Hacker News Reader

Comment by seanwilson | Hacker News Reader

I thought this community would be more scientific...instead of anecdotes I'd rather see links to rigorous studies e.g. showing acupuncture or chiropractors work especially since these two forms of intervention have been subjected to decades of research already, have no plausible scientific basis and claim to treat many more things besides back pain. They should be held to the same standards as other medicine.

View on HN

The HN community is interesting in that respect and it's not only with some medical issues. As an example, any article on pesticides and modern food production will be full of comments decrying them and claiming health/taste benefits from alternatives. As you say, the community is surprisingly unscientific. Maybe that shouldn't be surprising because most people here probably have a technical and not scientific background and by virtue of having a reasonable disposable income are also the target demographic for products based on alternative science.

It's disturbing. I suspect most of this community would be anti-anti-vax (pro-vax?). And if one were to suggest on this forum that "I vaccinated my child and they got autism, therefore vaccines cause autism", they would be met with the appropriate skepticism. But in this thread people are professing "I had acupuncture and got better, therefore acupuncture works" and somehow that's considered defensible.

"Anecdotal" is not the correct term. "Testimonial" is. People are relaying the truth they observed, which is testimony. Anecdote means more along the lines of hearsay, or repeating something from a third party as if it were testimonial.

Anecdotal sounds accurate to me:

https://en.wikipedia.org/wiki/Anecdotal_evidence

> Anecdotal evidence is evidence from anecdotes, i.e., evidence collected in a casual or informal manner and relying heavily or entirely on personal testimony. When compared to other types of evidence, anecdotal evidence is generally regarded as limited in value due to a number of potential weaknesses, but may be considered within the scope of scientific method as some anecdotal evidence can be both empirical and verifiable, e.g. in the use of case studies in medicine. Other anecdotal evidence, however, does not qualify as scientific evidence, because its nature prevents it from being investigated by the scientific method.

> Where only one or a few anecdotes are presented, there is a larger chance that they may be unreliable due to cherry-picked or otherwise non-representative samples of typical cases.[1][2] Similarly, psychologists have found that due to cognitive bias people are more likely to remember notable or unusual examples rather than typical examples.[3] Thus, even when accurate, anecdotal evidence is not necessarily representative of a typical experience. Accurate determination of whether an anecdote is "typical" requires statistical evidence.[4] Misuse of anecdotal evidence is an informal fallacy and is sometimes referred to as the "person who" fallacy ("I know a person who..."; "I know of a case where..." etc.) which places undue weight on experiences of close peers which may not be typical. Compare with hasty generalization.

Wikipedia != dictionary

Yes, it's a shame that this thread is full of both quackery and anecdotal evidence.

However, don't rush to condemn acupuncture as a treatment for long term pain. Here's one recent analysis showing advantages over placebo: https://discover.dc.nihr.ac.uk/portal/article/4000672/acupun...

> in this thread people are professing "I had acupuncture and got better, therefore acupuncture works"

Where are you seeing that? I didn't see it. What I see is "I had acupuncture and got better, therefore acupuncture works for me", which is a much weaker claim. Moreover, it contains a great insight: treatment for back pain needs to be highly idiosyncratic.

Here's a recent good study from a reputable source showing acupuncture is more effective than placebo for some types of long term pain.

I agree that it's frustrating to see so much anecdote in threads like these.

https://discover.dc.nihr.ac.uk/portal/article/4000672/acupun...? Kindutm_content=buffered7fb&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer

Maybe that's another signal (along with the linked article) that the "science" in this arena hasn't advanced to the point of being repeatable and predictable.

> Maybe that's another signal (along with the linked article) that the "science" in this arena hasn't advanced to the point of being repeatable and predictable.

Why would there be a treatment right now that works well but can't be shown to work in a controlled study? There's nothing special about back pain that makes it impossible to scientifically study.

Personally, I think the signal you mention is more because back pain is susceptible to placebo effect, coming and going by itself, clearing up by itself or because something else you recently changed around the same time is helping it. These reasons and more are why you have to do large scale controlled studies to find out if something works rather than relying on anecdotes.

I don't see many people giving anecdotes for how alternative medicine fixed their broken leg or cured their cancer.

Back pain actually is kind of special in that the back and spine are so complex that a wide variety of different underlying conditions can cause similar symptoms. That introduces a lot of noise into the data and makes it harder to scientifically study.

It's easy to see if someone has a broken leg or test them for cancer. For most back pain, you can't even prove that person is actually suffering from it.

> For most back pain, you can't even prove that person is actually suffering from it.

Why does this matter? You ask the patient to report on their subjective experience the same way you would in studies about cures for depression and headaches.

My point is that 1) only when no surefire scientifically backed cure exists will most people take alternative therapies seriously and 2) problems that can resolve on their own (self-limiting) are very susceptible to generating unreliable anecdotes about cures. Back pain fits these two so generates many unreliable anecdotes which you should be highly skeptical about.

Why? No one is telling anyone to stop chemotherapy and use acupuncture to treat their cancer. In the context of back pain, so what if acupuncture isn't better than a placebo? We know placebo works for making people feel better. That's the whole reason why we need to control for them.

Why on earth would you assume any lay audience would have an up-to-date knowledge of the state of research in any given field in which they are not a working researcher?

I'm assuming that hackers that expect things like benchmarks for software performance claims who are then also skeptical enough to point out microbenchmarks aren't meaningful to be more skeptical than the average person. I don't have up-to-date knowledge about back pain or other medical topics but I know enough to be skeptical of certain claims and to ignore anecdotes.

Yes but that doesn't really justify the expectation of your original comment. It's pretty unreasonable to expect any group of laymen to have up-to-date knowledge of current research in any field outside of their own.

Reply on news.ycombinator.com