Stat 418 – Day 3Last TimeConfidence interval for pLikelihood functionTest of significance?Likelihood ratio testTo DoStat 418 – Day 3Maximum Likelihood Estimation (Ch. 1)Last TimeMultiple confidence interval procedures for Main goal: Achieve stated confidence level, e.g., 95% of intervals for all possible random samples capture the value of the parameterAlso nice if narrowerConfidence interval for Exact binomialWald intervalScore interval (Wilson)Which values of  would not be rejected…Adjusted Wald (Agresti & Coull, 1998)Approximate is better than "exact" for interval estimation of binomial proportions, American Statistician, 52:119–126.Likelihood function=C(407,248) 248(1-)209Test of significance?Gives you a point estimateAlso allows you to compare to other valuesAnother value of  will be plausible if the likelihood function at that value isn’t too much smallerLikelihood ratio testLikelihood Ratio compares the unrestricted maximum likelihood to value under nullPearson is z2 from “score” test (two-sided)To DoFinish reading Ch. 1Investigation 1HW 1 solutions posted tonight?HW 2 posted this weekend?Final Exam survey (by