MSC DISSERTATION ECO000012M 2021-22
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
DEPARTMENT OF ECONOMICS AND RELATED STUDIES
MSC DISSERTATION ECO000012M
2021-22
INTRODUCTORY LECTURE - BIPASA DATTA
THE MSC DISSERTATION
The dissertation is an integral part of the learning process of the MSc programme. It gives you an opportunity to put the skills you have learned in your modules this year into practice – the ultimate test. This lecture provides a guide to the process and an indication of what the department expects you to achieve. More importantly, the aim of this lecture is to provide you with some guidance on how to carry out MSc Dissertation research. This lecture will provide you with a few tips, designed to ensure that you avoid the pitfalls and enjoy this part of the degree.
Aims and Objectives of the Dissertation
The aim of the dissertation is to allow you
To carry out your own in-depth analysis, or your own in-depth critical review of a topic, working
mainly independently.
The dissertation provides a first opportunity for you to do your own academic research in economics
or finance. A high degree of originality is not expected. But,
We do expect you to make a serious and honest attempt to put into practice the principles and
techniques which you have learned in the taught component of the MSc degree, in order to arrive at some carefully reasoned conclusions of your own. Time management skills are a crucial part of this process.
What are you expected to write in your dissertation?
In a nutshell, you will be investigating a specific research idea within your chosen field of research topic to provide a clear exposition of the problem and a solution. You will provide an account of the relevant theory in your own words, (possibly) carry out some original empirical analysis, and reach provisional conclusions.
Types of dissertation
Most MSc dissertations will be of one of three types:
(a) Purely empirical research
A piece of your own empirical analysis, typically involving the use of econometrics in order to study a set of data, or at the very least involving the analysis of the data using tables, charts, statistical
summary measures, etc.
(b) Purely theoretical research
A piece of your own theoretical analysis, typically involving the construction of a theoretical model using appropriate mathematical techniques, and the formal derivation of results from it.
(c) A combination of theory and empirics
A piece of your own work combining (a) and (b) where you first produce a theoretical model that you test empirically. Indeed, students often start out with the idea that they should first come up with a new theory about something, and then test it econometrically against the data. However, in general we would warn you against trying to do both of these things together in an MSc dissertation . It is nearly always too ambitious a task to achieve in only three months. Even in a PhD thesis, which takes 3 or 4 years, most students choose to specialise in either empirical work or in theoretical work, rather than in both. You are more likely to be able to go into depth, and to obtain worthwhile results, if you confine your efforts more narrowly.
(d) Critical Literature Survey
Your own critical survey of an area of the existing literature, structuring it in an illuminating way, making your own critical comparisons of the papers within it, and providing your own constructive commentary upon them.
NOTE: A catalogue of previous articles without evaluation does not constitute a literature survey. Similarly, simply regurgitating existing surveys, or textbook treatments of very settled areas of the literature do not constitute a critical literature survey.
How to structure your MSc dissertation?
Normally, your dissertation should have the following structure:
- The content page
- An abstract (usually no more than 250 words)
- Acknowledgement
- Introduction
- Literature review
- Theoretical and/or or empirical model
- Analysis of results from your model
- Conclusion
- References
- Appendix
NOTE:
(i) All dissertations should contain some element of literature review, i.e. (d). (ii) Dissertation Word count: 10,000 words excluding appendices, references, figures, tables.
(iii) A word about references: make sure that your references are not too dated: ideally they
should not be more than 10/12 years old except for some key papers on which the entire research idea is based (i.e. those classic pioneering papers).
How exactly do you carry out your dissertation research?
Start with a literature review. Normally, you will start your dissertation work by spending some time reading the relevant literature. Your written-up dissertation will typically contain an entire section in which you review this literature in your own words. The exact amount of space that you end up devoting to a survey of the literature in the submitted version of your dissertation is likely to depend on how well your ‘original’ research progresses. If it goes very well, you may want to devote as large a fraction as possible, out of the available 10,000 words (the maximum permitted length of the dissertation), to reporting your original research. You would therefore limit the survey section to the minimum that is necessary to provide the background and motivation of your research. On the other hand, if your own research makes only slow progress, you may want to expand the literature survey component.
Learn how to conduct a literature search efficiently.
Google scholar search. Recent papers on Google Scholar will contain bibliographies – look for overlaps which indicate the significance of an earlier paper. Learn how to assess an academic paper quickly. Most of the interesting ideas in a paper will be outlined in the abstract and introduction along with a statement about the contribution and its context. Take a look at these and if the paper is not relevant to the issue of your survey then don’t get stuck into the maths but move on quickly. However, do some of these other issues now look more interesting than your original one? If you make an early start you will have time to change horses. Otherwise you will have to stick with your original one.
Reference list of relevant papers and Authors’ websites. Once you have found some relevant papers, take a look at the reference list and authors’ website for papers in the same area. Unlike bibliographies which take you back in time, their more recent discussion papers will take you forward in time and bring your survey right up to date.
Recent issues of journals. It is always a good idea to flip through the recent issues of relevant journals. Use certain keywords to look for papers in similar areas. If you have found some related papers, also take a look at the references.
