Confidence intervals, t tests, P values

This was one of the first results in modern small-sample statistics. Confidence intervals, ttests, P values – p.15/31. Why we have to make the confidence interval wider We can’t just say that the estimated σˆ is sometimes 20% smaller than the true value, and sometimes 20% larger, so it all cancels out. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download