The dissertation foundation work

Everything needs to have a good foundation, from homes to education, if the foundation is awful, it’s all going to collapse on itself. Science is no different. Especially when we’re working with data that you can’t make sense of at a glance. Without the proper checks, you can rapidly set yourself up for failure. So today is all about building that foundation. And with DARPA and my next presentation coming up, we need to at least get the foundation done.

Time is not on my side, but when is it ever? I mean the whole blog can be summed up with, “AHHHHHHHHHHHHH!!!!!!!,” but also the constant lack of time. I’m not sure why that is, but here we are. As a fun reminder (/sarcasm) I’m still months behind on my original dissertation timeline. Right now I should be preparing my first manuscript for publication, getting data for the second aim (and by extension the third), and life would be great right now because I would have all the data and all the analysis done for the DARPA conference.

Unfortunately, I do not live in that universe and in this one, we’ve just barely finished data collection (okay, about a month ago). I’ve also finished the data pre-processing, got it all grouped together the way I want, and now I can do the group analysis that I really want to do. I’ve already started the first few steps with that. Using a similar framework I used for the last EEG paper I published (this one) I’ve got a lot of the data in the format that I want so that I can recycle a lot of the code for that project (which is what I’ve been doing now).

I’m hopeful that I’ll be able to finish a good chunk of the groundwork over these two days and while today is mostly over already, I have already started making decent progress. So what kind of foundation do we need for this? Mostly we need to know if there were any differences between conditions within participants (a fancy way of saying if we see something different while the same person is performing different tasks) and then we need to see if that finding holds across participants (mostly). That’s the basics of what needs to happen.

So you may have noticed the (mostly) above, that’s because maybe it doesn’t work for everyone. That’s okay as long as it works for most. That will be the key here. It doesn’t need to work for everyone and it may not, but we collected so much data for that very reason, to give us the proper statistical power to say that this the stuff we’re looking for is working or it is just noise. Now obviously I’m hoping we get the it’s working solution instead of finding out that I’m the victim of statistics. But we won’t know until we take a peak and that’s going to be the focus I think for my next presentation and probably for the DARPA stuff.

It’s not the big “groundbreaking” advancements school PI would probably want me to display when the time comes, but I want to have a solid foundation, the groundbreaking stuff can happen afterwards if everything is working. And by groundbreaking I just mean stuff that’s cool to show or visually appealing like the supplementary videos from the last paper I linked to above. Those were cool to look at even if you’re not someone in the field. The stuff I’m going to be showing is unfortunately, by comparison, boring to look at and very specific to people who know what they are looking at.

It’s okay because I’ll be there to explain, that’s not an issue, but it also means that I won’t get too many “wows” from people either. And to be fair between now and the DARPA deadline anyway I may find time to make something that is “wow,” once again I need to qualify that with, “if this works.” So while the figures will be very basic, at least I can go in with some confidence to answer the question, “does this work” and that’s the most important thing to me because that will dictate what happens next.

With that I should probably get back to doing the work that needs to be done. If I feel very productive I may even know tonight if I have something or not with the group data. It’s some tricky code that I need to write, so it may not happen. But if the coding gods take pity on me, well then anything could happen! Just so everyone knows, if the work goes well, I’m going to credit it to the new monitor.