,
39 tweets,
6 min read
Read on Twitter
At a #UMich Rackham Grad School training on best practices for graduate admissions, focused on diversity & excellence in grad admissions.
Sitting next to @twittkopp. @sbarolo decided to sit at another table. (I guess that's where the cool kids are?)
One thing that is being highlighted are the criteria for the Rackham Merit Fellowships. Proud to have two Rackham Merit Fellows in my lab! 
Social schemas (stereotypes) influence how we evaluate applicants. Sometimes we are aware of these, sometimes not.
Workshop is giving overview of CV studies showing preference for male applicants (w/identical CVs) & based on names (Jamal vs. Greg).
It can be hard to review a big pool of applications! We need to be aware of biases that can creep in (including based on stereotypes).
We need to make sure we don't rely too heavily on poor indicators of talent (e.g., undergraduate institution).
We need to think about biases that are associated with letters of recommendation (e.g., letters for men tend to be longer).
GREs are NOT predictive of success in grad school. Main thing they can predict: grades in first year in program (but UG GPA predicts better)
ETS itself (which administers GRE!) does not recommend using GRE as cut-off, says should only use as one factor.
GRE scores DO predict socioeconomic background and education of parents.
If we shouldn't use GRE, what should we use? Multi-faceted approach that considers **multiple criteria**.
We tend to focus on things like GRE & GPA, but we should also consider things like perseverance, motivation.
Now we're going to take some time to work in small groups to think about what we can do to better identify potential excellence.
Here's the @umpibs document with more on why they dropped GRE: medicine.umich.edu/medschool/site… (Thanks, Scott!)
I shared what I love to see: a student who has solid grades, research experience AND had some other major obligation (job, sport, music,etc)
Some of the alternates proposed seemed problematic to me (e.g., internship at the Fed). Fortunately someone else called this out!
Thing to do: look for promise. Use holistic evaluation practices. Best practices: 1. have multiple, diverse reviewers for each applicant.
2. Evaluate individually, independently. Evaluate first, then discuss collectively. (Do this to avoid groupthink.)
3. Have a clear, comprehensive rubric for file review. 4. Rethink "achievement" to include extra/curricular activities.
5. Level the playing field by offering phone, Skype, or even asynchronous video interviews.
6. Be self aware. Be careful about what you prize in experiences ("interesting" can require $$$!)
7. Don't rely on pedigree! Black and Latino students' enrollments in selective institutions are declining.
There are lots of great students at state schools, small colleges, regional schools. Make sure your biases don't lead you to overlook them.
How to diversify your applicant pool? Leverage your current PhD students! Have faculty visit institutions that you'd like to recruit from.
Do a preview weekend, possibly incl a workshop where you help students with application materials. (My addition: refer to #EEBMentorMatch!)
Turns out #UMich has a Minority Serving Institution coordinator (Edmund Graham). I should find out more about that!
Ask your current grad students for ideas of how to recruit students. Make sure your website is top notch, looks modern.
One way to recruit students: follow up with them after their visit!
.@sbarolo & I both noted that keeping your current students happy is a really great way to recruit future students. :)
FUND THE CAMPUS VISIT! (And, if at all possible, pay for everything up front!)
Ethnic minority students indicated they accepted offers from programs that funded their visits AND signaled that they could succeed there.
Contact the referees/advisors of applicants to thank them for their letters of recommendation.
Think about what your website signals about your department's culture, climate, and priorities.
.@sbarolo is asking if others have tips on how to extend deadline for students in Puerto Rico & the Virgin Islands. & now workshop is ending
One impt point raised in disc: If you get rid of/downweight GRE, make sure you don't replace it with something it strongly correlates with.
To me, some of the alternatives participants suggested (e.g., think tank internships) were probably just as problematic (if not more).
