Design of the PQE
The PQE is an essay exam constructed to assess the student's capability to design a research proposal based on a research problem scenario presented to them. It also tests the capacity for clear and logical expository writing. The student answers one question chosen from a set of four or five questions based on a policy or management problem. The PQE does not examine specific substantive knowledge from any given course; instead, it relies on the student's general knowledge and the appreciation of the logic of applied social science research.
The PQE is offered once a year in March. Students should sign up for the exam with the Assistant Director of the Doctoral Program. Sample exams are available to assist students in preparing for this exam. In addition, the program runs a PQE study group beginning in January each year.
Grading of the PQE
Each exam is graded independently by at least two faculty members. They are reviewed and graded anonymously. The faculty members then compare results and discuss any discrepancies in the grades. Students are notified of their final grades approximately four weeks after taking the exam. Students who do not pass are encouraged to speak to the relevant faculty to see how they can improve their performance. The specific areas considered in assessing the quality of the exam include: research context, articulation of a conceptual framework/causal model, statement of hypothesis, study design, sampling strategy/method of assignment, quality of measures, data source/data collection, analysis plan and limitations.
Aspects and Criteria for Assessing PQE Performance
|Aspect of The Examinee's Response||Criteria for Rating the Aspect|
|Research Context||Has sense of the policy context for the proposed work (no detailed knowledge expected).|
|Articulation of a Conceptual Framework/Causal Model||Articulates IV, DV, and Covariates. Includes short statement of the theory of the treatment (no detailed knowledge of substance required).|
|Statement of the Hypothesis||Clear statement of questions addressed.|
|Design||Clear statement of type of design to be used; design appropriate to topic. Unit of analysis identified.|
|Sampling Strategy/Method of Assignment||Rationale for choice of that sample (those subjects) as exemplars; realistic sense of where subjects might come from.
A plus: discussion of recruitment or identification issues. Method of assignment to treatment and comparison status discussed (when appropriate).
|Quality of Measures||Brief description of how constructs are operationalized through specific measures. Recognition that operationalization can be problematic.|
|Data Source/Data Collection||Realistic sense of where data might come from. No data ex machina; attention to spatial and temporal issues.|
|Analysis Plan||Plan should be brief and non-technical (no advanced knowledge of statistics expected).|
|Limitations||Measures- validity issues, particularly
Design- threats to internal validity
External Validity Issues - bias
A plus: discussion of how these might be mitigated in future work.
|Logical Organization of Answer||Answer is organized into logical sections and each individual part of response is clearly developed (separated by bolding, effective use of paragraphing, etc.).
Uses topic sentences effectively; employs smooth, logical transitions.
|Quality of Writing||Response written in a clear, grammatical style. Arguments and ideas are logical, well developed and clearly stated.|