The question, Do the Amish have a lower rate of autism? is looking like a road accident.

This question is to explain what happened, and to invite suggestions on how to fix it.

Let's work our way through it.

Version 1

The initial version simply asked whether autism was low among the Amish, and whether that was because of low vaccination rates.

Now, the question of whether MMR vaccines cause autism has been addressed already on Skeptics.SE, and then again when new documents came to light, but this was a new question. However, it had no notability.

Version 2

That was fixed in Version 2 by adding a notable claim - a reference to this nonsensical argument from Dan Olmsted.

The logical flaws in this article are numerous and it contains a number of JAQing off hedges. (i.e. that we should spend more effort investigating some journalist's whimsical questions.)

Worse, it doesn't explicitly state that the Amish are (a) lower in autism or (b) vaccine-deniers. It does quote one doctor as saying that he hasn't seen autism.

Answer 1

At this point, an answer was given by @jamesqf that (a) argued autism wasn't linked to vaccines (citing a meta-study) and (b) supposed, without any evidence, that the Amish community might be more accepting of children on the Autism spectrum without a diagnosis.

That answer was downvoted, flagged and received upvoted negative comments, before I saw it. I deleted it on the grounds that it wasn't an answer.

The reference did NOT address the issue of the Amish at all. Without evidence, the supposition that the Amish are more tolerant of mental health issues is little more than prejudice, and didn't add to the answer.

Further, it is inappropriate to respond to a claim that evidence exists with a generic "There is no evidence." If there is a lower level of autism amongst the Amish community and they have a lower vaccination rate and there are no confounding factors, including lower diagnosis specificity in their health regimes, then that would need to be included in the next version of the meta-analysis, not ignored out of hand.

(On the other hand, if you wanted to argue that such a study should be low priority, should not be sponsored by the public purse, and that the subjects of such a study should be informed before signing up that the study is highly unlikely to advance scientific knowledge, then that link would be useful.)

The answerer complained in a comment that the answer was still valid. I hope this explains my reasoning.

Version 3

Version 3 (disclosure: edited by me) improved it slightly but didn't really address the key flaws. The question was now focussed on an article that argued in circles.

Version 4

Version 4 just added a tag,

Answer 2

The second answer by @RobWatts doesn't answer the question either.

It points out the nonsense article makes some errors: the Amish do vaccinate, that the quoted doctor is probably exaggerating about his exposure to the Amish and that the authors peers criticise him.


So, now we have an answer eviscerating the Olmsted article... that wasn't part of the OP's question. We, as a community, added an article for notability, focussed the question on that answer, and then shredded that article as nonsense, leaving the original question from the OP unanswered.

I've closed the whole sorry mess. I hope the OP can come back and clarify where he heard the claim - was it from Olmsted?

Olmsted certainly makes claims that we can investigate (and, it seems) rip apart, but until it is clear that is the target, we should hold off.

If it is the target, we should be more explicit which of his claims are the ones we should investigate.

  • I'm on the fence on what should be done (I have another proposal, which I'll post as an answer), but I'm nominating this post for Moderator Oscars.
    – user5341
    Jan 21, 2015 at 16:59

5 Answers 5


So, now we have an answer eviscerating the Olmsted article... that wasn't part of the OP's question. We, as a community, added an article for notability, focussed the question on that answer, and then shredded that article as nonsense, leaving the original question from the OP unanswered.

When I answered the question I failed to look through the history of the question first. However, I think that "eviscerating the Olmsted article" is a reasonable way to approach the question even in its original form. Regardless of where the OP heard the claim, it appears that the Olmsted article is the original source.

I just edited the opening paragraph of my answer to hopefully make that a little bit clearer:

While researching this question, everything I found tied back to the Olmsted article. In Wikipedia's article about the "Amish anomaly" it mentions that these claims (of the Amish having a lower prevalence of autism) "originate primarily from columns by Dan Olmsted". Rather than simply demonstrating the claim's notability, that article is actually a source of the claim. I haven't found anything that shows whether or not autism is significantly less common among the Amish, but I have found information that gives good reason to be skeptical of Olmsted's claims.

