www.Tobold.com
Tobold@GMail.com
A blog with my thoughts regarding games I am playing and other stuff in life. Please read my Terms of Service
RSS Feed
Saturday, September 06, 2025
Science, probability and political bias
Imagine that your probability of getting struck by lightning is 1 in a million in a given year. Now imagine that carrying a mobile phone with you increases that chance to 2 in a million. Would you stop carrying a mobile phone? With so many people carrying mobile phones, how would you even scientifically prove that there is a causal effect between carrying a mobile phone and getting struck by lightning? And if you get struck by lightning and want to sue the mobile phone company for damages, how do you prove that you aren't the 1 in a million that would have gotten struck by lightning anyway, but are the additional 1 in a million that only got struck because they carried a mobile phone?
The good news is that I made the link between mobile phones and lightning up. Not the chance to get struck by lightning, it is around that number, although it obviously depends a lot on where and how you live. A link between carrying a mobile phone and getting cancer has been discussed, but the chance to get cancer is higher than the chance to get struck by lightning, and the possible increase to that chance is so low as to be considered not scientifically proven. Right now, in the EU official classification, mobile phones are in the same category of increased cancer risk as pickles. In any case, my list of questions in the first paragraph is relevant to any claim of "product X causes undesirable outcome Y". The question is always how high the chance of undesirable outcome Y is anyway, by how much product X changes that probability, and how a link can be proven.
While links between products and undesirable outcomes have been an issue for a long time, it is a relatively new development that this is now a matter of political bias. This week in the news is a possible link between the use of Tylenol during pregnancy and the child being born with autism. In the United States the probability of a child being born with autism is estimated to be 1 in 36. That is certainly higher than a century ago, but it is unclear whether this is due to external factors like Tylenol, or simply to an increased awareness of autism. Now there are scientific studies that link Tylenol to autism, and there are scientific studies that show no such link. Just as there are scientific studies that link glyphosate to non-Hodgkin lymphoma, and there being scientific studies that show no such link.
The difference now being that these things have been politicised, and in the USA a Republican would be more likely to believe in the link between Tylenol and autism, while a Democrat would be more likely to believe in the link between glyphosate and non-Hodgkin lymphoma. It should be obvious that this isn't how probability or science works. But humans are notoriously bad with probabilities. And thus anything that science would consider to be an event of low probability can find its way into our belief systems. If you thought my example of a mobile phone getting you struck by lightning was ridiculous, how about the belief that it was your sins that would get you struck by lightning? That was pretty wide spread in humanity over many different cultures and religions. And our belief systems are evolving away from being religious to being political. Especially in two-party systems like the USA, party affiliation today has a stronger influence on what you believe to be true than your religious affiliation.
For all questions of "does product X causes undesirable outcome Y", I would urge you to consider that whatever you read might be a result of political bias. You still should ask the relevant scientific question of how high the probability of undesirable outcome X is with and without product X. How strong or weak is the link? If the probability is relatively high to start with, and increases only by a small amount, how justified would it be to sue the producer of product X for damages? Although in a legal system where the probability of getting over 1 million dollars in a lawsuit is higher than the probability of winning 1 million dollars in the lottery, scientific proof is maybe the last thing you want.
- posted by Tobold @ 9:02 AM Permanent Link
Comments:
<< Home
Newer› ‹Older
Do you realize that the problem you highlight is the same as the problem of proving the efficacy of the medical drugs and treatments (only in inverse, with the drugs the question is what ratio of people get better on their own without the drug versus what ratio of people get better when administered the drug)? This problem was kind of "solved" by respectable authorities like FDI with double-blind drug tests.
These test are designed in the following way: two groups of pacients are tested, one groups gets the drug and the other gets placebo. Double-blind means that not only the pacients do not know what they get, but the personnel that administers drug/placebo and subsequently checks on patients does not know what it administers, so that it cannot affect their perception of treatment's success. After the test is over and some people in both groups do get better, two ratios are compared and tested to have a "statistically significant" difference (as a scientist, I'm sure you know what that means).
The same can be done with mobile phones and lightning, although I doubt that double-blindness can be achieved in this case (but it may not be necessary). The only requirements are that bot groups of people (with and without phones) are sufficienly large to account to small odds of lightning, and the experiments is conducted in the areas where thunderstorms are not rare.
These test are designed in the following way: two groups of pacients are tested, one groups gets the drug and the other gets placebo. Double-blind means that not only the pacients do not know what they get, but the personnel that administers drug/placebo and subsequently checks on patients does not know what it administers, so that it cannot affect their perception of treatment's success. After the test is over and some people in both groups do get better, two ratios are compared and tested to have a "statistically significant" difference (as a scientist, I'm sure you know what that means).
