David S. Fink
@davidfink.bsky.social
Epidemiologist focused on society, drugs, and mental health.
Let’s just say a 5% false positive rate, which would be pretty low I imagine…could you imagine being one of those people a computer looks at and says will die?
May 9, 2025 at 12:18 AM
Let’s just say a 5% false positive rate, which would be pretty low I imagine…could you imagine being one of those people a computer looks at and says will die?
These two issues seems totally different to me. Sure, more money for science can do a lot of good, it can pay for replication research, it can create different incentive structures to allow more patience, but it can’t fix challenges of interdisciplinary conflicts or measurement. At least to me.
April 25, 2025 at 1:08 PM
These two issues seems totally different to me. Sure, more money for science can do a lot of good, it can pay for replication research, it can create different incentive structures to allow more patience, but it can’t fix challenges of interdisciplinary conflicts or measurement. At least to me.
That’s actually not why. The reason why is that science is really hard. There are times when biologists work in one unit of measurement and chemists use different measures. In my field, epidemiologists and economists call the same method something totally different. Interdisciplinary science is hard
April 25, 2025 at 1:06 PM
That’s actually not why. The reason why is that science is really hard. There are times when biologists work in one unit of measurement and chemists use different measures. In my field, epidemiologists and economists call the same method something totally different. Interdisciplinary science is hard
I’m having trouble seeing the link between this comment and the discussion on measurement challenges and identifying meaningful findings versus “significant” findings from large data sets.
April 25, 2025 at 12:58 PM
I’m having trouble seeing the link between this comment and the discussion on measurement challenges and identifying meaningful findings versus “significant” findings from large data sets.
How do we overcome this problem? Anyone who has worked with administrative data knows that this is real. We often find small results that are very precise. But, even a small result on a population level could affect millions of people, so it isn’t unimportant if true. So what do we do?
April 25, 2025 at 11:05 AM
How do we overcome this problem? Anyone who has worked with administrative data knows that this is real. We often find small results that are very precise. But, even a small result on a population level could affect millions of people, so it isn’t unimportant if true. So what do we do?
Reposted by David S. Fink
Just putting this here so we can use it when it happens.
March 26, 2025 at 4:57 PM
Just putting this here so we can use it when it happens.
One of the only NECESSARY components of an addiction and any mental health disorder for that matter is functional impairment, which isn’t a given for billionaires or uber-wealthy because their wealth actually insulates them from that impairment.
April 20, 2025 at 1:20 AM
One of the only NECESSARY components of an addiction and any mental health disorder for that matter is functional impairment, which isn’t a given for billionaires or uber-wealthy because their wealth actually insulates them from that impairment.
They are using the wrong term for what they are trying to communicate. Most would call it standardization, right? It’s the same as plotting the avg price of milk over the past 50 years in 2025 dollars to account for inflation. I assume the authors are attempting to account for demographic shifts.
April 19, 2025 at 1:25 AM
They are using the wrong term for what they are trying to communicate. Most would call it standardization, right? It’s the same as plotting the avg price of milk over the past 50 years in 2025 dollars to account for inflation. I assume the authors are attempting to account for demographic shifts.
Not exactly. I worked for the VA as a researcher from 2007-2010, then the US Army and National Guard. Now I’m at Yale and have VA approval to work with data for research.
April 9, 2025 at 8:23 PM
Not exactly. I worked for the VA as a researcher from 2007-2010, then the US Army and National Guard. Now I’m at Yale and have VA approval to work with data for research.
The VA EHR is an amazing data source for understand what is really happening in the patient populations. Like any administrative dataset, it has its downsides (patients aren’t directly asked about use too), but there is a lot of great work that can be done in these data.
April 9, 2025 at 6:46 PM
The VA EHR is an amazing data source for understand what is really happening in the patient populations. Like any administrative dataset, it has its downsides (patients aren’t directly asked about use too), but there is a lot of great work that can be done in these data.
