Top 5 Tips For Writing Your Thesis.

BACKGROUND

There comes a point in every Ph.D students life where they have to leave the lab and chain themselves to a desk for a couple of months for the final write-up session. This can result in a mixture of emotions, initial elation at the thought of getting everything finished, anxiety when you can't seem to write more that 50 words in a day, and sheer frustration as you try convince MS word to allow you insert a single landscape page amid a multitude of portrait orientated cousins. Despite completing my Ph.D 4 years ago I remember my own experiences of writing up as if it was yesterday. During my write-up I discovered a couple of techniques that helped me keep writing, and manage my data well. I'd like to share my own experiences of what helped me. I'm omitting the obvious things like setting targets, proof-reading, or getting your peers to review your work. Instead I will discuss the less obvious. 

TIP 1 - USE MENDELEY FOR REFERENCE MANAGEMENT

I've devoted a whole article to this topic before so I won't go into detail here. What I will say is that Mendeley is a superb piece of software for reference management. It's far more easy to use than Endnote, it's free, and it's packed full of useful features. That's pretty much it in a nutshell.

TIP 2 - TRY TO BREAK MS WORD

This sounds counter intuitive I know but I promise you this is a good idea. Allow me to explain. When I was writing my thesis I was using word 2007, which at the time was a big improvement on word 2003. I had written my undergraduate thesis on word 2003, and I ended up changing over to openoffice because it was such a hateful experience otherwise. But I was assured by my colleagues that Word 2007 was actually usable so I thought I'd give it a go. I was still wary because I intended to have a lot of images and equations in my thesis, and one of the main problems with 2003 was how poorly it responded to trying to insert/re-size images and equations. So, before I embarked on using Word 2007 at all I decided I need to know its limitations. What forces it to stop working, and in what way does it stop working? For example, here are some of these tests I did. 

• Equations: I began by entering a simple equation. I wanted to see how Word handled problems like inserting Greek letters, centering an equation in the middle of a page, or labeling it "eqn" for example. Word 2007 could handle most of this, I could use Greek lettering, I could center the equation in the middle of the page, and I could label it, but the label did not line up with the equation itself! That was pretty frustrating. Eventually, a clever colleague of mine realised that if you inserted a one row, two column table the width of the page, you could insert an equation in the left hand column, and the label in the right hand column. You could then remove the lines so it looked like it was aligned correctly. We never did find a better way of doing this. But thankfully we discovered this in the early stages of writing up so inserting equations was only slightly painful.

• Document Structure: There's a reason people use word processors. They are designed to speed up the process of writing, make editing quick and easy, and automate some of the more tedious processes. But actually most people who used Word to write their thesis did not take advantage of many of its useful functions. For example, being able to automatically generate a table of contents and a table of figures is incredibly useful. To do this you need to use section headings and you need to let word manage the naming of your figures.  It can be tricky to set up but open a new document and experiment with the section heading formatting. Check what happens to your numbering when you add or remove sections, check what happens when you add your table contents. Once you set up your document correctly you will be able to generate and more importantly update your table of contents and table of figures in a few seconds! Two friends of mine writing up their Ph.D at the same time as me never did this. This meant they were manually writing and numbering section headings. Every time they added or removed a page they were manually updating their table of contents to have the correct page number. This is not how a word processor is supposed to be used. It's extremely time consuming to operate this way, and it takes valuable time away doing what you are trying to do, writing your thesis.

TIP 3 - USE DROPBOX/GOOGLE DRIVE

This one is non-negotiable as far as I'm concerned. There are a multitude of options when it comes to backing up data, TimeMachine for Mac, or File Recovery for Windows, SD cards, USB keys, and email. Trust me, Dropbox or Google Drive are the best options for you. Personally I have a better experience with dropbox so this is what I will discuss.

