Hi all 🙂
The last two months have brought some analysis of preliminary results, my second departmental presentation, and a holiday which couldn’t have been better timed.
If you want to go present at a conference you need something to present. In general, this means data and with data comes stats. Stats brings anxiety to many people and, whilst I thought I was making progress, I was no exception. I came up against multiple problems and some seemingly simple questions I had I couldn’t find any information online. (I should add here that yes my supervisor is great and walked me through the things I was stuck on but I really wanted to give the full picture a shot on my own first). Combined with a few very long days, some sleep deprivation, and a general sense of being fed up I had a bit of a break-down; feeling defeated, it was the data which broke the camel’s back.
A little down the line and after a rest I went through the analysis with my supervisor and found a tentative hit! I had a significant result! Whoop! Conference here I come! The result in question was an interaction between condition and the sex of the individual but the important next question was what was that interaction? To illustrate what I mean I’ll provide an example.
Imagine that you’re interested in whether being stressed affects memory. So you give a group of people a list of words to remember twice: once relaxed and once when you’re putting the pressure on. You then ask them to list all the words they remember. Now you may expect a difference based on the situation (condition) but would you expect them to differ based on their sex? A third option is that both could happen in different ways. For example, males may perform better under pressure whereas females may perform worse or any other different response you can imagine. This is called an interaction because the results depend on not only the individual factors but how they combine. Now the stats tells you that you have a significant effect of condition combined with sex, but what is that interaction? Do females get worse under pressure but males don’t change? Do males get worse and females get better? To work this out you perform post-hoc tests which identify the differences between pairs of measurements (females stressed v female relaxed, female stressed v male stressed,…). To cut a long story short, when you do this you raise the threshold at which you consider the results significant, in essence making it more difficult to find one but leaving you more sure that it’s a real finding, and in doing so my significant result disappeared. “Damn you Bonferroni!” I exclaimed at my desk.
Anyways, whilst that was a rather simple tour of my recent stats, no conference for me. I’m nowhere near finishing my experiment though so any exciting results were quite unlikely.
Each year every PhD student in the department gives a talk about their work. My last one wasn’t great; it was poorly timed, not very well constructed, and I wasn’t able to deal with questions well. One of the biggest lessons learned was: practice it! This time I arranged to give the talk to my lab group before for some feedback. Suffice to say, that didn’t go well. I mean, it went well regarding getting constructive feedback but I started with quite a messy presentation. But that’s why we practice! I took the feedback, simplified the slides, refined a few ideas, and added in several more slides to make points clearer. It was a good exercise and made me think harder about what are the real take home messaged for each part of the research.
Then it came to the actual presentation. I was quite nervous. I felt jittery and ended up grasping some white-tac whilst I spoke as comfort. Once I started though I felt a lot more comfortable. I went in telling myself “you are the expert in the room” and that mentality made me flip from feeling like it was a performance I was being judged on to being someone who is essentially teaching a collection of interested viewers. That settled me a lot. Overall the talk went well, I was only about a minute over, and I had very interesting questions which I was told I handled incredibly well. The only negatives I received were that my background was a little long (for someone from my lab group) and to avoid having something to fiddle with! Curse that distracting white-tac…Anyways, it was a good boost to my confidence both in my knowledge but also my presentation skills. Sadly I didn’t have any data so I’m eagerly looking forward to the next one when I can actually tell my story!
By the time I put this out I’ll have been on holiday for a while and it could not have come soon enough! Suffice to say that the last few months have just been getting harder. I’ve felt quite defeated and deflated and that is in no small part due to the fact that I haven’t had a solid break in almost a year. I don’t have a great deal to say about this to be honest. We all need breaks and sometimes what we do make that difficult. For me it’s a combination of having to work PAYE to pay my bills and having an experiment I can’t take much time away from. But it does slowly exhaust you and I’m pretty sure that even now when I’m feeling rested that I’m still fatigued. I’m off to South Africa to see Cape Town and then going to Ulusaba for the safari. It’s going to be amazing I’m sure and if it’s anything like my holiday last year I’ll be totally relaxed by midday on day one! For now though (when I’m writing this) I’m just counting down the days until it begins.
So that’s me. The next few months will involve enjoying this holiday, resting up, and finishing the last 30% of my testing (and seeing a certain Avengers movie).
Thanks for reading ☺