard currently separates by predominate degree and I think it that it should stay that way.
5) Monthly Payments (use - PREDDEG): Comparing the monthly payment for a completer of a 4 year school against a completer of a 2 year school doesn’t make any sense. Of course the 2 year schools will look better here. They should continue to be done by predominate degree.
6) Repayment Rates (use - All schools): On the current scorecard default rates are compared against the national average for all schools. This makes sense to retain a comparison against all schools here.
7) Retention Rate (use - All schools): This is a universal metric… it could be all schools or predominate degree type.
@jjoteal - Can you figure out how you would display these distinctions in the UI?
@LisaGee -- As a simple solution for launch (though not sure how much development time this would take), how about a simple toggle interface for 2-year and 4-year? All 2-year schools would display in toggle view A and all 4-year schools would display in toggle view B?
I know this would require some testing... is the goal to launch this in the MVP?
Based on recent usability testing, I would recommend implementing this change in conjunction with introducing the filter left column back into the design. I also recommend adding a top "Edit Search" area (similar to Kayak). The 2-year and 4-year "toggle" options would appear in the "Edit Search" area. Additionally, if a user is viewing both 2-year and 4-year together in the search results, I recommend adding some messaging at the top of the page along the lines of "You are looking at both 2-year and 4-year schools. For a more accurate comparison, you may want to view [4-year schools only] or [2-year schools only]" that allows the user to quickly toggle. Can do a mock-up of this once this issue is prioritized in a sprint.
@nguyenist I think this is yours now to actually mock up. It's actually sneaking up pretty high into the priority list and we will be ready to have development start on it pretty soon!
This applies to:
This will require a bunch of trial and error, designs and usability testing to figure out how to achieve it well.
???What is the exact goal? Asking Clare.