To give you an idea of why cloud storage is much more effective than anything else allow me to demonstrate my computer setup. During the last year of my Ph.D I went through three laptops while I was writing up. The first two just stopped working without any warning. The screen broke on one, and the motherboard broke on the other. Laptop no.3 was a gift from my supervisor and while it worked very well, and allowed me to continue to have some mobility it did have a tiny 13" screen. This is not ideal when you often want to look at multiple sources of information. So sometimes I would work from home on the weekends and use my own desktop to avail of a much bigger screen, and sometimes I would work from my Linux desktop in work since it was to only way for me to generate data from computer simulations. This means I used a total of five computers, three of which were in regular use depending on my needs. Trying to manually sync your data between all of these would have been a nightmare without cloud based storage.

By using Dropbox all I had to do was have one folder called "Thesis" and save my active word document in there. Once I opened my word document from Dropbox and set word to autosave every 5 minutes it didn't matter if any of my computers broke. I could simply move to any other computer on the planet, login to my Dropbox account and continue to write with all my images, data, and research papers right where I left them. I didn't have to spend any time making backups, moving files to SD cards, or emailing anything to myself. I didn't have to do any version control and I never once lost any of my work, despite the best efforts of MS word. It was easy and stress free. In addition to all of the backup benefits there is one more very useful benefit. If your supervisor wants to see the latest version of your document you can simply share the relevant folder with him. He can access it anytime he wants, and any changes he makes will be saved on the version you are working on. This setup has worked really well for me. Do it!

One final word on this. I have written about the reference management tool Mendeley in the past, see here. Mendeley give you a small amount of cloud space (500Mb) to store and sync your research papers. However, I quickly went over this limit, so to circumvent this I found it was much better for me to create a Dropbox folder for research papers and save all my journal articles there. I could then point Mendeley to this Dropbox folder and it would always have access to all the relevant papers no matter which computer I used.  

TIP 4 - USE PLENTY OF IMAGES

Pretty much everyone I knew got writers block at one time or another. It doesn't just happen to professional full time authors. It happens to anyone writing on a particular topic for a long period of time. Different people have different solutions for this, but personally I found that inserting a image really helped me to re-focus my attention and clarify what I wanted to say. The problem with writing a thesis is that's it is very specific and sometimes you cannot find an image you need to describe a particular process. In this case I would actually make my own images using PowerPoint. I know this probably sound labour intensive, but actually you can make great images using very simple shapes and the more you do it the faster you will become. Also, because you are making the image if forces you to think about the process you are trying to describe, and this often leads to a re-write of the relevant text. The resulting combination of a specific image, and well written text means both you and the reader understand the topic much better.

Even if you don't make your own images it is important to utilise them to help you make your point. They should always be there to add clarity, demonstrate an abstract principal, or summarize information you have said, or are about to discuss. Well used images can really demonstrate your ability to explain and understand complex ideas. 

TIP 5 - WRITE WITHOUT RESTRICTION

As a scientist writing about peer reviewed scientific theories and concepts I found it necessary to qualify every sentence I wrote, unless I was making it clear that I was stating my own opinion. Even then my opinion must be based on fact established somewhere else in the literature. This meant that for the first few weeks of writing up I would end up writing one sentence like, "All proteins are made of amino acids of which there are twenty naturally occurring types" and then agonizing over how we/I know this to be true! I'd trawl through the literature looking for similar statements and after hours would only just about be in a position to verify this (it turns out there are actually 22 naturally occurring amino acids). However, it is just not practical to question every piece of information you think you know. After a few weeks I was writing without this kind of restriction, I trusted that the statements I was making must be based on evidence, and I was definitely going to provide evidence for all the claims I was making, but that would come after I completed that particular section. So after finishing a couple of paragraphs I would then go back over it inserting the relevant references. If I came across something I could not find any evidence for in the literature I would have to re-write it based on the evidence that was available in the literature.

Overall, by trusting that you do know what you think you know you can actually get stuck into describing the really important and heavily theoretical/abstract aspects of your work. Modifying a sentence to say there are 22 amino acids instead of 20 is really really quick! And this was something I definitely didn't appreciate before I started writing.

CONCLUSION
There are some valuable tools out there to help you write up and remove some of the unnecessary stress. Get to know the software you use. Know their weaknesses, and be prepared to spend a small amount of time at the start to prevent a lot of time being wasted later on. Writing up can be fun. It's very satisfying to close your laptop at the end of the day and know you've made good progress that will sill be there in the morning no matter what happens.