If you are a prospective Bachelor or Master's student at Goethe University thinking to write a thesis with my supervision, please consult this page prior to reaching out .
I am happy to supervise both bachelor's and master's theses. To ensure a productive and rewarding experience for both of us, I have outlined the following guidelines and expectations:
1. Allocation Process
Bachelor's Students: Allocation occurs annually through the registration process for students studying under the Chair of Michael Kosfeld.
Master's Students: I am open to supervising students who have taken one of my courses (e.g., MSc Empirical Political Economy or BSc Econometric Analysis of Microdata), conditional on my having sufficient capacity to do so at any given moment. Students who have not taken one of my courses should first look elsewhere.
2. Pre-Registration Phase
The pre-registration phase is crucial for establishing the foundation of your thesis. The primary aims during this period are to agree on the following:
Thesis Type: Decide whether the thesis will be empirical or literature-based. For master’s theses, an empirical analysis is expected.
Research Question: Identify and define the central research question.
Thesis Outline: Agree on a general structure for the thesis.
Data and Methods: Discuss any relevant data sources and establish an overall empirical methods.
**Important Notes**:
Pre-registration supervision concludes once the points listed above are finalized. I will not review results, data, or code prior to registration, unless there are exceptional circumstances (i.e., problem with the data).
Typically, one meeting and a follow-up email are sufficient for this phase. After our first meeting, I expect students to prepare a brief thesis proposal (about one page), outlining the sections of the thesis and the key points for each. Once we agree on this proposal, you register your thesis and start.
3. Data Preparation
For empirical work, students are expected to confirm the availability of all necessary data during the pre-registration phase.
If additional data requires an application process, this must be clarified and addressed prior to registration.
After registration, I cannot assist with applications for unanticipated data sources.
4. Student Preparation and Ownership
I am happy to help students identify research questions, data sources, and to hash out empirical methods. However, I expect students to approach the first meeting with a degree of preparation.
Please come with some initial ideas about what you want to study and how you intend to approach it.
Please do not apply if you are solely seeking ideas or data from me; my supervision is most effective when students take ownership of their thesis.
5. Post-Registration Supervision
Once your thesis is registered, I am available to assist with clarifications.
Generally, one or two emails are sufficient to address questions that may arise. But of course, if there are real issues, a meeting can also be arranged.
Regardless of the form of the meeting, please consolidate your questions into a single, context-rich email, or come prepared with your questions in context of the project. Fragmented questions or multiple emails with isolated queries in which you are "thinking as you go" are discouraged and may not receive a response.
In other words, if you have a question, wait a little to see (a) if you can solve the issue or (b) to see if other questions arise which you can ask all at once, either in person or with an email.
6. Expectations
For empirical work:
Sharp causal identification is not a requirement. While studies that push more in a causal direction (or exploit natural experiments) will naturally be graded higher, creativity and clear thinking can yield very good results even when using standard OLS regressions.
Originality, rigour, and relevance of the research question, as well as thoughtful interpretation of results, are key factors in your evaluation.
Your grade is not contingent on obtaining statistically significant results, though papers that demonstrate a clear pattern (significant or not) will do better. The worst will be a paper that simply shows a range of results without tying the results into a coherent narrative that relates to the overall research question.
If you are producing your own tables and figures, please consult papers from leading journals (American Economic Review, QJE, AEJ, ReStat etc.) to get a sense of how to present results.
For literature based reivews:
In these papers, I also expect students to delve into empirical methods. A literature based thesis should really engage with the science first and demonstrate that you are familiar with the methods and approached used in modern applied econometric studies.
You can use the introduction and conclusion to give your own thoughts and motivation as to why the question is important for policy (or in general). But I expect these opinions to be well grounded in evidence, which you discuss in your thesis.
For both empirical and literature based:
Ensure sections and subsections are clear, easy to follow and informative (i.e., try to avoid, as much as possible, generic section titles)
7. Format and Structure
The written thesis should follow a clear and logical structure. Below are the key elements your thesis should cover:
Introduction: The introduction should explain what you do and why you think it is important. This means opening with the motivation for the study — what gap in the literature or real-world problem prompted the research — before stating the central research question clearly and concisely. The introduction should also include a brief summary of your main results, giving the reader a sense of what the thesis finds before they work through the analysis.
Literature Review: The literature review is the place to contextualize your paper within the existing literature. Be generous in your citations, though it is not necessary to cite every paper in the field. The main purpose of the literature review is to place your work in the context of the three or four closest current papers, and to give proper credit to those whose work directly informs yours.
Background: Provide the relevant context needed to understand the study. This includes any important institutional details, a description of the research setting, and — where applicable — an account of the intervention or policy being examined.
Data: Describe the dataset(s) used in the analysis. This should cover the nature of the data (e.g., panel or cross-sectional), the unit of observation, the source, and any relevant features or limitations of the data that bear on the analysis. You may also wish to present some basic summary statistics in this section.
Empirical Methodology: Walk the reader through the regression framework underpinning the analysis, including the key equations where relevant. This section should make clear how the empirical strategy relates to the research question and what identifying assumptions are being made.
Results: Present the core findings of your empirical analysis, drawing on your figures and tables. Do not simply display output — narrate the results and guide the reader through what the evidence shows and what it means in the context of your research question. Note: all tables and figures should be replicable and should be "stand-alone", meaning a reader should be able to look only at the table or figure (without the accompanying text) and understand exactly what is being presented. Table and figure notes should be used accordingly.
Conclusion: Close by distilling the main takeaway of the thesis. What is the bottom line, and what does it imply for policy or for future research?
8. Final Review and Grading
I will not review your results or thesis prior to submission.
Your grade will be based on the rigour of your analysis, the quality of your research question, and your ability to motivate, interpret, and contextualize your work.
I look forward to working with motivated and prepared students. If you are enthusiastic about your topic and willing to engage deeply with the process, I am here to guide you through the journey.