Thesis: General remarks

1. Thesis: General remarks • The goal of these meetings? • To get you started! •  To provide ideas of possible research topics. •  To provide some general writing tips. • Why can’t I wait until spring? • A good question should be based on earlier research, and surveying the literature takes time. • Even if you know what you would like to do, it is not 100% sure that you will be able to find a supervisor that is willing to take on your project. •  Even if you think that you have found the best question ever, your supervisor might not agree, and so it might be necessary to revise the question. •  Even if the question is good, you might not be able to find the data, which brings you back at square one again. •  Many students start with a too broad question, and ‘reducing’ it can be time consuming, even if it is done in dialog with the supervisor.

2. Thesis: General remarks • The importance of the thesis • Writing reports is likely to be an important part of your job. • The Master thesis is the single most important document you will write during your studies. • A really good thesis can outweigh any mediocre course grades. • If you would like to apply for the PhD programme, the Master thesis is a good indicator of your suitability. • If you apply for a job in the private sector, the employer will use the thesis as an indicator of your ability to conduct analysis and to communicate the results in an effective manner.

3. Thesis: General remarks ● You may pick any topic (given limitations implied by your track). • Once you have an idea of the area, try to find a suitable supervisor. Together you decide on a more exact research question. • Talk to me if you don’t know who to talk to • It is good to write in pairs (three is not allowed), partly because you are more likely to finish in time, partly because the end result is usually better. • At a later meeting you will present your thesis topics. Everybody must present. This is important because it forces you to think deeper. • Do not be too ambitious -- the thesis project should be doable in less than 20 weeks. • The length should be about 25 pages for a single authored thesis, and about 40 pages for a coauthored thesis. Use 12 points font and 1.5 line space. • Don’t write much more! Quality is much more important than quantity

4. Thesis: General remarks •  All theses must be sent to Urkund for control. •  When the thesis is finished, you should defend it at a seminar. At the seminar, another student will act as opponent, and give his or her judgment. (more details later) • In addition to presenting your own thesis, you must act as opponent on another student's thesis. • You need to attend at least three presentations, excluding you own • Reading for the thesis • Westerlund, J. (2009). The Recipe for a Successful Thesis in Applied Economics. Available via the course web page. • DO READ IT!

5. The Master’s Degree projectImportant dates • Mid-April • Last day to send in abstracts for the Yearbook (word file sent by email) to the GS Student Office. •  Seminars: late May • Late May • Last day to hand in the final grades of the thesis to the GS Student Office. (From supervisors or thesis coordinators) • Late May • Last day to notify GS Student Office who gets the Malmsten Awards. • Late May-Early June • The students hand in three paper copies (double paged, non stapled) to the GS Student Office in the reception desk. • 14 June (Thursday) • Graduation Day

6. Choosing a topic • Many students believe that research is only made up of ground-breaking discoveries like the wheel, the light bulb, the theory of relativity or the cure for cancer. •   This is wrong!  • The majority of research makes only very marginal contributions to our understanding of things, and it is almost never made directly from scratch, but is made based on the work of others. • So contributions are marginal, and you do not even have to work everything out yourself -- easy!

7. Choosing a topic • Important items: • A good question. • A testable hypothesis. • A good data set. • An econometric approach that fits the problem and that can be used to test the hypothesis of interest. • Note, you are allowed to write a theoretical thesis, but not a survey • Base your choice on • Interest, specialization and future work • The type of topic. Is it economics? • Difficulty. Avoid studying very recent events or the future

8. Choosing a topic: The question • Ask, what is at stake. (Does anybody care about the answer) • Survey the literature.   • Reading articles help you see where the ``holes'' are. • Some questions to help you spot potential shortcomings in studies: • Is the motivation of the question convincing? •  Are the data used appropriate for the task? •  Is the econometric approach satisfactory? •  Are the assumptions reasonable? •  Do the conclusions follow from the evidence presented, or is the author overstating the results? •  Another way to start is to try to mimic the study in an article that you like. •  Use the same econometric approach and pose the same question but with another data set. •  Keep the data but replace the econometric approach with one that you think is better suited.