If you are carrying out Empirical research (i.e. your dissertation is of type (a))
If you like econometrics, try to find a tractable model or interesting hypothesis that either
(i) has not been tested against the data, or
(ii) try to find a new data set to test it. For example a model originally fitted to US data might be fitted to
UK data, if available. Are there any other explanatory variables that are relevant and could be tried?
Empirical work is usually less risky relative to the theoretical research. This is because, in the three months that you have available to work on the dissertation, and given that you have to work mostly independently, a short empirical study is usually easier to complete than a theoretical modelling exercise. It generally gives you more freedom and more scope for originality.
If you are carrying out Theoretical research (i.e. your dissertation is of type (b))
For a theoretical dissertation, an extension or modification of an existing theoretical model is the most likely way of arriving at a precisely-defined exercise, instead of writing a completely new mathematical model (which will be a very ambitious work and indeed is more suitable for PhD research). Indeed, since
published theoretical papers are often very terse, and omit to present fully laid-out proofs of all their results, it can sometimes be enough to provide your own explicit proof of an existing result, or even of a slightly more special result, by taking some more specific functional forms. You can, for example, try the following:
(i)
(ii)
(iii)
NOTE:
(i)
(ii)
Change an assumption and redo the model
Provide some specific functional forms to solve a model
Try some numerical simulations or provide examples to provide a variation on, or illustration of, existing results. Again, a good dissertation should not just make arbitrary variations: you need to provide good ‘motivation’ for your own investigations, by explaining why your variations are interesting, and by thinking of hypotheses which your formal model will test.
Theoretical research perhaps is the most ambitious task.
Theoretical research means doing original mathematical models or extending an existing mathematical model in a meaningful way e.g. adding or relaxing new assumptions and/or features. DO NOT confuse between producing a theoretical piece research with something that is purely a long descriptive piece of work!
If you are writing a dissertation which is a (critical) survey of the literature (ie. Your dissertation is of type
(d) category)
Here, the objective of this kind of project is to acquire more detailed knowledge of an area and to develop
analytical and evaluation skills. Given your chosen field of research, you must
identify and critically evaluate the key issues in this area.
try to find gaps or areas of weakness that could be addressed by future research
N.B. this is the minimum requirement - a catalogue of previous articles without evaluation is not sufficient.
be very careful to attribute rather than to plagiarise (more on this below)
the literature survey may lead naturally to a type (a) or (b) project.
it could also furnish a topic suitable for research at MPhil level.
For a critical literature survey, the greatest novelty is likely to be achieved by selecting a new and/or recently quite active specific area within your chosen field of topic . You need to place the contributions in some sort of order, to bring out their similarities and differences, and to highlight questions which remain unanswered. Avoid simply regurgitating existing surveys, or textbook treatments of very settled areas of the literature.
Summary
These categories are not mutually exclusive. A brief review of the literature is still required and could
afford a fallback position if the project is not feasible or fails for some other reason.
It is possible to combine theoretical and empirical modelling, indeed it is important to test new
theoretical models against the data. However that is very ambitious for an MSc dissertation and we would warn you against that.
At this level, we are mainly interested to see that you deploy these techniques correctly. So for
example in testing a hypothesis empirically you do not have to find a high level of significance or even ‘the sign’ indicated by the theory – gratifying though that may be!
Make sure that you are not ‘reinventing or re-calibrating the wheel’: survey the journals and the
internet carefully for similar work – see the section on plagiarism below.
Explain your results and put them into the context of the literature. Pay special attention to the
introduction and conclusion. Ask a friend to look at them to see that they are self-contained and make sense.
Common pitfalls in dissertation work
There are a lot of pitfalls ready to trap those attempting type (a) or (b) projects.
Too ambitious! The most common pitfall, i.e. strategic error, is to be too ambitious. Being too
ambitious is a natural tendency: even if you have chosen your dissertation area of research, chances are that you are still relatively ignorant of the subject area. So you have the tendency to do something that is too broad. The most successful dissertation is likely to be the one which is narrowly focused. Having a precisely defined question, and a clear plan of the method which you will use to answer it, is the best way of ensuring that you use your limited amount of time effectively. As you delve deeper into a topic, you will usually find that it comes to seem bigger than you first thought. When you only have (approximately) 3 months to work on it, it is vital to choose something that is feasible.
(Un)Availability of empirical data. A common pitfall in empirical work is being unable to obtain
relevant data. In planning your research, you should therefore check the availability of suitable data at an early stage: don’t leave this to last. You must discuss this with your supervisor in your initial meetings. If you cannot find data for the country or time period you are looking for, consider whether you could perform a similar study for some other country or time period.
Hard to solve theoretical models. A common pitfall in theoretical work is being unable to solve
the model whose assumptions you have written down, or being unable to derive unambiguous and easily interpretable results from it. To reduce the risk of this, always start with the simplest possible model e.g. use a specific tractable functional form, make simplifying assumptions that captures the features that you want to capture. Don’t be tempted to add in extra complications unless they have a clear role in enabling you to answer the theoretical question which the model is intended to answer. It is much better to start with what may at first appear to be a ridiculously simple model, and then extend it, than to start with a model containing a long list of ‘realistic’ features, and then ‘get stuck’ in your ability to analyse it. It is OK if you cannot prove propositions or theorems. You can still provide conjectured results and explain how you will tackle the proofs. You can provide heuristic arguments.
