Statistical assistance for a master’s or doctoral thesis if usually thought of as a mere statistical analysis. But a good statistician can prove useful to you at several other stages of your struggle to obtain an academic degree.
A correct and well thought-out research methodology is the foundation for success of the entire project. And mistakes made at this stage are often irreparable at the later stage of data collection or analysis, even by the most skilled statistician.
On the other hand, it is also not true that absolutely every master’s or doctoral thesis requires consultation with a statistician at the methodology development stage.
The most common areas of statistician’s assistance in determining the methodology of a master’s or doctoral thesis are:
When can a statistician’s help be particularly useful at the stage of developing the methodology of a master’s or doctoral thesis?
This is where a statistician can help you, for example, in preparing recruitment lists if you are planning a blinded study, or in entering the collected data into a database, or in preparing the database for work in statistical software.
Statistical analysis is the main stage where a statistician is involved in the project. The terms such as “statistics for a doctoral dissertation” or “survey development for a master’s thesis” will not sound terrifying for you if only you plan the research methodology well and collect the data properly. Learn more about statistical analysis.
The role of a statistician doesn’t have to be limited to just making statistical calculations and producing hundreds of tables in statistical software. Such tables contain a lot of information, but they are never suitable for simple pasting into your work. Statistical elaboration involves selecting appropriate sections from these tables, and then often using them as a basis to create new tables and graphs altogether – ones that will provide and justify the direct answer to the hypotheses formulated at the beginning of the study. The challenge is even greater if your thesis is prepared as scientific publication and you have to compress your long research paper’s statistical analysis into a mere few-page article. I such case, these tables should be appropriately formatted and described in compliance with the standards of data presentation. Appropriate commentary and interpretation are also necessary. Everything matters here and it all may contribute to rejection of your paper by a serious journal.
It is worth making sure that the statistician who did the statistical analysis for your master’s or doctoral thesis also accompanies you at the stage of questions from the thesis supervisor and/or reviewers. This is particularly important, because some of these questions can sometimes only be answered by the person who did the statistical analysis. At many universities, questions about statistics during doctoral studies are commonplace. You will simply feel more confident and relaxed having the support of a statistician at every stage, including the final one.
Waiting until the last minute to do a statistical analysis is not wise. Remember that a good statistician won’t be able to do a statistical analysis “for yesterday,” especially during the thesis writing season (i.e., spring/autumn), because they simply have their schedule jam-packed with various assignments. So don’t wait until the last minute and either order a statistical analysis in advance, having the comfort of time to finish writing your thesis, or at least book a date.
Additionally, a full statistical analysis takes time. From familiarizing yourself with the goals and hypotheses of the research paper, through selecting appropriate statistical tests, to presenting the final results – if the whole thing is to be done well, there can be no cutting corners.
Choosing the cheapest offer on the market without checking the terms and conditions first is buying a proverbial pig in a poke. Make sure what exactly you will get, because the term “statistics for a doctorate” or “full statistics for a master’s thesis” can mean different things and can veil very different content, depending on who prepares it. Unfortunately, we are often approached by Clients asking us to “translate” the statistical analysis prepared in this way into their own language (Statistical Analysis).
Another mistake is to outsource the statistical study to a random person, “because there’s little time left”, or “because it’s cheaper”. We wrote about this in more detail the ”Statistical analysis” section
Involving a statistician too late is a basic and really serious mistake. Unfortunately, a very common one.
Intuitively, one thinks of a statistician when the data has already been collected and the time has come to analyze it. Unfortunately, sometimes it is too late and it turns out that what you have collected does not allow you to answer the basic questions of your project. We wouldn’t wish this on anyone ☹.
And sometimes things that seem insignificant and trivial have huge impact on the data analysis later on. The key to success is a well-designed study and a well thought-out methodology. If you have experience in planning research projects, you can of course plan them yourself. You are often fortunate to have a clever, experienced and helpful supervisor to help you with the methodology of your dissertation. Do not hesitate to use such assistance!
However, if you neither have experience in research work, nor can you rely on your supervisor for help, we STRONGLY suggest involving a statistician already at the stage of developing the methodology for your project.
We will conclude this with a cliché, something perhaps obvious. Neverthless, it is a cliché which our Clients too often forget about.
FIRST develop hypotheses and research questions, and only THEN collect the data.
Never the other way around…
And ALWAYS consult your research questions and the analysis plan with your thesis supervisor BEFORE you start collecting the data.
Is that all? Yes – only that, and as much as that.