So, still waiting to hear from some fantastic institutions who may or may not offer me jobs. The nervous tension is killing me, but hey, I’m still channeling it into some pretty high productivity levels, so that’s good.
Given there is a possibility that I may start my own lab someday, I’m trying very hard to nail my own way of doing things down to ensure I can start things out using the best practices possible. In conversations with senior scientists, one of the things I find that always comes up is tackling inertia to change practices. Many people are open science advocates in principle, I find- but once the infrastructure to do things in a certain way is there, it’s much harder to change.
A good example of this is the SAS vs R thing*: both packages come with steep learning curves, and for the casual stats user**, let’s be honest, both can do the job equally well. R is preferred by the open science community for a number of reasons- not the least of which is it’s free. But if you’ve got money for a SAS license, it’s already installed on your computer, you feel reasonably competent at scripting what you need to script in SAS, you’ve got a colleague who teaches the “Intro SAS for Grad students in organismal biology” in your department, even if you have the best intentions, there is a heck of a lot of infrastructural inertia to overcome to try something new (and often, if someones insists you should do [task x] in R instead, you may regard them as an annoying time waster). This is why I feel it’s very important for me to be thoughtful about building my science house’s foundation on open practice.
I’ve got a new paper that I’m preparing to submit, and I’ve just about got all the bits wrapped up. Oh, the bits.
For this paper, I’ve decided to put my money where my mouth is with regards to the whole Open Science thing, as much as possible. I have a fully scripted analysis, right down to scripting the stacking of the multi-paneled figures together. I’d originally planned to provide the code and data as supplemental docs when I submitted (as per policy in the journal I’m planning on submitting to), but I took it a step further and put it on github and figshare. Honestly, I feel a bit naked, I’m emotionally invested in this paper and here I’ve put it ‘out there’ without first getting to do the victory dance of acceptance. I’m trying to get over that- but I wanted to be straight with you- it’s not always easy.
There was a few non-starters in the making of this manuscript, mostly related to my learning to use github.*** Also, I should note that this is not going to be a 100% open science piece of work- the journal I’m planning on submitting to is subscription-based. There are a variety of reasons for that, but the core one is that the journal’s scope really fits the work.**** So, as you can see, this whole thing is a process.
I’m pretty excited about this paper. I feel like it is a quirky, interesting study that represents something truly novel. Wish me luck in the review process!
* Yes, this is a thing. A THING. You can actually gauge how receptive people will be to your open science advocacy by asking them what their preferred statistical application is.
** i.e., the grad student who just needs to do an ANOVA so s/he can get a p-value and write her/his danged thesis- but many are in this situation
***Namely, I committed without pushing and I pushed without committing. I’ve resolved this issue by using what I call the “Dr. B Brute Force Method” which primarily consists of mashing things until they work. I’m not sure how reproducible that is, but I’m working on making it so.
****Also, it is an established journal with a respectable impact factor. This is important at my career stage, so I’m really hoping it sticks.