Crowdsource question about pivoting F31 aim

[jdawg2017]

Hi all,

I am trying to get some (anonymous) advice about pivoting an F31 aim. Apologies if parts of this seem vague and if this post is a little long, but I really appreciate any thoughts. I am going to talk to my mentor about this, but they are not the best at communication and I think are strongly trying to steer me more toward a certain career objective I am increasingly not wanting to do, so I figuring crowd-sourcing some other advice and thoughts might be helpful.

As we all know, the last year with the pandemic has been a mess for researchers across the board, including me as a trainee. I am going to give my dissertation prospectus soon, so I want to know what my game plan is for all of the studies.

I had originally planned to use a task I had developed with my mentor to look as a certain aspect of cognition and its relation to task-based fMRI BOLD activity. It's a simple task behaviorally, but a nightmare to code for an fMRI task. This is especially compounded by the fact I have been WFH since the pandemic began and that many workshops/classes I had intended to attend last summer/fall got cancelled or moved online to a time that did not work with my schedule. I am, in general, someone who learns a lot better in-person with hands-on examples, so learning coding/fMRI design and analysis has been consistently challenging now... In addition to challenges on my end, there is a very real chance if the pandemic goes south again with variants, people not getting vaccinated, etc. that we may not be able to scan participants (or too few before the time I intend to defend my dissertation) given their inherent health vulnerabilities and other challenges. So, this alone impedes my motivation, in addition to not having the support I thought I would have when I applied for the F31 and was awarded in pre-pandemic times.

For context, I have been gathering via Zoom research visits reliability/validity data on the task as a behavioral measure in healthy adults, a project that is almost done. I had intended to add this as a dissertation paper anyways, since it would have supported the fMRI work. I had intended to pass to another student the study of this measure behaviorally in our clinical population of interest, also over virtual visits d/t the pandemic and other considerations, but I am not sure if the other student is really even interested. As part of my dissertation/F31, two other papers are going to use extant data we and large, publicly available datasets have (i.e., retrospective).

In sum, I am wanting to basically abandon this fMRI task adaptation and data gathering of the behavioral measure and focus on other papers that I can for sure do, including the healthy adult and clinical population data on this new measure (likely two papers), and the two other F31 aims using extant data as proposed.

My main questions are:

1) What are the implications for not completing 1 of my 3 F31 aims at NIH?

2) Are there any impacts on my future grant applications, or my mentor's reputation, assuming that I actually complete the other aims?

I would like to learn more about neuroimaging task design, truly, but accomplishing my dissertation and going off to internship is WAY more of a priority for me.

Thanks in advance for any comments/advice!

---

Members don't see this ad.

 

[Ollie123]

1) Talk to program, but I imagine implications will be relatively non-existent. Right now they are concerned with investigators who have 20 million dollar P01's and can now complete zero aims. I can't imagine hitting 2/3 on an F31 during a pandemic is a big deal. I'd ask about "switching" aims as this is the more common approach for other grants (versus simply not doing one). I had to switch things up some on my K23 and program could not have cared less.

2) I imagine consequences to be minimal to either of you (if present at all), besides the hopefully obvious fact that you won't be ready to compete for larger fMRI work without it.

That said, while there are always exceptions I wouldn't worry too much about starting a scanning study right now. Tons of scanning studies are being run right now. Given we're less concerned about surface transmission at this point, risk actually seems like it would be lower than many other types of research as long as protocols are in place. We've been back up and running since August without incident and I know others in the same boat. It might be <slightly> different if you are at one of the handful of places with a scanner actually in the psych department where you might not have as much capacity or experience sterilizing the heck out of things like a traditional hospital radiology setting would. If your interests have shifted and you are just less excited about scanning, that is another matter.

Feel free to PM me if you want to discuss in more detail. I admit I'm a bit puzzled by the programming challenges since adapting a behavioral task for fMRI is normally pretty straightforward. Is it an issue of the task length being variable so you need to feed task info back to the scanning computer to modulate the sequence? That can be a nightmare and most of the time we usually just change the task to avoid having to do it (though in some cases that is admittedly not possible). This is very much in my wheelhouse so I'm happy to offer pointers if I can. I mostly use E-prime, so I'm less useful with actual coding if you are using PsychoPy or PsychToolbox but am happy to help think through things at a conceptual level.

 

[jdawg2017]

1) Talk to program, but I imagine implications will be relatively non-existent. Right now they are concerned with investigators who have 20 million dollar P01's and can now complete zero aims. I can't imagine hitting 2/3 on an F31 during a pandemic is a big deal. I'd ask about "switching" aims as this is the more common approach for other grants (versus simply not doing one). I had to switch things up some on my K23 and program could not have cared less.

2) I imagine consequences to be minimal to either of you (if present at all), besides the hopefully obvious fact that you won't be ready to compete for larger fMRI work without it.

That said, while there are always exceptions I wouldn't worry too much about starting a scanning study right now. Tons of scanning studies are being run right now. Given we're less concerned about surface transmission at this point, risk actually seems like it would be lower than many other types of research as long as protocols are in place. We've been back up and running since August without incident and I know others in the same boat. It might be <slightly> different if you are at one of the handful of places with a scanner actually in the psych department where you might not have as much capacity or experience sterilizing the heck out of things like a traditional hospital radiology setting would. If your interests have shifted and you are just less excited about scanning, that is another matter.

Feel free to PM me if you want to discuss in more detail. I admit I'm a bit puzzled by the programming challenges since adapting a behavioral task for fMRI is normally pretty straightforward. Is it an issue of the task length being variable so you need to feed task info back to the scanning computer to modulate the sequence? That can be a nightmare and most of the time we usually just change the task to avoid having to do it (though in some cases that is admittedly not possible). This is very much in my wheelhouse so I'm happy to offer pointers if I can. I mostly use E-prime, so I'm less useful with actual coding if you are using PsychoPy or PsychToolbox but am happy to help think through things at a conceptual level.

Thanks for your advice and insight! This makes me feel a lot better about the situation. I know in the end it will all be okay and that the world keeps on going, but it's reassuring that probably the biggest implications would be future grants on imaging (which I am totally okay with, given that the pandemic kind of pushed my training and interests in an entirely different direction anyways).

No need to go into too much detail about the task I think RN but basically it was not originally designed in e-prime or psychopy (huge mistake) but another software entirely which makes taking it to these software a nightmare. The timing is entirely off (stimuli presented for too short of a period for BOLD in the behavioral version; also the timing not set well to being easily broken into typical run lengths) and we would change a lot of the stimuli anyways, so unfortunately besides having a task concept the design process is pretty much starting from scratch. I talked with a former post-doc who was kind of stumped about what to do (and they have a lot of task building experience in general) so that raised a red flag for me lol.