RESPONSE TO THE 2-4 PROJECT
Yale University Department of Physics - December 8, 2006
Jon Butler, Dean of the Graduate School, has asked every department to review and evaluate their PhD program this fall semester, concentrating on years 2 through 4. This is the 2-4 Project.
To oversee the physics department’s review a faculty committee was established comprising Bonnie Fleming, Steve Girvin, Walter Goldberger, Dan McKinsey, Simon Mochrie, and Nick Read. Because we believe it is impossible to sensibly separate years 2 through 4 from year 1 and from later years, we took it upon ourselves to review and discuss all aspects of the program.
A key element of the 2-4 Project review was student input. To this end, both the Graduate School and the Physics Department’s 2-4 Project Committee invited responses to questionnaires asking for students’ opinions concerning various aspects of our graduate program. There were 9 responses to the GS questionaire and 15 to the committee’s, out of a total of about 120 graduate students in the Physics Department. In addition, draft proposals were discussed at the Chairman’s Town Meeting with graduate students on December 6, at which the attendence was about 50 students, and at a meeting of the Teaching Faculty on December 7. The proposed changes described in this document were approved by faculty vote at the December 7 meeting.
At these separate meetings, in essentially all cases, there was broad concensus among the students and among the faculty about how to modify the Physics PhD program. In addition, there was also good agreement between faculty and student opinion.
The questionnaires indicated that more contact time between TA and their students and less grading would be highly welcomed by many respondents. Some of our graduate students may want this and be suitable for this, some may not. However, there are many classes and we encourage a student-faculty discussion of how to accomodate graduate students who want to lecture/tutor, rather than grade. Perhaps, in some classes, there should be TA-led recitations?
One suggestions was the following, on a trial basis with willing faculty and willing TAs: Instead of having TAs grade all (6 or 8) problems on a problem set, they grade only 3 or 4 of the 6 or 8, chosen at random. In the remainder of their TF time, the students lead recitations down the hill the evening before the problem set is due.
Rationale.
For many incoming students, the current core classes are appropriate. For better-prepared incoming students, however, repetition of material that they’ve previously mastered elsewhere delays the taking of more advanced and interesting courses and delays entry into research. It emerged in several of the questionnaire responses that students found some of the material of the core courses repetitive of undergraduate material, leading to considerable dissatisfaction.
Currently, it is possible at the discretion of the DGS for a student to replace a core course with an elective. However, this is a subjective judgement with lots of possibility for error either way. An exam, based on material from the previous years’ course, that permits students to place out would be more objective and fairer. This motivates Proposal 1.
In the questionnaires, a significant fraction of the respondents indicated that some of material in PHYS 500 was repetitive of what they had already studied as undergraduates. This motivates Proposal 2.
Proposal 3 is motivated by the uncertain focus of CFT, and the opinion among student respondents that it was of doubtful value in later research.
Proposal 4 follows from the observation that, on one hand, General Relativity is increasingly important in a number of subfields, while, on the other, currently, the teaching of GR within the department has become distributed among several courses, including Math. Methods, Classical Field Theory, PHYS 538a, and PHYS 680a.
In essence the QFT/MBT requirement is a requirement to have taken a class that involves Feynmann diagrams. Proposal 5 expands the list of courses that are consistent with this, but which may be preferable for some subfields, in particular, for soft-condensed matter physics.
At one level, Proposal 6 addresses the student-expressed concern that there is a mismatch between what is taught in the core courses and what is tested on the Qualifying Exam. Beyond this, however, in view of the increased diversity of Physics and the department, it seems valuable to indeed specify the essentials.
Since physics is based on experiments, it is imperative for all physics PhD students to have some laboratory experience. PHYS 504Lb is appropriate for incoming PhD students without sufficiently advanced undergraduate laboratory experience. Currently, students can be excused from PHYS 504Lb at the discretion of the DGS if they can demonstrate an equivalent previous class. This procedure too is fraught with objectivity and fairness issues. For students who have had prior laboratory experience, Proposal 7 propels them into a real research environment earlier than previously and it gives them credit for it. This is not a “soft option”. The reason for insisting on a proposal and a final presentation is to ensure that the SI is a genuine educational experience, in which the student is invested in the research of the host group. Having a faculty group examine the outcome of the SI addresses one of the perceived weaknesses of the SI/rotation process in other science departments, namely that students who perform poorly in SIs are not given proper feedback, so that they remain unaware of what is expected of them in a research context. This problem was particularly emphasized in the Engelmann report. Such a group review also permits a wider range of potential advisors to get to know the student.
