Organise your thesis according to a hierachy of information

You should treat your final thesis as a research article that may be useful and of interest to other students, researchers and even businesses. But different readers may have different interests in how much time they want to invest in reading your thesis. Therefor you should organise your article according to a hierarchy of information:

The title

The title is the part of your thesis that most people will read: Make it good. When people are searching for literature on a topic, the title will be the first, and perhaps only part of your thesis that they will see. A title should therefor be descriptive and specific and tell the reader what your thesis is about, without being overly wordy. You can be playful and clever, but not at the expense of how informative your title is. A bad made-up example: Tossing out the baby with the bathwater?. A better example: Tossing out the baby with the bathwater? How firms can improve layoff decisions using tree-based machine learning. In two short sentences, we have managed to convey the general subject, the research question, methodology and even some of the results. Sometimes a good title simply states the main result of the thesis. A real life example: Why High-Order Polynomials Should Not Be Used in Regression Discontinuity Designs (Gelman and Imbens 2017). Another: Why You Should Never Use the Hodrick-Prescott Filter (Hamilton 2018).

The abstract.

After the title, the most read part of your thesis will be the abstract. This should be about a paragraph long. It succinctly tells what the research question is and motivates the question. It tells about the methods used. It provides the main results and it states why the results are important.

Examples of good abstracts:

The $800 Billion Paycheck Protection Program: Where Did the Money Go and Why Did It Go There?

David Autor, David Cho, Leland D. Crane, Mita Goldar, Byron Lutz, Joshua Montes, William B. Peterman, David Ratner, Daniel Villar and Ahu Yildirmaz

The Paycheck Protection Program (PPP) provided small businesses with roughly $800 billion dollars in uncollateralized, low-interest loans during the pandemic, almost all of which will be forgiven. With 94 percent of small businesses ultimately receiving one or more loans, the PPP nearly saturated its market in just two months. We estimate that the program cumulatively preserved between 2 and 3 million job-years of employment over 14 months at a cost of $169K to $258K per job-year retained. These numbers imply that only 23 to 34 percent of PPP dollars went directly to workers who would otherwise have lost jobs; the balance flowed to business owners and shareholders, including creditors and suppliers of PPP-receiving firms. Program incidence was ultimately highly regressive, with about three-quarters of PPP funds accruing to the top quintile of households. PPP's breakneck scale-up, its high cost per job saved, and its regressive incidence have a common origin: PPP was essentially untargeted because the United States lacked the administrative infrastructure to do otherwise. Harnessing modern administrative systems, other high-income countries were able to better target pandemic business aid to firms in financial distress. Building similar capacity in the U.S. would enable improved targeting when the next pandemic or other large-scale economic emergency inevitably arises.

Will Studying Economics Make You Rich? A Regression Discontinuity Analysis of the Returns to College Major

Zachary Bleemer and Aashish Mehta

We investigate the wage return to studying economics by leveraging a policy that prevented students with low introductory grades from declaring a major. Students who barely met the grade point average threshold to major in economics earned $22,000 (46 percent) higher annual early-career wages than they would have with their second-choice majors. Access to the economics major shifts students' preferences toward business/finance careers, and about half of the wage return is explained by economics majors working in higher-paying industries. The causal return to majoring in economics is very similar to observational earnings differences in nationally representative data.

Formative Experiences and the Price of Gasoline

Christopher Severen and Arthur A. van Benthem

Formative experiences shape behavior for decades. We document a striking feature about those who came of driving age during the oil crises of the 1970s⁠—they drive less in the year 2000. The effect is not specific to these cohorts; price variation over time and across states indicates that gasoline price changes between ages 15–18 generally shift later-life travel behavior. Effects are not explained by recessions, income, or costly skill acquisition and are inconsistent with recency bias, mental plasticity, and standard habit-formation models. Instead, they likely reflect formation of preferences for driving or persistent changes in its perceived cost.

The introduction

The introduction is often the most misunderstood part of a master thesis. I would argue that this is the most important part of the thesis. Once a reader has read the title and abstract and decided they are still interested, the introduction will be the next part they will read.

Build-up and motivate your research question.

The reader should get a full idea of what the thesis is about from the introduction. I think about it as a popular-scientific summary of your article. In the introduction you have some background and motivation, and perhaps also a review of some literature that builds up to your research question. Then of course you should state your research question (as a part of a normal paragraph--not pulled out into a blockquote, or italicized or bolded. )

Tell how you will answer the research question.

This is often where many students stop. But this only gives a partial picture of your thesis. You should also tell how you will try to answer the research question. That is, you want to give a good summary of the methodology and motivate the methodology. Here you can also give a summary of the data that is used.

Tell the reader what the results are. Really.