So my answer is attempting to show that:

  1. Dan Olmsted is the reason why people have heard that autism is less prevalent among the Amish
  2. Dan Olmsted's claims do not have any credibility (as @Oddthinking put it, he's "JAQing off")
  3. Because of 1 & 2 the claim should be viewed as notable in the sense that it is widely known, but that it is also baseless.

Given that the claim is baseless, I feel like the lack of a study determining the prevalence of autism among the Amish should be viewed as soft evidence that it isn't a real issue.


So, now we have an answer eviscerating the Olmsted article... that wasn't part of the OP's question. We, as a community, added an article for notability, focussed the question on that answer, and then shredded that article as nonsense, leaving the original question from the OP unanswered.

I suggest reverting back to Version 1 of the question.

When the question was first posted I did my own Googling to verify whether it's a notable claim, discovered/decided that it is a notable claim (e.g. there's a WIkipedia article on the subject), and decided not to add a reference to any specific claim to the question.

You argued earlier in Must every questioned 'notable claim' include a referenced citation with a quote? that a quoted reference is not required.


Another option might be to do the Solomon thing, and cut the baby in halves:

  1. Revert the question to V1, sans Olmsted claim

  2. Post a SEPARATE question(s) about specifically Olmsted claims, and migrate Olmsted specific answer(s) there.


If you want to simultaneously try to exclude the current (accepted) answer, and add a reference to a notable claim, perhaps you could do that by adding the following to Version 1 of the question:

For example, Wikipedia's Amish anomaly article says,

The 'Amish anomaly' refers to claims of unusually low rates of autism spectrum disorders among the Amish, which originate primarily from columns by Dan Olmsted

Excluding (ignoring) Dan Olmsted's claims, is there any other relevant study showing that the autism is low among the Amish?

Given that the current answer was accepted, though, perhaps it adequately answered the OP's concern.

  • But at the point I posted my original answer (to a question quite a bit different than the current edited version), there were no other answers, accepted or not.
    – jamesqf
    Jan 18, 2015 at 4:04

So a couple of meta-questions:

1) How do you properly respond to a question like this, which postulates a link in some situations/groups when there is overwhelming evidence that no such link exists, except by stating the fact and linking to (some of) the evidence?

2) Why isn't it proper to offer alternative hypotheses that might explain the supposed link?

  • 2
    1) If I say "There is a dragon in my garden." you can't merely say "There are no such things as dragons." because clearly my claim, were it true, would overthrow all existing evidence for the non-existence of dragons. (Although it might be good argument to suggest my claim shouldn't be accepted without extraordinary evidence.) To answer this claim, we have to determine if there is a dragon in my garden.
    – Oddthinking Mod
    Jan 18, 2015 at 4:20
  • 1
    2) Generally, speculation is off-topic. We sometimes will accept conjectured alternative hypotheses, to help understanding, AFTER showing the original claim is true/false. But not INSTEAD of showing the original claim is true/false, or you just get "It might have been swamp gas" explanations that don't help anyone. We are looking for definitive answers, not speculation and conjecture.
    – Oddthinking Mod
    Jan 18, 2015 at 4:24
  • @Oddthinking: But we can strongly suspect that there are other explanations. E.g. Oddthinking is telling whoppers, Oddthinking is doing really good drugs, somebody put a good fake dragon in Oddthinking's garden as a publicity stunt for a dragon movie... I think it's reasonable to speculate on the greater liklihood of all these, while asking for actual evidence of the dragon.
    – jamesqf
    Jan 18, 2015 at 18:38
  • 1
    Imagine the question is, "Are temperatures declining in Finland?" and the answer, "Global warming is a well-proven theory so, I don't know whether temperatures are declining in Finland (are they?) but if they are that's not caused by a lack of global warming." A big part of the problem is that it doesn't answer the question, which is asking about what facts/observations/evidence are, about the Amish (or about Finland)? A secondary part of the problem is that it's asserting theory before/without having seen what the specific facts are.
    – ChrisW
    Jan 18, 2015 at 21:36
  • All of those speculations are baseless opinions. The fact you came up with four of them without trying shows each one has a remote chance of being right. The Internet already has plenty of places where wild speculation is welcomed. Here we are looking for definitive answers.
    – Oddthinking Mod
    Jan 18, 2015 at 23:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .