📱 A confidence interval is not what you thinkSource
📱 A confidence interval is not what you think
SHORT FORM
🎯📱 A confidence interval is not what you think
🚦 Status🗂️ Library
❤️ Favorite⭐ Favourite
📱VIDEOS
The true value does not move. It is fixed and unknown. What varies from sample to sample is the interval, so the 95 percent has to describe the method that draws intervals, not this one interval.The picture: a hundred samples, a hundred horizontal bars, one vertical line for the truth. About ninety-five bars cross the line and about five miss. You are holding one bar and you do not know which kind it is. A wider bar is not a worse result. It is an honest report of a small sample.A generic clip from a statistics teacher, about ninety seconds, saved on the tram on the evening the midterm came back at 68. I lost marks on exactly this. For three weeks my notes had said, in my own confident handwriting, that there is a 95 percent chance the true value lies inside the interval I calculated. It sounds right, it is what everybody says in meetings, and it is wrong. Hale Winter had written 'nearly' in the margin, which from him is severe. The repair, as I now say it: if I repeated the whole study many times and built an interval the same way each time, about 95 in 100 of those intervals would contain the true value, and this is one of them. The old sentence is still in 🔤 Confidence interval, crossed out, so that I can see what I used to think.Why a clip worked where lecture 6 had not: I was a week behind and half asleep in that lecture, and the hundred bars are a picture. I can hold a picture. Used since in the final in June, where the question came up and I drew the bars in the margin first, and at work whenever a dashboard shows a single number with no range. I ask how wide the bar is. People have started to see me coming. Filed with the rest of 00. Statistics for Product Decisions I Hub.