Many students think that writing a thesis is like writing a mystery novel--that you should wait until the end before revealing the results. Nope. In addition to succinctly stating your main result(s) in the abstract, you should again state your results, now in more detail, in your introduction, perhaps also discussing why the results are important. Repitition is fine. You can (and should) state your results in the abstract, in the introduction, in your results section, and in your conclusion.

What else you can have in your introduction.

Other elements can also be included in your introduction, depending on preferences and the lay-out of your thesis. Many like to include an explicit discussion of their contributions. Something like, "We contribute to the literature on machine learning in bankruptcy prediction, showing the importance of macroeconomic factors." Many also include a preview of the organisation of the rest of the thesis. If you do this, it should be more informative than, "section 2 is the methodology, section 3 is theory".

The best way of learning to write a good introduction.

Don't read other master thesis to inform the way you write your introduction. Even those that have received a high grade often have poor introductions. Instead, read introductions in top academic journals.

You can find some interesting and high-quality articles in, for example American Economic Journal: Applied Economics. Here is the full article about the wages of students studying economics, which has a nice introduction.

The body of the thesis

For the body of the thesis--what I will define as everything between the introduction and conclusion--some students get the impression that you must have an organisation consisting of sections for methods, theory, and results. This organisation can be fine, but you have a good deal of choice of how you choose to organize your thesis. An organisation consisting of elements specific to your thesis can be more effective. Instead of a section titled "Methodology", you could have "A tree-based boosting model of bankruptcy." Instead of "Theory", you could have "Mechanisms for macroeconomics affecting bankruptcy." In general, a thesis does not need a methodology and theory sections, but it should include discussion of methodology and theory.

A few tips on the body of the thesis:

• Remember the "red thread." If it is hard to understand what you are writing relates to the central argument or the "red thread" then you should question whether it should be included in the thesis.
• Be specific. Avoid generalisations in your thesis. In the description of your methodology, you should not have extensive, general description of broadly used methodologies like linear regression. You do not need to describe the difference between quantitative and qualitative research. You do not need two pages on the usefullness of data analysis for business. In general, if you can find the same information in a text book or other source, you should refer to that source. Instead you should carefully describe how and why you use a methodology for your question. What is your model? If you are doing a regression, write out your regression in mathematical format with variable names.
• Use directly relevant literature. When discussing theory and literature, find and discuss the most directly relevant literature related to your thesis. If your thesis is on the macroeconomic effects on bankruptcy in the Nordic countries, find academic articles on the macroeconomic effects on Bankruptcy in the Nordic countries, and perhaps from other countries. You do not need 10 pages of literature on Bankruptcy in general.
• Avoid copy-and-paste. In a masters thesis you should aim for high quality figures and tables. This means making your own figures and tables in latex, word, or whatever software you are using. Copying and pasting tables, figures, and other components often looks ugly, and puts into the question the originality of your work. If you do use a figure or table from another source, make sure you have cited the source clearly - including in the figure/table caption.
• Longer is not better. Most high quality academic papers have a main text (Introduction through Conclusion) that is between 20-40 pages. There is usually no good reason why a master thesis needs to be 70, 80 or 100 pages. If you have many pages, look for parts of your thesis that can be rewritten more concisely. There may be elements of your thesis that you can put into an appendix

How to conclude.

A conclusion serves a few roles. It will include a few of the same elements that you have in your Introduction: A statement of your research question, and the answer/results that you have found. The conclusion is often also a place to discuss some of the potential limitations and scope of the study. For example, is the study of bankruptcy relevant to all firms or just the small firms in your study? But don't get too carried away with talking about the weaknesses of the study and the things you should have done. A reader may ask why you didn't do those things if they were important. Finally, the conclusion is a place where you can allow yourself to speculate a bit. What are the potential implications of your study? What are some avenues for future research?

Style and usage

In the semester before your master thesis, you should read Economical Writing by Donald McCloskey

You should also look at other writing and style guides. Some that I like include:

• The classic Strunk and White Elements of Style
• The Economist Style Guide
• The Chicago Manual of Style

My advice for all students, but especially for those where English is not the native language is as follows:

• Keep sentences simple and short. See if you can split compound sentences into separate, simpler sentences.
• Get rid of unnecessary words – especially adjectives, adverbs, superlatives and filler words. Some made-up examples:
• I find a VERY large effect -> I find a large effect.
• Of course, the findings in my opinion are differentiated extensively from earlier papers -> The findings are different from other papers.
• The scientific statistical model is completely robust to various assumptions -> The model is robust to assumptions.
• Try to write sentences with an active voice: Subject-Verb-Object:
  • The thesis has been completed using statistical techniques -> I used statistical techniques in this research.