I think the implications of what you describe for depression treatment are a little more hopeful than you make it sound. It’s a wonder we find evidence anything helps at all given how incredibly crude our standard measures are. If you conceptualize “depression” as the functional impairments described by the client, there’s much more cause for optimism. We know a lot about modifying behavior! And of course that's much easier to measure.
Indeed. If treatments make the symptoms better, it doesn't really matter if there is an underlying 'condition' that is causing the symptoms or not.
Many psychological diagnoses seem to be really a shorthand way of describing clusters of symptoms that often go together so are nothing more than labels. It seems, however, that people find reassurance in this. Hence all the articles you see written by people saying things like ' I didn't understand why I couldn't concentrate on work / finish anything but I now know its because I have ADHD and am much happier' when really all 'having ADHD' means is (potentially) that you can't concentrate on work or finish things. This doesn't mean ADHD doesn't exist it just means it could be a label for a set of behavioural traits and not a cause of them.
100%. The vast majority of people diagnosed with depression or anxiety conceptualize their condition with the same diagnostic literalism that holds back clinical research. I think you’re right that most people find comfort in a sort of diagnostic essence account -- e.g., it reduces stigma and self-blame. But there are also huge downsides to people having simplistic understandings of what’s going wrong. Effective therapy helps people understand the mental and behavioral processes that maintain their condition. It’s harder for people to understand those processes if they’re committed to seeing themselves as having some sort of permanent structural brain defect (rather than a vulnerability to falling into particular dysfunctional processes).
Oh, and I recall from Maths lessons an aeon ago that the first, classic, example of a Poisson distribution was the number of Prussian cavalry officers who die each year from being kicked by a horse.
"Bortkiewicz graduated from the Law Faculty in 1890. In 1898 he published a book about the Poisson distribution, titled The Law of Small Numbers.[1] In this book he first noted that events with low frequency in a large population follow a Poisson distribution even when the probabilities of the events varied. It was that book that made the Prussian horse-kicking data famous. The data gave the number of soldiers killed by being kicked by a horse each year in each of 14 cavalry corps over a 20-year period. Bortkiewicz showed that those numbers followed a Poisson distribution. "
Thanks for sharing. The Prussian horse-kicking data is both helpful to understand Poisson and a funny turn of phrase. Is it Prussian horse-kick deaths (People killed by horse kicks in the State of Prussia)? Or Prussian-horse kick deaths (Prussian horses doing the kicking)? Just so long as it's not Prussians kicking horses!
Interesting as always. I am not sure the temperature analogy quite works though. Clearly any instrument turns something we can't directly see (temperature, electric voltage, blood pressure) into something we can (height of mercury in a tube, position of a needle on a dial, etc) via a physical effect that has a theoretical basis for its validity. In the case of temperature its two theories - the laws of thermodynamics (I think its the zeroth law ?) that says that in thermal equilibrium the gas and the thermometer will have the same temperature (roughly) and the law underpinning the co-efficient of thermal expansion that tells us that gives us a calibrated scale.
So my point is that in the case of depression there is no theoretical underpinning of the measurement scale . So we have no basis for determining if the scale of the measured score bears any relation to the scale of any underlying property.
blame Stuart, it was his stupid analogy (this is an interesting and thoughtful comment that deserves a better response but it's late and I'm going to bed. Glad you enjoyed the episode!)
Thank you for another wonderful show. As a practicing psychologist, I would add that another problem with the self-report questionnaires you mentioned in your episode is that they don’t contain symptom validity measures. In today’s world where most people are familiar with depression symptoms (or can easily find them listed online), it is important to know whether the person completing the questionnaire is exaggerating or minimizing their symptoms, or being consistent in their endorsement of the questionnaire items. There are questionnaires that contain symptom validity measures (the MMPI-3 is an example) but such measures are lengthy and tend to be underutilized in both research and therapy settings.
(Grew up with Celsius and switched to Fahrenheit only after moving to US for grad school, so have a lot of experience with both)
Fahrenheit is superior to Celsius for everyday use (at least in non-arctic climates), and this is the hill I will happily die on
1) it's a true 0 to 100 scale. 0F is very cold, 100F is very hot, and this is the range that covers most people's everyday experience.
2) it almost entirely eliminates the conversation: "-what's the weather like?" "it's 5 degrees." "plus or minus?" which happened ALL THE TIME when i lived in Northern Europe (where its around 0C for a significant proportion of the year). With C the only way to avoid it is to always specify plus or minus but that is an extra word and a lot of people chose not to say it. With F you can have the same problem, but in the range that is much more rare, and, frankly, difference between -5F and 5F is much less important.
3) 1 degree of C is just a bit too much. As a result, a lot of the time you need to think of it in whole degrees even for everyday stuff, and certainly can't just round to nearest 10 (difference between 5 and 14 or 11 and 19 is massive). With F, most of the time the first digit of the temperature gives all you all the info you need to know.
Basically, it is almost perfectly tuned for the temperatures that people actually use. And guess what, scientists, always switch to the temperature scale that is tuned for their specific need, whether it is Celsius, Kelvin or electronvolt.
1) Fahrenheit is a legitimate scale and and partisan adherence to Celsius is arbitrary and silly.
2) Aren't there some diseases where the response to a cure is part of the diagnosis? My sister has lupus (I guess) and she told me part of the diagnosis is that the symptoms improve after certain prescriptions. Could that not be the same for depression? The fact that a course of treatment is successful could also be a diagnostic tool?
(Water boils at 212 and freezes at below 32. It's not that hard, Stuart. If you measure angles in degrees instead of radians you must confess your sins!)
What a great show! Thanks so much for doing this.
