Writing a thesis is hard work: a lot needs to happen in a little time. I've created this document to clarify responsibilities and meeting procedures, adding general advice as well as advice specific to particular chapters. My hope is that this will streamline the process and free up more time to discuss content. So please read this document before our first meeting. You may also find it useful to consult throughout the process. Check this site out before you email me questions. I update it to incorporate more useful tips and advice based on past experience.
Of course be sure to check out the masters thesis Canvas page with examples of good theses, the thesis manual, and advice pertaining to specific methodologies.
Good luck with your thesis!
My job is to give you advice based on the written product you deliver. I only look at finished written chapters. The ideas and writing are your responsibility. It is not my job to ensure that you graduate. I may identify problems, but the solutions are for you to discover. Of course, if it is something beyond what is covered in class, I will step in and offer more guidance. But if it involves things you should know about, based on marketing classes, or general knowledge, then it's your responsibility to identify the solutions.
A few remarks about this:
Questions like “what is the right ____ (e.g., model, theory, explanation)?” are mostly for you to figure out, not me.
As another example, I may say you need to fix your hypothesis, meaning, to write them in a way so that the variables and relationships are clearly described. But you must figure out how to write them the correct way.
Feedback I give may require global changes to your thesis. For example, if my advice is to change your statistical model in chapter 3, you may need to revise what you write about it in chapter 1. These are your responsibilities to fix.
Lastly, I focus on the most critical errors as I see them. Hence, I cannot correct everything. Therefore, it is entirely possible that a chapter I reviewed contains errors that a co-reader spots. The responsibility is on you to keep revising (and hopefully improving) your thesis.
I am not the guarantor of quality, you are.
We will have 4 meetings in total. We cover a chapter each meeting, except for the last meeting when we will cover chapters 4 and 5.
The meetings will typically be 4 weeks apart (once per month).
I will read the new chapters you have submitted, plus the revision of the last chapter, not all the earlier ones. I read through the entire draft on the final meeting.
A one-on-one meeting will last no more than 30 minutes. If you are in a thesis group, it will last no more than 90 minutes.
We will hold meetings remotely on either Teams or Zoom.
If your thesis is based on a project for a company, you are encouraged to invite your company supervisor to the first meeting. Students often complain that they find themselves in the middle of a conflict between their academic and company supervisors. You can better manage this if we agree in the beginning about what the aim and scope of your thesis will be.
At least 2 days before the meeting:
You must upload your draft of the agreed upon chapters to a Google Drive folder. If there is no drive folder with your name, create one.
All thesis chapters must be in one file.
For the chapters which you have revised (but not the new one):
Use a highlighter to indicate which parts you have changed. If there is nothing highlighted, I will only look at your newest chapter
Next to those changes, use a note to explain why you have made those changes
No thesis, no meeting.
Tip: prepare a list of questions and issues you would like to discuss. Use this time efficiently. It is your job at these meetings to ask questions. Email is not the place to ask me questions. In response to your question I may ask you to consult a resource or guide you in the direction of where the answer may be found.
First provide a summary of what your thesis is about. Then, describe what you have done. This helps me and other people in the group give you feedback.
You are leading the meeting: do you have any questions about my comments? Are there other issues about this chapter or others you want to raise? Now is the time to ask. I will not go over my comments unless you ask.
In the last 5 minutes, I describe what is required for the next step and plan the next meeting.
I provide a thesis template provided on Google Drive with Tables and equations embedded: I suggest using it.
Start every chapter with a few sentences about the chapter's contents. First, I will discuss .. followed by.
In each of your chapters, try to thoughtfully organize what you want to say into subsections and paragraphs. Actually it's better to start with an outline: what subsections will you need, etc. If it's a jumbled mess, no one will understand what you are saying and give up reading; this leaves you with a low grade.
In a thesis every point you make must be defended: you can argue for it, giving examples or statistics; you can cite previous literature. But you cannot assert that something is true without any supporting arguments. This is not like an opinion column in the newspaper.
Resist using the acronyms or words that are only used in the company. Students writing their thesis in conjunction with a company tend to adopt that company's particular vocabulary. This makes it hard to read for the supervisor and co-reader. Use standard academic or common terms instead. If you must use it, of course be sure to define exactly what it means.
Resist inventing your own variables if there are already perfectly good variables that do the job.
Be consistent in naming your variables: keep the same name throughout.
Use the APA format to properly reference and cite material. You can see more here: "How do I reference?"
Also, make your tables and figures in APA (American Psychological Association) format.
Remember the page limit is 35 (see the thesis manual).
Number pages and the distance between rows should be 1.5-space text (see masters thesis manual step 3)
Check your document for spelling and grammar.
Consider visiting the Scriptorium in the Language Center, which offers free advice on writing and information search. Do this multiple times; my experience is that it always results in improvements!
Note: this is meant to supplement the information that's in the masters thesis guide on BB.
On Google Drive, see the Business Research videos: specifically defining a business problem and formulating problem statement and research questions.
A clear problem statement has (1) a dependent variable, (2) independent variable(s) and/or moderators, and (3) a context. It should address the discussion in the preceding text, the problem indication.
