Comment by anon____ | Hacker News Reader

Comment by anon____ | Hacker News Reader

Parent Full thread

> Your comment, and the way people tend to utilize these reviews, clearly conflates the concept of "not enough evidence" with "evidence against".

When there's plenty of opportunity to search for and find evidence, and multiple researchers do try, and repeatedly fail to find evidence (or they have mild "success" every one try out of ten, by pure luck, see: https://en.wikipedia.org/wiki/Data_dredging), that strongly suggests that there's no real effect.

Especially when you know that these people don't only do this with garlic, but with every single piece of spice, herb and other plant material you can imagine. Yes, some of them are effective; contrary to what many people believe (and you can read this in this thread, too), most, if not all our medicine is of "natural" (plant or animal) origin. Yet, at the same time, most "natural" stuff is not medicine. Why cling to the ones we can't prove effective however we try, when there are many more candidates?

> When there's plenty of opportunity to search for and find evidence, and multiple researchers do try, and repeatedly fail to find evidence (or they have mild "success" every one try out of ten, by pure luck

You're doing it too! Again, this review looked at a SINGLE study. They discounted other studies. Those other studies were not taken into account at all. Those other studies didn't "fail to find evidence"

No, they did review them, but you didn't read the article. The excluded studies were either investigated something else, not common cold, or were so bad that they failed right away. They still gave short summaries of them, and gave the reasons for exclusion:

> Andrianova 2003: The definition of ARD included influenza, thus excluding it from our review.

> Rafinski 1974: We excluded this study because there was no comparison group and it was a non‐randomised controlled trial.

> Ushirotake 2004: As the study was not randomised or blinded, there is a high risk of bias.

> Hiltunen 2007: This study did not meet this criterion as the cellulose could not be considered a standard treatment or a placebo.

[...]

I won't copy all of them, please go and read it yourself!

If it's not a randomized, controlled, blinded trial then it's not a trial to be considered any further, their results are unreliable.

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

I have indeed read the review. I think there's some confusion here.

Your comment: When there's plenty of opportunity to search for and find evidence, and multiple researchers do try, and repeatedly fail to find evidence

The review: *Explicitly discluded all but one study.

Here's the other thing. The ONE study they DID include did actually find a positive effect:

> Key results > The included study found that people who took garlic every day for three months (instead of a placebo) had fewer colds. That is, over the three‐month period, there were 24 occurrences of the common cold in the garlic group, compared to 65 in the placebo group. When participants experienced a cold, the length of illness was similar in both groups (4.63 versus 5.63 days).

The reason the analysis concluded that there's "not enough evidence" is simply because there was not more than a single study they deemed valid based on methodology. The thing is that even those discluded studies found a positive effect. In fact, all evidence points to a positive effect. The review is literally just saying there's not enough high quality evidence

Comments like yours confound "not enough evidence" for "evidence against". That's all I'm pointing out

Reply on news.ycombinator.com