I think the implications of what you describe for depression treatment are a little more hopeful than you make it sound. It’s a wonder we find evidence anything helps at all given how incredibly crude our standard measures are. If you conceptualize “depression” as the functional impairments described by the client, there’s much more cause for optimism. We know a lot about modifying behavior! And of course that's much easier to measure.
Indeed. If treatments make the symptoms better, it doesn't really matter if there is an underlying 'condition' that is causing the symptoms or not.
Many psychological diagnoses seem to be really a shorthand way of describing clusters of symptoms that often go together so are nothing more than labels. It seems, however, that people find reassurance in this. Hence all the articles you see written by people saying things like ' I didn't understand why I couldn't concentrate on work / finish anything but I now know its because I have ADHD and am much happier' when really all 'having ADHD' means is (potentially) that you can't concentrate on work or finish things. This doesn't mean ADHD doesn't exist it just means it could be a label for a set of behavioural traits and not a cause of them.
100%. The vast majority of people diagnosed with depression or anxiety conceptualize their condition with the same diagnostic literalism that holds back clinical research. I think you’re right that most people find comfort in a sort of diagnostic essence account -- e.g., it reduces stigma and self-blame. But there are also huge downsides to people having simplistic understandings of what’s going wrong. Effective therapy helps people understand the mental and behavioral processes that maintain their condition. It’s harder for people to understand those processes if they’re committed to seeing themselves as having some sort of permanent structural brain defect (rather than a vulnerability to falling into particular dysfunctional processes).
Oh, and I recall from Maths lessons an aeon ago that the first, classic, example of a Poisson distribution was the number of Prussian cavalry officers who die each year from being kicked by a horse.
that's amazing
How many Prussian officers die each year?
Wikipedia has the following:
"Bortkiewicz graduated from the Law Faculty in 1890. In 1898 he published a book about the Poisson distribution, titled The Law of Small Numbers.[1] In this book he first noted that events with low frequency in a large population follow a Poisson distribution even when the probabilities of the events varied. It was that book that made the Prussian horse-kicking data famous. The data gave the number of soldiers killed by being kicked by a horse each year in each of 14 cavalry corps over a 20-year period. Bortkiewicz showed that those numbers followed a Poisson distribution. "
Thanks for sharing. The Prussian horse-kicking data is both helpful to understand Poisson and a funny turn of phrase. Is it Prussian horse-kick deaths (People killed by horse kicks in the State of Prussia)? Or Prussian-horse kick deaths (Prussian horses doing the kicking)? Just so long as it's not Prussians kicking horses!
Interesting as always. I am not sure the temperature analogy quite works though. Clearly any instrument turns something we can't directly see (temperature, electric voltage, blood pressure) into something we can (height of mercury in a tube, position of a needle on a dial, etc) via a physical effect that has a theoretical basis for its validity. In the case of temperature its two theories - the laws of thermodynamics (I think its the zeroth law ?) that says that in thermal equilibrium the gas and the thermometer will have the same temperature (roughly) and the law underpinning the co-efficient of thermal expansion that tells us that gives us a calibrated scale.
So my point is that in the case of depression there is no theoretical underpinning of the measurement scale . So we have no basis for determining if the scale of the measured score bears any relation to the scale of any underlying property.
blame Stuart, it was his stupid analogy (this is an interesting and thoughtful comment that deserves a better response but it's late and I'm going to bed. Glad you enjoyed the episode!)
Thank you for another wonderful show. As a practicing psychologist, I would add that another problem with the self-report questionnaires you mentioned in your episode is that they don’t contain symptom validity measures. In today’s world where most people are familiar with depression symptoms (or can easily find them listed online), it is important to know whether the person completing the questionnaire is exaggerating or minimizing their symptoms, or being consistent in their endorsement of the questionnaire items. There are questionnaires that contain symptom validity measures (the MMPI-3 is an example) but such measures are lengthy and tend to be underutilized in both research and therapy settings.
(Grew up with Celsius and switched to Fahrenheit only after moving to US for grad school, so have a lot of experience with both)
Fahrenheit is superior to Celsius for everyday use (at least in non-arctic climates), and this is the hill I will happily die on
1) it's a true 0 to 100 scale. 0F is very cold, 100F is very hot, and this is the range that covers most people's everyday experience.
2) it almost entirely eliminates the conversation: "-what's the weather like?" "it's 5 degrees." "plus or minus?" which happened ALL THE TIME when i lived in Northern Europe (where its around 0C for a significant proportion of the year). With C the only way to avoid it is to always specify plus or minus but that is an extra word and a lot of people chose not to say it. With F you can have the same problem, but in the range that is much more rare, and, frankly, difference between -5F and 5F is much less important.
3) 1 degree of C is just a bit too much. As a result, a lot of the time you need to think of it in whole degrees even for everyday stuff, and certainly can't just round to nearest 10 (difference between 5 and 14 or 11 and 19 is massive). With F, most of the time the first digit of the temperature gives all you all the info you need to know.
Basically, it is almost perfectly tuned for the temperatures that people actually use. And guess what, scientists, always switch to the temperature scale that is tuned for their specific need, whether it is Celsius, Kelvin or electronvolt.
2 things:
1) Fahrenheit is a legitimate scale and and partisan adherence to Celsius is arbitrary and silly.
2) Aren't there some diseases where the response to a cure is part of the diagnosis? My sister has lupus (I guess) and she told me part of the diagnosis is that the symptoms improve after certain prescriptions. Could that not be the same for depression? The fact that a course of treatment is successful could also be a diagnostic tool?
(Water boils at 212 and freezes at below 32. It's not that hard, Stuart. If you measure angles in degrees instead of radians you must confess your sins!)