Data analysis

A French writer called Jean de Lafontaine once wrote a fable about an oak and a reed: the oak makes fun of the reed for being shaken by the winds, and so frail and vulnerable. The oak feels indestructible. Then, a huge storm arrives and takes the oak down, while the reed bends, but does not break.

This story is a tale of resilience.

oak and reedResilience is a skills that I have tremendously improved in the last few months, the months of the dreaded (dramatic horror movie jingle) “data analysis”.

The bulk of my PhD data being quantitative, I ended up with massive data files from an online survey I had launched earlier this year. Although dealing with quantitative data is very different from qualitative data analysis, I think there are similarities in the processes, so hopefully everyone can relate to this story.

So, data in hand, around May-June, I figured: “How hard can it be to analyse all this? you just have to know which statistical analysis to run, learn how the software works and pretty much do it. This should be done by the end of the summer”. Such wild misconceptions on the ease of the task ahead have been common in my PhD journey. Let me tell you practically why data analysis was such a high mountain for me to climb.

First, because when you perceive a task to be hard, it is really easy to beat around the bush for ages before actually getting started. I had done data analysis in the past, but I had become highly unfamiliar with the first software and procedures I had to employ, and had to learn a whole set of new techniques and second software on top of that. Resources to help me where multiple and readily available, but I hard to pretty much teach myself to do it, which proved a long and tedious process. I tend to be stimulated by new and unknown tasks, and I love learning, but this time, it just seemed “too much”. I went slower than I normally do.

Once you get the ball rolling, you think it’s going to go smoothly. You have a fair grip of the statistics you need to run and how the software work. Things are in control. And then, you hit a wall: the data is not performing as you intended. The stats are wrong. It’s a disaster. You go back, redo everything twice, tweak your approach, iterate, modify…and most importantly, stare at your screen for hours, binge eat, freak out, lash out at your friends, and on the most glorious days, wake up at night thinking about it.

Along this mind-wrecking journey, which is still in progress, I have learned several things:

Make decisions: it’s fine to try things out, iterate, try and find the very best solution or approach, but at some point, you just need to stop, focus, and make a decision.

Seek advice: don’t remain stuck on a problem you cannot solve. Asking a quick question to someone and getting feedback will put you back on tracks.

Keep going…: even if today is not going to be the most productive, stay engaged with your analysis. Don’t leave your data hanging for a week or two without touching it, you will forget where you were and have to start again.

…but take breaks: for sheer sanity purposes. Taking a short break from your data will make you come back to it in a better state of mind and with a fresh pair of eyes.

Then one day, you’ll get past the hump and be done with it.

In the storm of data analysis, be like the reed: bend, do not break!

And you, what did you learn with your data analysis?

9 thoughts on “Data analysis”

  1. My question to you, as someone who’s recently been admitted as a PhD candidate (passed my MRes with distinction BTW! – woo!) is, what’s so hard about it? I don’t get it…

    1. Hi Laurence! First of all congratulations on your distinction and official admission, that’s great news! I guess what made it so hard was the fact that I have just really bad at statistics to start with 🙂 the more you build up the idea that something is hard in your head, the more it actually turns out to be!

  2. Thanks Lau, I totally agree with you and thanks for the tips, I will remember and use them for the writing up process as well! My data is qualitative and it was no easier job. Just about getting there and got lost in the woods a few times. The metaphor of not seeing the wood for the trees really comes to mind. The sheer amount of it made my head spin! But getting organised through well structured tables based on a good conceptual framework is what saved me in the end. And getting great advice from both supervisors – and a very kind someone further ahead in the process – saved my sanity (he will know who he is…).
    I think that we all start in year one thinking that a PhD is just going to be a longer version of a Master’s dissertation and how wrong can one be honey! I also graduated with distinction in my MSc and it made no difference; or maybe that’s why I am still here. Good luck to the other Laurence with her PhD. I think we all eventually ‘get it’ through experiencing the journey (and am I glad I decided to do this!) but there is no one size fits all; it’s a very personal journey for sure.

    1. Hi Carole! Thanks for the read. I am glad you enjoyed it and could relate to it. I think all types of data can be extremely confusing and involve a lot of back-and-forth. Supervisors are definitely of great help, but as you say, other people can also shine a light on the whole process: one should not hesitate asking for help and get different advices. And definitely, a PhD is not just a longer version of a Master’s dissertation; no way! Take care, analyse away and I will see you very soon in the office 😉

  3. The PhD journey is a real emotional roller-coaster of a ride. Getting a distinction in the MRes is a great start but if you romp through your PhD perhaps it is not a PhD at all! On the other hand, it is not a nobel prize either! Your blog post is really good.:)

    1. Thanks for reading and for your kind comment, Marian. You are absolutely right, it is an emotional roller-coaster: some parts are extremely stressful, others utterly enjoyable. I guess this is part of every job and you must experience ups and downs as a professor too. Hopefully maturity and experience help 🙂

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s