How To Analyse A Questionnaire / Survey For A Dissertation Or A Research Paper?

Buy me a coffee: buymeacoffee.com/r3ciprocity

Listen to my new podcasts: podcasts.apple.com/nz/podcast/r3ciprocity-podca…

In this video, I detail how I would go about analyzing a questionnaire / survey for a research paper or dissertation. First of all, before you do much analysis, you should try to think about your research question, and how the survey / questionnaire relates to the research question. The key thing is to specify as much of the analysis as you can before you touch the data. Why do you want to specify your analysis before you touch the data? We have a tendency as humans to look to confirm our thesis, and the goal in science is to falsify your hypotheses that relate to your research question. By specifying as much as the analysis upfront as possible, you prevent yourself from being human.

The other reason is far more practical - the more you understand what you are doing upfront, the less time you will waste with analysis. Your goal is to write, and not analyze with nothing written.

What I would start with is to do construct validation - does the data seem to hold together as you would expect it to hold together. What if it does not hold together? You should try to think about your context to identify possible moderators or controls that prevent this from occurring.

2. The next step is to plot the data. Always plot the data! Try to think about simple scatterplots (x-y plots), or line plots. You will learn far more about your data by looking at the data, than looking at some correlation matrix or statistic. Most / all statistics should be ‘eye-balled.’ Does it make sense given what your data looks like?

3. Think about how your data fits with past studies? What should you include or not include in the analysis? Everything should make easy sense. If it is not easy to understand, you have made your study too difficult.

4. This should be your last step. Run your regressions by adding each variable at a time. You will often see patterns in the regression runs when you add a variable. If the variable causes everything else to be unstable, think about why this is unstable. It may not be a problem, but you have to justify why that is occurring.

5. I often will just run different post-hocs just to make sure that the results are more or less stable. No data is going to be perfect, and the results change with the new post-hoc, specify why that is occurring and how it is occurring.

6. If there are limitations to what you did, or the data is not performing as expected, this is a good thing! You just need to specify why that is occuring. You can get other people to research why you did not find what you thought you would. An example of this is the exploration of ether to explain how light travels through the air. It was a puzzle why no one was confirming the theory - if you were the person that suggested that the hunt for ether was misguided by a robust set of failed tests, you would have been very famous!

7. Also, remember that you don’t need to use all of the data in the survey in your analysis. Save some for future research.

Check out:

How Do Professors Grade Papers And Evaluate Manuscripts?
youtu.be/RFj4FkezFdc

Mistakes That Most PhDs Make (Mistakes In Doctoral Research and Our Scientific Careers)
youtu.be/xwzNtQdffoY

Why Get A PhD In Strategic Management? And, Questions About Strategy Doctorates.
youtu.be/rvE-0dAefsM


***

David Maslach is a research professor of entrepreneurship, innovation, and business strategy, I discuss topics, such as behavioral science, strategy, innovation, and entrepreneurship, and apply these to my new peer proofreading and editing platform. Topics include the sharing economy, altruism, investing in technology, starting a business, and bounded rationality. My favorite videos pertain to incentives, goal setting, and learning from failure to drive behaviors such as weight loss, stopping telemarketers, creating novel technologies, and creating new movements.

r3ciprocity.com: Peer proofreading and editing platform

A new platform where you can earn credits by editing other people's documents. Use these credits to have your own work edited. If you do a good enough job, you can convert these credits to money.

The goal of the platform is to get people to 'pay it forward' and help other people out by creating incentives for people to give back.

Check out r3ciprocity.com

Please subscribe to the Youtube channel:
youtube.com/channel/UC5spxk7bNDMGPSHjW_8ndZA