Writing a dissertation or any other kind of research paper in statistics is quite different from almost any other discipline, for a variety of reasons.
Firstly, statistics is not exactly an independent discipline; it is more of a research approach that can be used in many different fields. This means that you have to be ready to operate with many different kinds of information.
Secondly, statistics all but dictates the type of approach you have to use in your writing: your dissertation has to be a quantitative research put into words, which means that the verbal part of your paper is going to be heavily dependent on numerals, tables, graphs and other similar elements.
Thirdly, the fact that you are dealing with a dissertation does not make things easier, because it is an extremely complex type of academic assignment. Unlike an essay or a typical research paper, a dissertation is much larger (while an essay is rarely longer than 2000-3000 words, there is no real upper limit on the dissertation’s size) and is based on your original research. Differently from most other academic assignments, you have almost complete freedom on the choice of a topic (you have to get it approved by your supervisor, but otherwise can write whatever you want). All this can be the source of many problems, especially for the students who are not used to independent research.
In our guide, however, you will find all the information you need to successfully write an assignment of this type.
Statistics is a peculiar subject that is different from most other disciplines because it often plays a supplementary role in other researches and can be used in virtually any line of work, from sociology and marketing to history and biology. This opens up almost unlimited prospects for different types of research; here are just some examples of what you may write about:
But how does one find a suitable topic for oneself? Here are a few suggestions:
Information sources and raw data are the lifeblood of any research, but it is doubly so for statistics. In other disciplines, the information your research is based on is important – it serves as a foundation for your work, it backs up your theories, it helps to create a context for your research. However, it is always your original research that takes precedence. In statistics, it is the other way around – the data is at the forefront, and your own writing simply shows what you have found out by studying the sources. In other words, you should look for a topic that not just has some data on it, but a lot of it. Even if you know there are viable sources of information, check if you will be able to access them beforehand. Alternatively, choose a topic you are sure you can gather enough data about through fieldwork.
In other disciplines, it may be alright to leave the part of data collection for later. When you deal with statistics, this approach means pushing your luck. Collect the vast majority of necessary information before you submit the final version of your topic to the supervisor.
One can often hear that students should strive to write such a big and important assignment as a dissertation on a topic they are genuinely passionate about. After all, you are going to spend months (at the very least) working with the subject, and it is better to write something you are genuinely interested in. However, it may not be the best idea to take this advice at face value. Yes, you should have some interest in what you write about, but it does not have to be an abiding, all-consuming interest. If you choose such a topic, especially for your first truly independent research, you will be too tempted to treat it as your magnum opus and spend too much time perfecting both the topic itself and the paper based on it. Treat the dissertation as a stepping stone, not the most important thing in your career.
When choosing a dissertation topic, make sure the research will be doable first and consider everything else later. Of course, your paper will have to fill a gap in the existing research on the subject, but this should not be your primary concern. It may sound contrary to everything you have heard before, but you should worry about writing a complete and viable dissertation, not achieving a breakthrough in your chosen field. Once you have found a doable topic, you can virtually always find a gap in literature associated with it and justify doing research on it. Is there already existing research on the subject? You can, among other things:
In other words, pick a feasible topic and find a rationale for using it later.
Once you have determined the general area you would like to work on, it is often worth looking at the available data and asking yourself ‘Can I glean a viable topic from it?’ It is especially true for the situations in which you have easy access to the sources of information that are not readily available to the general public.
Unlike an essay, whose structure is usually the same, a dissertation can be vastly different even within a single discipline. Therefore, we will not provide a detailed description of how you should organize your dissertation – the instructions you receive from your college and instructor will override it anyway. Nevertheless, we will give some recommendations on how you should write certain sections that are normally present in most dissertations.
An abstract is a short (usually no more than a few hundred words) summary of your research. It should mention:
Despite its size, one usually dedicates a lot of time and work to writing it, because you have to condense thousands of words’ worth of data into less than a page. An abstract should be enough on its own to represent your research, and if your dissertation ever gets registered with a database, it is going to become an independent document detailing the contents of your research.
Although it is located at the beginning of the dissertation, it is better to write your introduction after the rest of the paper is already done and you know what exactly you are introducing. The primary purpose of this section is to introduce your topic and research question, provide basic context for the research and lead up to the general direction of your work.
This section is supposed to show that you know where your own research fits into the existing body of literature on the subject. Here you should do the following:
The order in which it is to be done is not set in stone. Feel free to experiment and see what works best in your case, so that you can build a logical and coherent structure.
Methods of data collection and analysis are extremely important in statistics, so make sure to be as clear and meticulous as possible in this section. Mention any tools, equipment, processes, and methods you used to carry out your research. Make your description detailed enough for other scholars to be able to replicate your work.
Here you give a detailed description of all results and findings of your work. Separate them into individual points and be as straightforward as possible.
In this section, you turn back to the context of your work, this time considering the results of your research. Refer to the reason you did it, dwell on whether you have managed to answer your research questions and whether you expected the results you have received. Elaborate on what your research added to the literature on the subject.
This short section enumerates the main points that emerged as a result of your research and analysis of its results and tries to define what they mean for the discipline in general. You can also point out promising venues for future research uncovered by your work.
The sources providing statistical data often accompany it with interpretations. If the interpretation contains the conclusions that are relevant for your research, do not waste time redoing the calculations and use the interpretation as is (at least if you believe the source is trustworthy).
Statistics is a discipline that makes heavy use of visuals, so make sure you introduce a fair share of them into your dissertation. Whenever a graph, a chart, a diagram or a table is appropriate to prove your point, do not hesitate to add one.
Maybe you have received a word limit from your instructor, maybe you have decided upon the length of your dissertation on your own. Anyway, you should plan for length before you write the first word. A typical dissertation is at least 5,000 words long, and it is much harder to estimate such quantities by eye than with much smaller essays. Therefore, you should decide how long each section is going to be at the outset. You can alter these values later on if necessary, but this rough estimate will help you see if you are going in the wrong direction if any individual section turns out to be too long.
A dissertation, especially a dissertation in statistics, is a type of academic work that is heavily dependent on finding, processing and analyzing the sources of information. To write a high-quality paper of this type you will have to read dozens of books, magazine publications and other sources, scour multiple samples of statistical data. With the impressive amount of time you have to write a dissertation, it is very easy to get into a trap of going overboard with reading up for it. As a result, you can end up with too little time to do the actual writing. To prevent this, start writing as soon as you get a basic idea of your dissertation’s structure. You will have to rewrite and edit huge swathes of this preliminary text, but having it ready will add security to your position.
Whatever you read in this manual or anywhere else, the instructions received from your instructor and college take precedence. Take time to read and reread them multiple times before you start writing. Make sure you understand everything. If you have any doubts, ask your instructor directly – it is better than to find out you have to rewrite half of your dissertation later on.
Like any other type of writing, a dissertation is not written in one go. In fact, if you are satisfied with your dissertation after finishing the first draft, you are doing something wrong. To write a good paper, you have to do at least four drafts:
Follow these principles, and you will be able to write a top-notch dissertation on statistics even if it is the first time you do it!