I need help with this statistical autism measurement result

[ambushed]

From: A Population-Based Study of Measles, Mumps, and Rubella Vaccination and Autism, here is the summary conclusion:

@Madsen

Results Of the 537,303 children in the cohort (representing 2,129,864 person-years), 440,655 (82.0 percent) had received the MMR vaccine. We identified 316 children with a diagnosis of autistic disorder and 422 with a diagnosis of other autistic-spectrum disorders. After adjustment for potential confounders, the relative risk of autistic disorder in the group of vaccinated children, as compared with the unvaccinated group, was 0.92 (95 percent confidence interval, 0.68 to 1.24), and the relative risk of another autistic-spectrum disorder was 0.83 (95 percent confidence interval, 0.65 to 1.07). There was no association between the age at the time of vaccination, the time since vaccination, or the date of vaccination and the development of autistic disorder.

Now, I understand that’s saying that there were small differences in the likelihood of developing autism between the vaccinated vs the un-vaccinated group, but I’m too dumb to know how to express that in layperson (like me!) language while still remaining fairly accurate.

Little help?

[footballisplayedwithyourfeet]

This pretty much means that the differences that were found are quite likely to be caused by chance (you are not looking at all the people in the world, it is just a subset). The confidence intervals mentioned indicate that if the relative risks were 0.68 and 0.65 *(or lower) you would be for 95% sure there is a real difference (still leaves 5% possibility it was chance); with the relative risk numbers are nowhere near these scores you should condlude there is no significant (jargon for we are not more than 95% sure) difference.

*or more than 1.24 and 1.07 for th upper end

[ambushed]

@footballisplayedwithyourfeet

This pretty much means that the differences that were found are quite likely to be caused by chance (you are not looking at all the people in the world, it is just a subset). The confidence intervals mentioned indicate that if the relative risks were 0.68 and 0.65 *(or lower) you would be for 95% sure there is a real difference (still leaves 5% possibility it was chance); with the relative risk numbers are nowhere near these scores you should condlude there is no significant (jargon for we are not more than 95% sure) difference.

*or more than 1.24 and 1.07 for th upper end

Thanks for your help!

May I ask for a bit more?

Would you be so kind as to fact-check this if I were to put it this simpler way?

“There was no statistically significant or meaningful difference in autism rates between the two groups. The small variance in the raw numbers was in the noise level, i.e., it was statistically shown to be due solely to random chance.”

Is that accurate and honest, or am I fudging too much?

[footballisplayedwithyourfeet]

I think this sentence is ok, but I would take out solely. The thing is that we can’t tell whether it is chance or a real effect and since we are not sure, we play safe by saying there is no significant effect. It is more a convention then anything else to have the cut off point at 95% confidence, but off course even the smallest difference could be because of a real effect; it is just that we can’t find out if that is the case. Therefore we calculte (probably through the t-test) the probability of there being a real difference and in this case that probability is small, you cannot say you are sure there is no effect (which is what you claim when you say it is solely due to random chance). You might want to say the difference is likely to be due to…etc.

[ambushed]

@footballisplayedwithyourfeet

I think this sentence is ok, but I would take out solely. The thing is that we can’t tell whether it is chance or a real effect and since we are not sure, we play safe by saying there is no significant effect. It is more a convention then anything else to have the cut off point at 95% confidence, but off course even the smallest difference could be because of a real effect; it is just that we can’t find out if that is the case. Therefore we calculte (probably through the t-test) the probability of there being a real difference and in this case that probability is small, you cannot say you are sure there is no effect (which is what you claim when you say it is solely due to random chance). You might want to say the difference is likely to be due to…etc.

Thanks for the fact check on that. I can see that I should definitely not have used the word “solely”. It’s a clear overstatement, and I don’t want to overstate the case or otherwise misrepresent the scientific results. I can’t mess this up!

Thanks again.

[Quercus]

If you wanted to be really careful, you wouldn’t say it was shown to be due to chance, you’d say there’s no statistical evidence that it’s anything other than chance.

But if you’re debating someone who believes there is a link, then being careful and subtle might be like speaking Latin to a cat.