9. Choosing a topic: The question • The best way to search for articles is to use Scholar Google http://scholar.google.se/ • Combine words and dates to narrow down the results of the search. Then follows citations • PDF] HIV/AIDS and fertility • [Fulltext@GUJG Fortson - American Economic Journal: Applied Economics, 2009 - aeaweb.orgThis paper studies the response of fertility ... Citerat av 16 - Relaterade artiklar - Se som HTML-version - Alla 12 versionerna • Also check out authors you know work on interesting topics. Most researchers have papers on their personal WebPages.  • Many papers are hard to read. • The jargon will oftentimes become clearer as you continue reading, thereby putting everything in context. • The math is only for giving rigor to something that can be explained in words. Think of it as a ``proof‘’. But the proof is not necessary if you only want to understand what is being proven. • If econometrics is unfamiliar to you, then it is just to look it up in one of your old econometrics text books.

10. Choosing a topic: The question • The research question should boil down to a testable hypothesis that can be tested. This is crucial! • The data • As soon as you have an idea, start searching for data. One cannot stress this point enough! • Many students are very enthusiastic about their topic and through themselves into the writing process, only to find out that there are no data available. • This is not only time-consuming but also devastating for morale, because being forced to start all over again is never funny. • A good tip when it comes to finding data is to pay attention to the data used in the articles you read. Check their sources • And type of data used. More than 40 observations are needed for a study • Control variables are usually not crucial and can in many cases be easily replaced. • The immediate task is therefore to identify those variables that are necessary for the test, and then to check the availability of the data.

11. Choosing a topic: The question • Some variables are very easy to find but others are harder to come by. For example, while aggregate macroeconomic and financial data are very easy to find, detailed data of household behavior are more rare. • Many students are able to find most of the data they need, but some variables are missing. In such cases you may replace the missing variables by proxies. • There are many databases available, publically (World Bank, etc) in the library, and in the department. • In most case, there is no time to collect your own data

12. Analysis • Empirical approach • Once you have secured some data, the next step is to check them • Make plots of the data. • Do they make sense? Check numbers, outliers, missing observations, breaks, etc. • Check relationships . Try answering you question using graphs and correlations • Then, try to obtain some preliminary results. • Most hypotheses can be cast in terms of a regression model, which makes ordinary least squares (OLS) a natural startup estimator. • Do the results look reasonable, and how do they conform to your hypothesis? • Use do-files and document your data analysis.

13. Format of thesis • The visual impression of a thesis is important. It is a signal of quality! • Structuring the thesis • The purpose of the written thesis is to report the results in a way that makes the reader interested. • Before reading the thesis the reader does not know whether he or she is going to enjoy it. • It your job to make the reader interested! • But readers are busy people. No reader will ever read the whole thesis from start to finish. • You have to make it easy for readers to skim!

14. Format of thesis • The following format should (could) be used: • Title page. • Abstract. • Acknowledgements. • Introduction, • and literature survey. • Theoretical model. • Empirical analysis. • (Discussion) • Conclusions. • References. Not numbered • Appendices. Not numbered • But the format is not cut in stone, depends on topic

15. Format of thesis • Title page • The title page should contain the title, the date, and the names and personal numbers of all the authors. •  The title should be short and informative. Just by looking at the title the reader should get a good idea of what the paper is about. •  But it should not reveal too much. • Example: Compare ``Wage and income effects on the timing and spacing of births in the US'' with ``Fertility dynamics in the US.'‘ • And “Extremes of Body Mass Index Reduce in Vitro Fertilization Pregnancy Rates”, with “Body Mass and Pregnancy Rates”

16. Format of thesis • Abstract • The abstract should be short, about 100--200 words. •  The main function of the abstract is to communicate the contribution of the thesis. •  The abstract must be concrete. Do not write ``I analyzed data on exchange rates and found many interesting results.'' Explain what the results are. • Example Miguel et al. (2004): ``Estimating the impact of economic conditions on the likelihood of civil conflict is difficult because of endogeneity and omitted variable bias. THE PROBLEM • We use rainfall variation as an instrumental variable for economic growth in 41 African countries during 1981--99. METHOD/SOLUTION • Growth is strongly negatively related to civil conflict: a negative growth shock of five percentage points increases the likelihood of conflict by one-half the following year. We attempt to rule out other channels through which rainfall may affect conflict. Surprisingly, the impact of growth shocks on conflict is not significantly different in richer, more democratic, or more ethnically diverse countries.'‘ RESULTS

17. Format of thesis • Abstract • Fortson (2009) “This paper studies the response of fertility to the HIV/AIDS epidemic in sub-Saharan Africa. Standard models of fertility have ambiguous predictions in this context. Because childbirth requires unprotected sex (which increases the risk of contracting HIV), we might expect HIV/AIDS to reduce fertility through an infection avoidance motive. Adding to this decline, HIV infection may itself reduce fecundity among infected women. However, because HIV/AIDS affects the expected longevity of children, a quantity-quality model of childbearing would predict an increase in fertility. I use repeated cross-sections of the Demographic and Health Surveys for twelve countries in sub-Saharan Africa to examine this question empirically. Using individual birth histories from these data, I construct estimates of the regional total fertility rate over time. In a difference-in-differences approach, I compare regional HIV prevalence to changes in total fertility rates from the late 1980s to the present. My results suggest that HIV/AIDS had very little impact on fertility, both overall and in a sample of HIV-negative women. • Problem, data, method, results

18. Format of thesis: Introduction • The introduction should motivate and describe what you do in the thesis. It should be concrete and to the point. • If you do not know how to structure the introduction, take a look at an article you like and adopt a similar approach. • Your readers are busy and impatient, and unless you are able to pursue him or her to continue, most will stop at the introduction. • A good motivation is therefore absolutely crucial! • The motivation should address three issues: • What is the nature of the problem under investigation? • Why is it interesting? • What does your thesis attempt to do? That is, what is the contribution?

19. Format of thesis: Introduction • The next step is to describe in more detail what you do in the thesis. • Structure • Problem in general terms • Quick review of previous studies • What you do and why you do it • Method • Data • Results • Roadmap. • A good rule is to at least write down the empirical results before you begin with Introduction!

20. Format of thesis: Literature review • The literature review has three goals: • It should identify and explain the main findings of the closest studies. • It should point out the main deficiencies of these studies. • It should explain how your thesis contribute by overcoming the deficiencies of the previous studies. • Most theses have a separate review section. • But this is not necessary; oftentimes it more natural to have the review in the introduction, when you motivate the problem. • In any case, do not write overlong reviews! The review is not a contribution of your thesis, save your efforts until the empirical analysis. • Note, the review has to synthesize the literature, i.e., do not just list studies and results

21. Format of thesis: Theoretical analysis • The purpose of the theoretical analysis is mainly to help understand and to put into context the model you take to the data. • In most theses the theoretical model is just an equation. • Your job in the theoretical analysis is to derive this equation and to formulate testable hypotheses. • But this is not always easy, or even possible • The following items should preferably also be included: • An explanation of the direction of the causality. • A discussion of which variables to include. Each variable should have its own little ``story.'‘ • Choice of functional form.

22. Format of thesis: Empirical analysis • The purpose of the empirical analysis is to provide empirical evidence for your research hypothesis. • This is obviously an important part of the thesis! • The empirical analysis should include the following elements: • A description of the data used. • A discussion of the empirical model. •  A discussion of the econometric analysis. •  An explanation of the empirical results and how they compare to the hypothesized ones. •  The empirical analysis is the contribution of your thesis, and should take up most of the space. •  However, it should still be focused. Through out everything that are not central to your thesis!

