This week I started my research for my masters thesis. To say “I’m struggling” would be an understatement.

The research part of any thesis is probably the most crucial part of the whole odyssey. It’s the part that sets the tone for the next 6 months. I guess struggling with the research will mean less struggle with the rest of the process. Hopefully…right?

Photo by James Pond on Unsplash

My biggest issue is staying positive. There’s a lot of self-doubt involved: My mood switches between “Nobody’s interested in that crap and this is the pettiest thing I’ve ever worked on” and “Oh my, I’m not the first one to think about that. Might as well stop now, since this has been done before”.

I have yet to figure out how to snap out of this…but it’s ok. I mean it’s been five days…

My main focus the past few days was to figure out the perfect workflow for my research and this should be the focus of this little post. When I wrote my bachelors thesis my research was extremely unorganized. This wasn’t too bad as the topics I had to cover were less complicated. However, when I was finally done with the programming and started writing the thesis, I very much despised my 2-month-ago-self for not organizing all of the sources correctly. I lost a lot of time finding stuff again that helped me when I first started my work. This is something I really want to avoid this time which is also why I wanted to start this blog in the first place. I pieced together the workflow I described below. I feel like this will work really well for me. The emphasis lies on “for me”. This might be an entirely wrong approach for someone else.

The Art of Googling

The first place I usually go to when researching a topic is Google Scholar. The biggest challenge of using Google Scholar is that sometimes you don’t know enough about a topic yet, so you have a hard time finding the right keywords. In my case, I needed to find research items that are not only associated with computer science but also have a strong presence in the educational sector. The educational sector is something I’m very unfamiliar with, so I had to make sure I was looking for things the right way. I started with searching for papers about general education trends and collected a nice set of keywords by going through different papers and their references. Once I had the right keywords for the educational side of things, it was fairly easy to pair these with my computer science related keywords like “Machine Learning” or “ANNs”.

I started composing a list of all the keywords I used so far which will probably come in handy when I have to re-search (not research ha!) a few things. 6 months is a very long time, especially with topics like Machine Learning where new work pops up every day.

The Process

The first round of research took me about five days. The first three days I worked through my keyword list and just googled away, read through papers (mostly skimmed them). Anything I found remotely interesting made it to my bookmarks. Once I bumped into the same papers over and over again I stopped.

The fourth day I started going through my bookmarks and assessed every single item. When I found a paper/post was good, I added it to Zotero, a free research assistant. This is going to make writing the thesis a lot easier thanks to their BibTeX support. Moreover, this tool provides a great way to organize your research which brings me to my next point: Categorizing.

While assessing my bookmarks I started categorizing in my head. In the process of adding the papers and posts to Zotero I came up with the following categories:

• Education
  • General
  • LEGO Mindstorms
  • Machine Learning
• Excercises
• LEGO Mindstorms
• Machine Learning
  • Basics
  • Tools
• Visual Programming

Additionally, I tagged papers that are in some way significant (cited a lot, lots of good information,…) with “important”.

I now have about 50 research items in my Zotero library. Categorizing and tagging all of them definitely makes the whole thing a little less confusing.

And less confusing equals less self-doubt.