Notes for future years from the 2010/2011 season
  • Website issues:
    • Fix the transcript problem – change the language so that students enter all their grades in chronological order.
    • Fix the issue of what school shows up
    • Add more terms to search process – make it easier to find unreviewed/uncompleted folders. Also, how do we search for numbers of reviews, number of completed reviews, etc.?
    • Can we merge in recommender statistics from EECS previous years?
    • Remember to get statistics from USAF in advance of starting review process.
    • The research areas are hard to decrypt. What is ACE?
    • Can we allow users to hide the scores on the folder list page? (They scores are missing from the individual folder pages, but the list shows them. Can we get just: review completed, rather than scored.)
  • Evaluation issues:
    • Need to grow the GAC, make sure we get coverage in all areas. 6 and 2/2 GAC members is way too few.
    • Determine better TOEFL cut-off process. Also compare success rates vs. TOEFL scores of matriculated grad students.
    • Should we have a GRE cut-off like the TOEFL? No: stick to GRE recommendations, not cut-offs. Note for next year: the recommended GRE scores were not known by GAC. We should make sure we inform the faculty/GAC of these.
    • Revisit the issue of the 3rd reviewer. Folders are getting a lot of reviewing now. If a folder gets 3 GAC reviews and is AD, then it will get at least 2 from the sector, for a minimum of 5 reviews. Maybe in the future if the reviews conflict, it should go the sector to resolve the conflict. Dependent on size of GAC. Right answer is probably to stick to 3 reviews on conflicts, but have more people.
    • Visit the issue of applicants who failed the quals twice. Are they allowed to apply, and how do we evaluate them fairly?
    • We evaluate SM applicants differently from PhD applicants. Should we evaluate SM applicants who clearly only want an SM (e.g., USAF) differently from applicants who might go on to get a PhD? If not, we should make this explicit to reviewers. If so, we should enforce the difference at quals time.
    • Can a faculty member force a fast-track even with low GAC scores? Suppose the low GAC scores were given by reviewers outside the area? Fast-track is a competitive advantage. If the candidate is admissible, and someone wants the fast-track, what's the argument against?
  • Others
    • Get Stanford open house date early, or set ours early, or negotiate with Stanford early. It turns out that our entire schedule can be affected by this.
    • Change aa-gac mailing list to be a mailman list with [AA-GAC] prepended to subject and also Wiki and applyweb URLs in footer.
    • Consider changing GRE score deadline
    • Consider adding an explicit deadline for recommendation letters, consider sending recommendation letter reminders.
    • Consider adding automatic generation of recommendation letter reminders.
    • Need to fix the process from the grad admissions office, some application materials come from them late/not at all.
    • Dave recommends not waiting transcript before allowing a folder to be reviewed. Beth has concerns about this.
    • EECS system may become the institute-wide system, and we should stay aware of changes that may result.
    • Need to formalize fast-track policy, in particular find a way to ensure that all relevant faculty have been given the opportunity to review a folder.
    • Need to think about how better to manage multiple RA offers in relation to sizing process. There may not be an answer here.
    • Is there a way to get lists of Fulbright fellows ahead of time, along with types of fellowships? Also, what is an IS&T alternate? Is it like being waitlisted? How often do these get upgraded to principal?
    • Need a policy for GAC members that don't own laptops.
    • Need to formalize admissions policy in relation to qualifying exam with grad committee
    • How close is the tie between admitted student and initial advisor? Admissions letter says "switch to another advisor" rather than "choose an advisor".
  • No labels