This is the problem with the science of policy making, we don’t do enough work with the policy makers and the public to learn how to scale it up. Americans have been taught that “hand outs” are bad-lazy scamming people get hand outs. This is what needs to be changed. Cost effectiveness isn’t enough
April 9, 2025 at 5:40 PM
This is the problem with the science of policy making, we don’t do enough work with the policy makers and the public to learn how to scale it up. Americans have been taught that “hand outs” are bad-lazy scamming people get hand outs. This is what needs to be changed. Cost effectiveness isn’t enough
Complex systems are incredibly difficult to understand and even more challenging to predict. Feedback loops aren’t always obvious, threshold effects aren’t known before hand. It would have been hard to predict had they tried to model this, and I feel confident they didn’t even try
April 9, 2025 at 1:42 PM
Complex systems are incredibly difficult to understand and even more challenging to predict. Feedback loops aren’t always obvious, threshold effects aren’t known before hand. It would have been hard to predict had they tried to model this, and I feel confident they didn’t even try
The real question: would you rather be an American Idiot or a 21st Century Digital Boy?
April 8, 2025 at 1:37 AM
The real question: would you rather be an American Idiot or a 21st Century Digital Boy?
Funny, I do the same thing. With the sorting algorithms, I live in a liberal bubble, so I’ve noticed going to the local news comments I can get a tap on the range of view pointe and I’ve definitely noticed the same thing.
April 4, 2025 at 6:40 PM
Funny, I do the same thing. With the sorting algorithms, I live in a liberal bubble, so I’ve noticed going to the local news comments I can get a tap on the range of view pointe and I’ve definitely noticed the same thing.
Not too long ago, I got in a discussion with a biologists who said people are over reviewing papers. In his words, the goal of peer review is to determine if the paper advances the science enough to merit publication, which should be done in about a 2 paragraph response. I see SOME merit in this
April 4, 2025 at 11:32 AM
Not too long ago, I got in a discussion with a biologists who said people are over reviewing papers. In his words, the goal of peer review is to determine if the paper advances the science enough to merit publication, which should be done in about a 2 paragraph response. I see SOME merit in this
Very nice paper! Like most things, we agree!
March 31, 2025 at 1:35 PM
Very nice paper! Like most things, we agree!
There is also an alternative, they don’t put out a call, but they actually pay researchers to ask a specific question they are interested in, would that be “okay”? Or do you think that leads to poor research too?
March 31, 2025 at 1:25 PM
There is also an alternative, they don’t put out a call, but they actually pay researchers to ask a specific question they are interested in, would that be “okay”? Or do you think that leads to poor research too?
Can I agree and disagree. There is actually a reasonable case for asking scientists to use data, particularly for descriptive work. For example, if I’m collecting data to study adolescent drug use and I created a good representative data set, why not encourage people to publish these statistics?
March 31, 2025 at 1:22 PM
Can I agree and disagree. There is actually a reasonable case for asking scientists to use data, particularly for descriptive work. For example, if I’m collecting data to study adolescent drug use and I created a good representative data set, why not encourage people to publish these statistics?
On a personal levels, I’m interested in whether the k99s are being rolled over to the R00 phase. Trying to get that answer from my PO
March 29, 2025 at 2:18 PM
On a personal levels, I’m interested in whether the k99s are being rolled over to the R00 phase. Trying to get that answer from my PO
I was explaining why the money from the government have nothing to do with the cost of tuition. Most federal funds the university received goes to funding research, not teaching. They are complete different revenue streams. If I the funding went to paying profs, then it would be tied to tuition
March 28, 2025 at 11:03 AM
I was explaining why the money from the government have nothing to do with the cost of tuition. Most federal funds the university received goes to funding research, not teaching. They are complete different revenue streams. If I the funding went to paying profs, then it would be tied to tuition
This funds research. Professors at universities need to cover their salary. They do this through research, teaching, clinical care or some combination. I don’t teach nor do clinical care, so my salary is 100% covered by research grants and I do research 100% of the time. Hope this helps. Pls ask ?
March 28, 2025 at 1:25 AM
This funds research. Professors at universities need to cover their salary. They do this through research, teaching, clinical care or some combination. I don’t teach nor do clinical care, so my salary is 100% covered by research grants and I do research 100% of the time. Hope this helps. Pls ask ?
Yes, sorry, that was one of my reasons for the email too because all funding for travel and such has been frozen at my university. I was told the executive committee was already discussing how to handle the issue. I hope they take the hit and provide remote access at normal cost of registration
March 27, 2025 at 1:46 PM
Yes, sorry, that was one of my reasons for the email too because all funding for travel and such has been frozen at my university. I was told the executive committee was already discussing how to handle the issue. I hope they take the hit and provide remote access at normal cost of registration