Dissertation – The first aim

Okay, with my data firmly in hand (so to speak), I’m ready to start tackling my first aim. Despite having all my data collected however, I’m incredibly behind schedule, by about four months if I check my handy proposed timeline. So it’s a race to catch up if I want to graduate on time and oh do I want to graduate on time! Then there’s the other issue, I have the DARPA Risers conference coming so it would be nice to show something new, that’s only a few months away so the race is on…

A quick intro. I’m a fifth year (almost wrote fourth!) PhD candidate in neuroengineering. My BS and MS are mechanical (two totally different fields!) and for the past year (almost exactly), I’ve been working full-time in a hospital doing research. I love my work so despite the increased stress, I stand by that choice. Early this year I successfully defended my PhD proposal, meaning the only thing standing between me and my PhD is my dissertation. As I’ve entered into the fifth year, I recognize the previous year has been a good year for me, but the first three was rough, really rough. It’s all documented pretty thoroughly though, so there’s that.

With that out of the way, we can talk about aim one (of three technically) for my dissertation.

Aim 2 and 3 are related (do one and I can do the other fairly easy… hopefully). That will require a second round of data collection, but I’m hoping to avoid it, or at the very least, shrink the number of datasets needed. We’ll see what happens with that, but there’s still a lot of work to be done so I don’t even want to think about it too much because it’s a lot to worry about and I don’t know if what I’m proposing works.

So aim 1. My entire dissertation rests on aim 1 even succeeding, which is a bit stressful. Aim 1 is the, “does this thing I think I can do actually work” aim. Meaning if it fails, the whole dissertation falls apart. We’re giving it our best shot though and the data I collected looked good as far as visual inspection can get you anyway. It all looks like squiggles until you do something with the data to make sense of it. But some squiggles look better than others and if the squiggles look correct, then at least the data is good. You don’t know if the thing you’re looking for in the data is there, but having good data is the first major hurdle to clear.

As of yesterday, I have collected data from 10 participants. If my qualifying exam data is (1) accurate, and (2) repeatable, this is more than enough to show what I’m hoping to show. The trick is we don’t actually know if the data I collected for my qualifying exam is accurate or not. On the brightside, that won’t be the case for long. There is however, a bit of work to be done and now I can talk again about my timeline.

Back in February or so (not even sure when I defended my proposal these days!), I said I would collect my data in about a month (or less) and spend approximately five months to do the analysis. That was the plan, BUT the equipment I needed took about twice as long as it should’ve to arrive. My timeline has been horribly out of order since. Now, how bad is it?

Pretty bad in fact. I had five months to do the analysis in my proposal after data collection, putting my completion of the analysis at the end of September. In my original timeline I should be about a month into writing my first paper detailing my findings (assuming it worked, there’s still not a journal of null results sadly). What does that mean for me?

Well a quick glance at the calendar tells me that we’re in September already. In fact, following my original timeline I would be collecting any needed data for aim 2 right now. Yeah I’m behind schedule. There’s still some hope, but it’s not good news at the moment. I still plan on trying to graduate on time, I just need to try to adjust my timeline accordingly. I gave myself four months for aim 2 analysis and two months for aim 3. This is probably where the condensing of my timeline will come from.

Aim 2 and 3 shouldn’t take that long, I gave myself a longer timeline for each aim to make sure I could complete it all in time and to help with delays. While this delay is a lot longer than I had originally planned, if I can work hard to get through aim 1, then I shouldn’t be too far off from finishing aims 2 and 3 from their original completion time, so there’s some hope (which is why I’m pushing so hard to get it all done).

Aim 1 was originally pretty low stakes for me. Either this works or it doesn’t. Either way I get my PhD and we learn something important about the limitations of the technology as it currently exists, which is a win in my book either way. With the DARPA Risers nomination, things feel a little more dire.

I obviously don’t want to go into a conference with my possible tale of failure while everyone around me is talking about the amazing advancements that they are making, but even if that’s the case I plan on making a case for taking chances like that even in the face of failure. I’ve had a lot of failure leading up to this event, so if it’s one thing I’ve learned it’s the best you can do is take it gracefully when it comes.

Okay, that said let’s talk about my revised timeline.

With DARPA around the corner, I have two months tops to get this done enough to show something for the event. I have slightly less time between now and the meeting with the person who nominated me (someone who works at DARPA). Basically things will need to move fast and I’ve already been planning the data processing and the analysis I want to do, so it’s not like I haven’t been putting in time on the project before the data was collected.

The next few weeks will be crucial to answering if this works and key to making sure I can graduate on time. But for today at least, I’m taking it easy. The data collection process has been exhausting and there’s no sense in trying to do data analysis if I’m not in a good place to get through the work quickly. No, instead data preprocessing and maybe (if I’m lucky) analysis will start next weekend. I don’t need to get it all done for the DARPA Risers event, but it would be nice to have most of it done. Ideally by November I’ll have all the data processed and ready for publication.

So today we rest! Then we get started on answering the big questions. Fingers crossed!