Rationale. An important recommendation of the Engelman report, that the GS is emphasizing, was that students’ research experiences should be formally evaluated and the evaluation results communicated to students in writing. Thus, substantive feedback to a student concerning his/her progress at the end of each year will shortly be required of us by the graduate school. This motivates re-evaluation of the department’s formal procedures for monitoring a student’s progress towards their PhD, and mentoring her/him along the way. Indeed, in the questionnaires, the reaction of most students was that they had never received any feedback concerning their research performance, and that they would like feedback. Students particularly mentioned lack of feedback after talks, and a few bemoaned the absence of instruction in how to give a good talk. Some students, however, who were already making a number of research presentations to collaborators, disdained the idea of having to do an additional, special presentation for the thesis committee.
The Graduate School is especially concerned that research performance and potential be evaluated early on. This motivates Proposal 1.
A student’s thesis prospectus is a very important document, representing an initial vision for the student’s thesis. Currently, the prospectus is prepared by the student, in collaboration with the student’s advisor. It is then submitted to the DGS and, if the DGS approves, it is posted on the Departmental web site for review by the entire faculty. The weaknesses in this procedure are two-fold. First, the DGS is not expert across subfields - certainly not the current one. Second, in most cases, expert faculty do not read the posted prospecti. Thus, the current procedure provides an inadaquate review of the thesis prospectus. By contrast, the proposed procedure provides the opportunity for mentoring and advice from more than solely the student’s advisor. In this context, it is worth noting that Physics seems to be unique at Yale in not establishing a thesis committee essentially until the thesis is finished. These considerations lead to Proposals 2 and 4.
Proposals 3, 5 and 6 are designed to more closely monitor progress and mentor students beyond Year 3, and also provide detailed written feedback to the student about their progress as required by the Graduate School. The allowable format and content for the “public presentations” should be viewed broadly, subject only to the participation of the core thesis committee. Especially early on in her/his research career, to satisfy this requirement, it may be that it makes most sense for a student to make a journal club-type presentation in the context of a group meeting, later on progressing to a research-based presentation in one of the regularly scheduled series. Yearly private meetings with the core thesis committee, in addition to yearly public presentations which the core thesis committee attends will provide ample opportunity for the core thesis committee to determine that a student has a sufficiently broad knowledge of the subfield and to insist on steps to correct any important gaps. These extensive meetings will include any function that the Field Oral plays currently, leading to Proposal 7.
Preparing for presentations is likely to improve students’ presentation skills and the ongoing monitoring of a student’s research progress will diminish the chances of surprises at the thesis defense. Such monitoring will also provide a protection to both the student and advisor that is presently lacking: First, if a student has sufficient material for a PhD, why then the committee can push a reluctant advisor to agree to a thesis defence. (One questionnaire respondent complained that his/her advisor was reluctant to let him/her graduate, even though his/her thesis was really done, because as a senior student, he/she was less expensive and more skilled than a junior replacement would be.) Alternatively, if a student’s research performance is inadequate, the committee can support academic sanctions on the student, i.e. that the student is not in good academic standing.
Finally, the use of a mutual assessment from, currently in use at Stanford, was considered but overwhelmingly rejected by the students.
The majority of students were more-or-less satisfied with the qualifying exam and by a ratio of about 3:1 would keep it as is rather than eliminate it.
Rationale. The student questionnaires overall did NOT reveal any great dissatisfaction with the current Qualifying Exam procedure, except a possible mismatch between material on the exam and that taught in the core courses. This motivates Proposal 1, and our earlier proposal to create guidelines for the topics appearing in the core courses and the qualifying exam. However, students agreed with the concern, expressed by several, that the current administration/grading of the exam was too opaque. On balance, there was no faculty concensus on a satisfactory means to address this concern.
Proposal 2 formalizes an ad-hoc procedure that has been followed occasionally.