統計の嘘

今日はネットで自閉症について動画を見ました。動画で南米の政府の声明が読み解かれました。声明の主張は自閉症の診断が増えて、同じ時にIQが高い自閉症の診断の割合が減りましたした。その統計は予想以外です。ほとんどの理解は自閉症の増加は診断の増加だけ、実の発生率は増えません。自閉症の理解が上がったから、微妙な自閉症が診断られます。

動画でハンク・グレーンが声明のソースを見ます。２０１８年のあとのソースにIQが知らない人ももっと頻繁に含まれます。そのため、IQが高い人とIQが低い人も発生率が下がりました。IQが知らない人の発生率だけが上がりました。声明を書くひとははずそれを知りました。英語でフレーズ「lies, damned lies and statistics」があります。意味合いは嘘より統計のほうが人を惑わせやすいです。

Today I watched a video online about autism. In the video, a statement by the US government was analyzed. The statement's claims were that autism diagnoses had increased, and at the same time the rate of high IQ diagnoses had decreased. Those statistics are unexpected. The general understanding is that the increase in autism is just an increase in diagnoses, not an increase in the rate. Because the understanding of autism has improved, more subtle autism can be diagnosed.

In the video, Hank Green looks at the report's sources. In the sources after 2018, people who's IQ is unknown are more often included. Because of that, the rate of people with high IQ, and the rate of people with low IQ have both decreased. Just the rate of people with unknown IQ increased. The person who wrote the report surely knew this. In English there's the phrase "lies, damned lies and statistics". The implication is that's its easier to mislead with statistics than with lies.