Now that you have chosen your dissertation research topic, what should you do?
Even before having the first meeting with your supervisor, start narrowing down your research topic
by identifying some key questions and very specific research ideas within your broad area. To do that,
Start your literature review search as described above.
Look up your MSc module reading list if relevant, and pick articles from a list of sources for ‘further
reading’ .
Have a quick read at some of the articles that you find interesting.
Remember: without further reading it is almost impossible to find meaningful research ideas and or
identify key issues to explore further.
What help can I expect from my supervisor?
First, your supervisor
(i) Will help you fine-tune your research topic more precisely.
(ii) Should advise you on further relevant reading and how to approach the analysis of your chosen
topic.
(iii) Will help you to decide (in the case of an empirical dissertation) what sort of data you might need
and what sort of regressions to run; or (in the case of a theoretical dissertation) what model to use and what questions to use it to analyse.
Second, your supervisor will
(i) Read your drafts and give you comments.
(ii) Give you comments on your (semi) final draft before submission.
Your responsibility
(i) The dissertation is mostly about learning to work independently and finding your own way to overcome problems. Hence, you should not expect a lot of detailed supervision – as in any research work, you need to also learn how to sort out problems if you encounter any . However, If you have serious problems, you should of course seek your supervisor’s advice.
(ii) You should also keep your supervisor informed about your progress. Don’t be tempted to drop out of
contact with your supervisor. This is not just a matter of courtesy: it also helps to avoid any suspicions of plagiarism (see further below).
Frequency of supervision meetings. There is no strict rule as to how many times you need to meet with your supervisor. We normally expect students to have a meeting with the supervisor at least once a month to discuss their work and get feedback. The frequency of the meeting can be more (or even less) depending on your progress. Your supervisor will have specific requirements and s/he will advise you how many times you should meet. You must show your drafts to your supervisor regularly. Do not leave your draft until the last minute, and allow sufficient time to receive feedback.
Plagiarism
Plagiarism means using other people’s work without explicitly acknowledging this. It is a very serious academic offence and you could lose your degree if you plagiarise and are found out. Academic staff are always alert to the possibility of plagiarism in dissertations, and are skilled at detecting it, so do not be tempted to present other people’s work as if it is your own.
Students sometimes seem to be tempted into copying and pasting whole sentences or passages of other authors’ text into their dissertations out of the belief that the student cannot express the ideas better him- or herself (especially if the student’s first language is not English). You must, however, resist any temptation to do this. It is essential that all explanations are given in your own words. When you are reporting someone else’s results, or giving an explanation of some standard, textbook-type, piece of theory, do not
have the other author’s work open in front of you: use your memory, or your notes, to reconstruct the argument in your own way. There can be a small number of occasions on which it is useful to include a direct quotation from another author in your dissertation, although these occasions are usually few in economics. In that case, the passage should be put in quotation marks and the exact reference from where it has been taken should be given.
An important way to give your supervisor confidence that the work in your dissertation is your own is to remain in contact with him or her about the progress of your research. If you disappear for most of the summer and then submit an apparently perfect piece of work in September, this will raise suspicions. If, on the other hand, your supervisor is familiar with the various stages of your work, and can see how the submitted dissertation reflects the discussions which you have had, this is evidence that the work is genuine.
(Approximate) Dissertation timetable
The approximate timetable for dissertation research is as follows: Once you have been notified of your topic and dissertation supervisor (by the end of April/early May),
May
Schedule your first meeting with your supervisor to settle upon a specific research idea that you both
are going to focus on.
You are now encouraged to get started with your dissertation research with a view to continue doing
some solid research after the summer term exams.
Meet with your supervisor as frequently as you both wish.
July
You have started working intensely on your research.
You should have prepared a detailed framework of analysis, done some solid literature review, and
have a timetable for preparing your dissertation for the next couple of months until the final submission date, agreed with your supervisor.
Your modelling is also well underway.
You should be in regular contact with your supervisor and show him/her your draft work (can be a
particular section or an part of the analysis)
August
By early August, you should have done some substantial amount of work and are ready to show your
first full draft to your supervisor.
By the second half of August, you should be ready to show a semi-final draft of your entire work to
your supervisor so that there will be enough time to receive feedback and incorporate them before you finally submit your dissertation by September 16, 2022.
The above is a suggested timetable; your supervisor will provide more information about what s/he expects from you.
Dissertation submission
All MSc dissertations must be submitted online at the Dissertation Submission Point on the VLE. Information and instructions about how to submit your dissertation will be posted on the VLE in due course. You will be notified when that information becomes available.
For further information on extensions on deadlines etc., please consult the dissertation memo.
The deadline for dissertation submission for All MSc dissertations:
3 pm on 16th September (Friday) 2022.
2022-09-02