I enjoy seeing what must be completely unique sentences in the English language being born. “I'm no fMRI fanboy” just scored a record of entry. Congratulations.
Continuing to love your breakdowns on neuroscience Tommy! I've been doing a bit of lecturing recently (management, leadership, etc) and I've started referencing some of these examples as a way for undergrads to understand how to better think about research, especially as the 'social science' subjects I teach are ones that are incredibly prone to misguided interpretations of research.
It lead me to a realisation that our schools aren't teaching students how to actually think and engage with research, and there seems to be a massive cognitive blindspot on the parts of the academics within the institution that students just inherently understand how they're supposed to engage with research.
What do you think is lacking in how things are being taught? Is it just a lot of only teaching results, not about the strengths and weaknesses of methods?
1) How to actually *think* about the academic research process, and thus how to read and interpret research. Let's say I'm teaching Contemporary Management. In class, I'm teaching students about the concepts and research, and thus how to be a better manager. Then, as an assessment, students are asked to analyse a case study, and back it up with research.
The piece I realised is missing is a fundamental understanding of what exactly the research they cite is about. Students focus on (and are encouraged to) reference their work according to APA standards or whatever, but they've missed the point about whether the research is actually any good or not, how the research came about, truly understanding what the phrase: "Research SHOWS..." means.
Then, given the sheer volume of research papers in circulation, they just grab whatever research backs up their analysis and go from there.
2) This then creates a huge gap to your point about 'only teaching results'. The missing word there is 'only teaching *academic* results'.
So whilst I'm teaching students how to be a better manager or developing the skills and frameworks for better management, the assessments (where I teach) are almost all about: "Now write me an academic paper on management."
Thus instead of demonstrating their grasp of the theories and principles, they're being assessed on how well they can write an academic paper, which is a completely different skillset.
I don't know if that's just an issue at the institution where I teach or a problem more broadly across academia.
Makes sense, thanks for fleshing that out a bit. I teach in a pretty different space (teaching data science) but I see a similar problem--a pretty superficial focus on engaging with research, but not much on actually evaluating it critically or tying it directly to real industry applications (where most of these students are trying to go)
"Brain activity" seems to be a portmanteau for magnetic resonance, electric current, or blood flow depending on the chosen measurement technology. The key word in your essay is "triangulation."
Yeah there are various methods for analyzing fMRI data that take into account the spatial autocorrelation in various ways. If you don't use those methods and instead treat each voxel as a completely independent test by just doing standard multiple comparison correction, you're being overly conservative--but that's the opposite of the problem the salmon paper was pointing out.
This is one of my favourite papers.
Up there with the parachute paper.
Do you mean the joke parachute RCT? (https://pubmed.ncbi.nlm.nih.gov/30545967/)
That's great.
Yes, that's the one. Would have dropped a link but was in the lift on my phone.
I enjoy seeing what must be completely unique sentences in the English language being born. “I'm no fMRI fanboy” just scored a record of entry. Congratulations.
Continuing to love your breakdowns on neuroscience Tommy! I've been doing a bit of lecturing recently (management, leadership, etc) and I've started referencing some of these examples as a way for undergrads to understand how to better think about research, especially as the 'social science' subjects I teach are ones that are incredibly prone to misguided interpretations of research.
It lead me to a realisation that our schools aren't teaching students how to actually think and engage with research, and there seems to be a massive cognitive blindspot on the parts of the academics within the institution that students just inherently understand how they're supposed to engage with research.
That's awesome!
What do you think is lacking in how things are being taught? Is it just a lot of only teaching results, not about the strengths and weaknesses of methods?
There are two broad issues:
1) How to actually *think* about the academic research process, and thus how to read and interpret research. Let's say I'm teaching Contemporary Management. In class, I'm teaching students about the concepts and research, and thus how to be a better manager. Then, as an assessment, students are asked to analyse a case study, and back it up with research.
The piece I realised is missing is a fundamental understanding of what exactly the research they cite is about. Students focus on (and are encouraged to) reference their work according to APA standards or whatever, but they've missed the point about whether the research is actually any good or not, how the research came about, truly understanding what the phrase: "Research SHOWS..." means.
Then, given the sheer volume of research papers in circulation, they just grab whatever research backs up their analysis and go from there.
2) This then creates a huge gap to your point about 'only teaching results'. The missing word there is 'only teaching *academic* results'.
So whilst I'm teaching students how to be a better manager or developing the skills and frameworks for better management, the assessments (where I teach) are almost all about: "Now write me an academic paper on management."
Thus instead of demonstrating their grasp of the theories and principles, they're being assessed on how well they can write an academic paper, which is a completely different skillset.
I don't know if that's just an issue at the institution where I teach or a problem more broadly across academia.
Makes sense, thanks for fleshing that out a bit. I teach in a pretty different space (teaching data science) but I see a similar problem--a pretty superficial focus on engaging with research, but not much on actually evaluating it critically or tying it directly to real industry applications (where most of these students are trying to go)
"Brain activity" seems to be a portmanteau for magnetic resonance, electric current, or blood flow depending on the chosen measurement technology. The key word in your essay is "triangulation."
Is spatial autocorrelation between voxels playing an important role here?
Yeah there are various methods for analyzing fMRI data that take into account the spatial autocorrelation in various ways. If you don't use those methods and instead treat each voxel as a completely independent test by just doing standard multiple comparison correction, you're being overly conservative--but that's the opposite of the problem the salmon paper was pointing out.
Hoel strikes me as an angry arrogant idiot. He might be a very nice person, but his public schtick always seems "Everyone is stupid except me."