CSD PhD Admissions: A Few Proposed Changes

1
CSD PhD Admissions A Few Proposed Changes

• Todd C. Mowry
• CSD Faculty Meeting
• Sept. 20, 2001

2
3 Steps in Getting Excellent PhD Students into
Our Program

• 1. Get the right people to apply
• 2. Admit the right people
• 3. Recruit those people to come here
• Steps 1 3 are where the game is won or lost
• there is room for improvement here!
• But today I will be focusing on Step 2

3
What Does the CSD PhD Admissions Committee Do?

• Phase 1 filter out obvious rejects
• remaining folders are placed in bins for faculty
  review
• huge public service!
• Phase 2 produce a ranking list (A1, A2, B1, B2,
  etc.)
• based heavily on faculty input
• Resulting list helps Randy make the actual
  admissions decisions

4
The Good News

• The quality of the output from the committee is
  quite high
• I felt very good about our list last year
• (Thanks, Steven!)
• Improving the quality further is not a major
  concern

5
The Bad News

• Time spent on admissions over ones career

6
Goals of My Proposed Changes

• Make serving on the admissions committee less
  onerous
• Give people who are not on the admissions
  committee more of a stake in the process

7
Why Should We Make the Admissions Committee Less
Onerous?

• We can do so without sacrificing quality
• We want people to be eager to serve on the
  admissions committee
• We should conserve our energy for recruiting
• the best recruiters are people who were involved
  with admissions!
• but these people are exhausted under our current
  system

8
Why Should People Who Are Not on the Committee be
More Involved?

• They can provide valuable input
• Give them more confidence in the system
• Get them fired up for recruiting!

9
Admissions Committee vs. Program Committees

• Program committees are also important
• In contrast with admissions, most of us serve on
  program committees on a regular basis
• There are some lessons that we can learn from
  well-run program committees

10
How Onerous is Our Current Admissions Committee
System?

• 6 committee members (3 faculty, 3 students)
• Phase 1 narrow down 600 folders to 150 folders
• each folder is read by one group of 2 people
• each person reads 200 folders over 3 weeks
• get together and discuss each case with partner
• Phase 2 read all folders that pass phase 1
• another 100 folders over 1.5 weeks
• discussed over a 3-day long meeting

11
How Can We Reduce the Workload?

• Make the committee larger!
• e.g., increase it from 6 to 12 people
• Make workload inversely proportional to committee
  size
• Phase 1 no changes needed
• Phase 2 each folder read by 6 people (not the
  full committee) more similar to a program
  committee
• Cutting the workload in half would make a huge
  difference in how much this impacts ones life
• more in line with serving on a program committee

12
Getting Other People More Involved

• Areas are encouraged to do the following during
  faculty review week
• Meet once or twice as a group (e.g., over lunch)
  to review folders
• can discuss them with your colleagues!
• Near the end of the week, meet to do a dry run
  of the ranking (A1, A2, B1, B2), including
  rationale
• representative from admissions committee will
  attend and present that info during admissions
  meeting
• this ranking will be taken very seriously!

13
Summary 3 Steps Revisited

• 1. Get the right people to apply
• 2. Admit the right people
• 3. Recruit those people to come here
• Admissions should be the easy part.
• Lets find new ways to improve Steps 1 3!