23. Format of thesis: Empirical analysis • Some data description and preliminary estimation results should always be included. • Tables of means, standard deviations, etc with cross section and microdata and graphs with time series data •  Modelling is not easy, be prepared to spend time of choice of variables, specification, diagnostic tests, etc. • If you have found the best model but still want to show some of your early attempts, then it is a good practice to report the results from the different models alongside each other. • Example: Although probit may be the most appropriate econometric technique to use, you may still want to report the results from OLS. • Sometimes there is no definitely best model • Once you have presented the results you need to interpret them to the reader. How do the estimated coefficients compare with the ones predicted by theory, and how do they compare with the results reported by others? • Always start with the most important results, i.e., not by discussing control variables

24. Format of thesis: Empirical analysis • Note the size of the coefficients, the economic significance. Illustrate the importance •  Follow up with some responses to obvious criticisms and robustness checks. • But do not make statements that undermine your analysis, save serious weaknesses for Conclusion and link them with need for future research • Some students have been taught that in the empirical analysis one is not permitted to draw any conclusions about the results. • This is wrong •  Note the size of the coefficients, the economic significance. Illustrate the importance •  Follow up with some responses to obvious criticisms and robustness checks. • But do not make statements that undermine your analysis, save serious weaknesses for Conclusion and link them with need for future research • Remember, a failed hypothesis does not imply a failed project! • Example: You test the Keynesian theory of consumption, but your regression results suggest that income is insignificant. Then it is okay to conclude that the hypothesis fails.

25. Format of thesis: Conclusions • If you did a good job of explaining your contribution in the abstract and introduction, then there is really no need for any conclusions. • The conclusion should therefore be kept short. • It should also not be repetitive; do not cut and paste what you wrote in the abstract or introduction. • One suggestion of how to bring some substance into the conclusions is to have a paragraph acknowledging limitations. You can also use the conclusions to discuss implications for policy, and future research. • See how well-reputed economists write Conclusion

26. Format of thesis: References • The style varies with type of journal/report/book • We have no standard but avoid non-economics styles, such as putting references in footnotes or numbering them • We recommend the simplified Harvard system • Some examples, • For books the required elements are: Author, Year. Title of book. Edition. Place of publication: Publisher. • Example: ``Gouriéroux, C., and Monfort, A., 1996. Simulation-Based Econometric Methods. Oxford, UK: Oxford University Press.'' •  For journal articles the required elements are: Author, Initials., Year. Title of article. Full Title of Journal, Volume number (Issue/Part number), Page numbers. •  Example: ``Bai, J., and Ng, S., 2004. A Panic Attack on Unit Roots and Cointegration. Econometrica 72(4), pp. 1127--1177.'' • For working papers it is common to use the following elements: Author, Initials., Year. Title of article. Type of working paper. • Example: ``Levin, A., and Lin, C., 1992. Unit Root Tests in Panel Data: Asymptotic and Finite-sample Properties. University of California working paper 92---23.''

27. Format of thesis: Tables and Figures • Each table and figure should have a self-contained caption so that the reader can understand the facts presented without having to go searching through the text for definitions of things. • No number should appear in a table that is not discussed in the text, although you do not have to mention each number separately. • Never copy directly the tables produced by the software program! • Tables take a lot of space -- use them efficiently. • Never report results that you do not know what they mean! • Figures illustrate patterns in the data much better than big tables of numbers. • Figures are suitable for preliminary analysis, but can also be used for illustrating a main result. • Whatever you intend to show later in your analysis should (hopefully) be visible in the data. • Figures are also useful when trying to understand things yourself. Students oftentimes get so ensnared in the technicalities that they forget about looking at the data.