The same can be done with mobile phones and lightning, although I doubt that double-blindness can be achieved in this case (but it may not be necessary). The only requirements are that bot groups of people (with and without phones) are sufficienly large to account to small odds of lightning, and the experiments is conducted in the areas where thunderstorms are not rare.
Yeah, but who is going to perform a large scale study like that to check a potential harmful effect, knowing that he would put half of the group into harm's way?
People fell into this same trap with ADHD. Is it really that rates of ADHD rose in the past 50 years because of diets or vaccines or whatever. Or is it the more likely scenario that we are simply diagnosing more kids with ADHD as the awareness of it grew?
Could Tylenol be linked to Autism? Maybe. Tylenol is also one of the most common pain meds pregnant mothers can take so is it really the source or are we just seeing a correlation? Studying things we thought we safe and reevaluating whether pregnant woman should take them is not a bad thing and I support that.
However when one side is also dissuading peoples from taking life saving safe vaccines and promoting Ivermectin as some sort of wonder drug you can excuse me for feeling a bit skeptical when they now claim Tylenol and MRNA vaccines are unsafe with hardly any new information to back that up.
I've experienced loved ones in my life who refuse to see doctors because they believe in quackery that they see on social media and think that herbal creams and oils can solve all their very real medical problems. It's very sad how people are so distrustful of science and medicine that they would rather suffer and eventually die than go see a doctor.
Could Tylenol be linked to Autism? Maybe. Tylenol is also one of the most common pain meds pregnant mothers can take so is it really the source or are we just seeing a correlation? Studying things we thought we safe and reevaluating whether pregnant woman should take them is not a bad thing and I support that.
However when one side is also dissuading peoples from taking life saving safe vaccines and promoting Ivermectin as some sort of wonder drug you can excuse me for feeling a bit skeptical when they now claim Tylenol and MRNA vaccines are unsafe with hardly any new information to back that up.
I've experienced loved ones in my life who refuse to see doctors because they believe in quackery that they see on social media and think that herbal creams and oils can solve all their very real medical problems. It's very sad how people are so distrustful of science and medicine that they would rather suffer and eventually die than go see a doctor.
The same people who test untested drugs on seriously ill people, including giving them placebo instead of cure, will as easily test lightning attractiveness of cell phones - for science!
I am not a fan of how they test for cancer as it almost always seems to result in a positive result. Massive overfeeding of almost anything results in cancer. Is it the a reflection that almost anything causes cancer or just that massive overfeeding causes cancer
On the other hand the fairly well established result that Americans generally die young despite the American medical profession generally being the most advanced in the world . This even applies to well off patients. Americans dieing now ,tend to die at the same age as they died in the 1950's, in comparison, Australians live another decade as compared to the fifties.
But exactly why American medical advances is not helping American live longer is much harder to understand.
I would put it down to diet and increase in contact with chemicals.
That and you can live a fairly long life in an unhealthy manner but it only really catches up with you past 70. Ie you can eat unhealthy amounts of fat and oil and meat and not have enough exercise and be slightly into the excessive morbidity levels of weight and generally be fine for many decades.
On the other hand the fairly well established result that Americans generally die young despite the American medical profession generally being the most advanced in the world . This even applies to well off patients. Americans dieing now ,tend to die at the same age as they died in the 1950's, in comparison, Australians live another decade as compared to the fifties.
But exactly why American medical advances is not helping American live longer is much harder to understand.
I would put it down to diet and increase in contact with chemicals.
That and you can live a fairly long life in an unhealthy manner but it only really catches up with you past 70. Ie you can eat unhealthy amounts of fat and oil and meat and not have enough exercise and be slightly into the excessive morbidity levels of weight and generally be fine for many decades.
I'm pretty sure life expectancy starting to tick down again in the US is down to large swaths of the country being incredibly skeptical of medical science (or really just science in general). The fact that fewer and fewer people have full access to our highly advanced medicine as inequality and poverty are also starting to tick up likely also plays a role.
In most countries the best healthcare, or close to it, available is free. So advances in medicine help everyone and continue to increase life expectancy. Here most healthcare is only as good as what you can afford. If you have to wait until something is bad enough to take you to an emergency room, where by law they have to treat you whether you can pay or not, you are already risking death.
Post a Comment
In most countries the best healthcare, or close to it, available is free. So advances in medicine help everyone and continue to increase life expectancy. Here most healthcare is only as good as what you can afford. If you have to wait until something is bad enough to take you to an emergency room, where by law they have to treat you whether you can pay or not, you are already risking death.
<< Home