Consider the scope and feasibility of the thesis: what are you focusing on and what is not being studied.
Avoid long descriptions of the company; only include what is relevant to the problem you are studying.
On Google Drive, see the Business Research webclips on conceptual model, building hypothesis, moderation and mediation (where relevant).
Need help finding enough literature on your topic? The Scriptorium has information specialists who can help you out.
What journals should I consult? Not all marketing journals are equal. Here's a list showing what the top and near top journals are.
Generally, I advise organizing this section as: (1) a short first subsection on literature related to the general context (if any); (2) subsection(s) for the variables in the conceptual framework (e.g., corporate responsibility); (3) finally a subsection on the relationships among variables, including hypotheses, and conceptual framework (CF) diagram.
Almost everyone tends to write too much in this part, perhaps to signal how much was read. Think Mark Twain. Take the necessary time to be concise, including only discussion related to your context, variables, and relationships among variables (hypotheses). Do not include everything you've read. Only what is relevant.
Organize your paragraphs by theme (e.g., variable or relationship), not by paper. One sign you're not organizing well is if your first sentences of the paragraph tend to begin: "X, Y, and Z (2004) find that ... "
Each hypothesis should be supported in the text immediately preceding it. Do not write, "the last 3 pages leads to the following 3 hypotheses: ... ".
Not every arrow in your conceptual framework needs a hypothesis. Only what is new for your thesis. Other variables can be called "control variables". See for example here, on Figure 1. There are 3 hypotheses (about groups of variables), and two control variables labeled in the figure.
Consider using a table: sometimes it's useful to summarize the literature (comparing what has been studied to what you are studying), or your hypothesis (and the arguments or papers they are based on), in a table.
Use clear language.
On the masters thesis canvas page, there is material on conjoint, surveys and experiments.
Here you should discuss (1) how you plan to measure the variables in your conceptual framework, including the relevant survey questions if they are new or different than usual; (2) what your population is and how you plan to sample from it; and (3) the model and how you will estimate the model.
Equations help clarify things, but make sure they have proper subscripts, and the variables are defined in the text.
Number your equations and write them in Word Equation editor. This is a very easy way to set this up: https://superuser.com/questions/594559/how-do-you-easily-add-equation-numbers-to-microsoft-word-2010-equations. There's a thesis template with equations provided on Drive folder.
The results generally follow this structure: 4.1, descriptive statistics about your sample, giving us some sense either formally (with statistical tests) or informally of how close your sample is to the population; 4.2 your main results, hypothesis tests; 4.3 validity, additional results, robustness checks, etc.; 4.4 a quick summary of what the results mean.
Put the most important tables and figures in the text, not in the appendix. For example, tables containing descriptive statistics, model estimates possibly with robustness checks, figures of marginal effects, simulations-- these should be in the main text. Stuff like regression assumptions can be relegated to the appendix.
When going through your main results, take the reader CAREFULLY through your results. Explain them well. Do not simply report the coefficients or test results. Give some sense of quantitative effect size (e.g., not just positive but how positive) either in words or a figure.
Many students are hung up about statistical significance: the effect exists if p < 0.05 and does not otherwise. (We teachers are somewhat to blame for this.) Actually, it's not good idea to do this; we should move beyond this binary rule and just report effect sizes and confidence intervals. See https://www.ama.org/2023/12/12/time-to-abandon-null-hypothesis-significance-testing-moving-beyond-the-default-approach-to-statistical-analysis-and-reporting/
The standard assumptions of regression are not equally important. Don't spend so much time on them. Non-normality of errors is almost never important, unless you have a small sample. Dependence is a problem if you have time-series or perhaps spatial data (rare). Same goes for outliers.
Another fairly common worry is multicollinearity. Some good advice, particularly point 2 which is common when moderators are present: https://statisticalhorizons.com/multicollinearity/
Often heteroskedasticity is present, which affects the standard errors of the effects but not the effects themselves. This is easily solved using robust standard errors or clustering at the panelist level in panel data.
Consider taking logarithms of all positive variables: it leads to a multiplicative model which often makes sense. (log(y) = bX.... + epsilon)
Consider a summary table with your results or hypothesis tests, outcomes, and interpretations at the end of the chapter (section 4.4).
Before (in section 4.4) you gave a focused summary of your results given your sample. Now you have permission to extrapolate. Taken together, what do your results mean? To what other situations would these results generalize? What are the implications for firms, consumers, policy makers, marketers, etc? Do not veer so far away from your results so as to make the connection tenuous. The implications have to result from the findings in your study.
What are the limitations of the analysis? Measurement, sampling, analysis, data issues..
What goes here?
Your questionnaire in the original language, and translated in English if necessary.
Pre-test interviews or focus groups. Protocols or details about the procedure.
Analysis that is central to your thesis (what I called the main results above) belongs in the main text not the appendix. If you take more than 1-2 sentences to discuss a result, it should be in the main text. The supplementary analysis, like robustness checks, correlation tables, outlier removal, ... belong in the Appendix.