One of the hardest things of the PhD is to get the thesis written up. I submitted mine a few months ago and can honestly say that the writing up was not as easy as I thought.
For me, the hardest part was knowing if I was saying things the right way. At times I realised there was something odd about my sentences, but I couldn’t figure out what it was. Other times I had a sentence in mind, but I just kept looking for that one word that would fit in nicely. It was these sentence-level issues that made the thesis-writing process slow and painful.
How did I deal with this? Like many of us do: by using Google as an accuracy check. That is, by endless visiting and re-visiting the Google webpage to enter chunks from my text, and checking how often these were found on the Internet. The more results Google gave me, the more ‘accepted’ the chunk apparently was, and the more confident I would feel to use it. And it’s not just me doing this – as soon as I started admitting this (slightly embarrassing) habit to my friends, I realised that most of them did the same.
And we aren’t crazy for doing it. The areas of language learning and teaching make great use of this frequency-technique to develop classroom materials. (The technique itself is called ‘corpus linguistics’).
But using Google for this purpose does have a few annoyances. The biggest one is that you constantly have to leave your text to check the frequency of chunks. Another is that you can’t really trust Google’s feedback. Something sloppy like ‘gotta’ might give you a lot of results but is really not something you should be using in your thesis.
So I decided to develop an app to make this frequency-approach more accessible and more accurate. I wanted something non-interfering; an app that would allow you (the writer) to get the frequency results without having to leave your document. I figured a popover would be best for this: You could select a chunk from your text, activate the popover, and get the feedback you needed (for example the frequency of that chunk on the Internet).
A tricky question was what type of feedback I could offer. I figured the first option would be the number of results, which would give you the same information as a Google search. The second option I decided to add was a ‘compare option’, where you could directly compare the number of results of two chunks (so this is basically doing two Google searches in one go). Besides these two, I added a few options tailored for writing. One was a display of examples of your chunk in different contexts, to give you an idea of how it could be used.
Another option was meant to help you in finding the right word for a context: by putting a star (*) in your selected chunk, you’ll get a list of the most frequently used words in that particular spot. For example, if you select ‘research should * this issue’ your list will show ‘research should address this issue’, ‘research should explore this issue’, ‘research should examine this issue’, etc. A last option I added provides synonyms: by putting a star before and after a word in the chunk you select, you’ll get a list of the most frequently used synonyms of that word in that context. For example: ‘a *difficult* situation’ will give you ‘an awkward situation’, ‘a serious situation’, ‘an embarrassing situation’, etc.
I decided to use the Google Books database to get the data. This is a massive database which includes more than 5 million published books. The best part of this is that you can trust that the language is accurate, and appropriate for academic writing. The downside of Google Books is that it only includes English books. This is why I created another edition of the app which uses data from the Internet, covering a lot of different languages.
And this is how I wrote my thesis. With a cute little popover that gives a surprising amount of information.
What’s next? I really believe in this app’s usefulness and I wouldn’t want it to be forgotten. I’m sure it can help a lot of other students with writing their theses (or their publications, grant proposals, reports, you name it). So I’ve created a website to make it freely downloadable for others. I’d love for you to check it out and give it a try: www.writefullapp.com.
Any comments are welcome – if not about the app itself, perhaps about the frequency-approach it’s based on. After all, I can be a huge fan of my own app, but that doesn’t mean it works for everybody. (Plus I do realise I might be slightly biased.) I’d love to hear what you all think!
Author Bio: Juan is a research associate in Computer Science at the University of Nottingham, UK. He has a strong interest in psychology, and in particular in natural language processing and acquisition.