32. Writing tips • Simple is better. Researchers sometimes think they have to dress up a paper or a thesis to look impressive. The exact opposite is true; the less math used, the better. The simpler the estimation technique, the better. •  Do not use difficult words just to impress. • Use the same word for the same concept, (not economic development, industrialization, modernization, etc) •  Avoid using adjectives to describe your work; ``striking results'' and ``extremely significant'' are not okay. • Write with a few words as possible, and try not to repeat things. Note in the text when you do repeat something (as already mentioned) •  Present tense is usually best. You may say ``Dickey and Fuller (1979) find that'' even though 1979 was a while ago.

33. Writing tips • Do not use future tense, • I will analyze … You already did it • Sentences, • Keep them short • Be specific, not vague or general. • many factors affect inflation • Avoid starting with This, if it is not 100% clear what This is referring to. • Paragraphs • Keep them short, a third or half a page is usually enough • Only cover one topic/issue in each paragraph • It can start big and then become focused or the opposite. It can also be a ’list of things’ • A new paragraph starts with an empty line or indentation (1 cm)

34. Writing tips • Beware of the difference between minus sign, em dash, en dash and hyphen • − minus sign, – en dash, — em dash, and cut-off rate • Blunt argues that other interest rate effects—expressed …— (US) • Blunt argues that other interest rate effects – expressed … – (British) • Note the difference between ; and : • 1) Use : when a list made up of sentences follows, separate them with ; • 2) Use : when a sub clause explains what was in main clause, (They brought presents: gold and …) • Use ; for a pause, larger than a comma

35. Some other suggestions when writing • Look at how others write, build a text corpus • Keep track of references from the beginning • Use one reference style from the beginning • Read Rick Wicks style guide on • http://swopec.hhs.se/gunwpe/abs/gunwpe0295.htm although it is aimed at Swedish speakers and I don’t agree on everything, particularly the use of past tense

36. The paper! • The length of the paper: about 25 pages for single authors and 40 pages for co-authored papers. Don’t write more than 50 pages. • Use 12 point font and 1.5 line space. •  The Harvard Reference Style is recommended, but you can use other reference styles; the important thing is to be consistent. There are many Style manuals available on the web (see course page) • Don’t use reference styles based on footnotes or numbers. • Don’t copy text from other sources unless it is a citation. All theses will be sent to Urkund

37. Discussing a paper • Give constructive criticism, • Focus on research question, purpose, method, analysis and disposition • Don’t spend too much time on formalities, typos, etc. Better to write them down and give them to the authors. Just mention major deficiencies. • Suggestion for disposition of the comments • What is the main question? • Is the purpose clear and adequately formulated? • Is the disposition adequate? Are there better ones? • Is the review of previous studies relevant? Is anything important missing • Is the method used relevant for the question addressed? • Are the results reported in a satisfactory way? Are there any alternatives? • What are the results, and do you believe in them?

38. Discussing a paper cont. • No study is perfect, what are the flaws (weak spots)?, find out some • Are the findings of economic substance? Statistical significance does not imply economic significance • Is the interpretation of the results correct? • Do the authors answers the research questions? • Additional comments • Do tables and figures provide support the discussion/analysis? • Formalities •  Are the references correct? (look for missing ones, wrong dates, etc) • Is the list of references correct? • Are tables and figures OK? • Language?

39. Structure of paper • 1) Abstract. one paragraph summary • 2) Introduction. In one or two pages, introduce the topic and state your thesis or argument. Include an explanation of why your research is important. • 3) Theory. Develop a deductive argument from generally accepted principles (e.g., supply and demand) and use that argument to develop testable predictions (hypotheses • 4) Literature Review. Summarize those aspects of existing literature that relate to your project. Use it to motivate your study. • 5) The Empirical Model. The empirical model should be described (e.g., regression model) and justified in this section. Link the model directly to the hypotheses, as well as you can. • 6) Results. This section presents the results of your analysis. • 7) Conclusions. Briefly summarize your major findings and develop the implications of those findings. What are the policy implications?