所屬科目:研究所、轉學考(插大)◆統計學
4. Consider the linear regression model: y = ??0 + ??1x1 + ??2x2+. . . +??mxm + ??. The data set is{yii, xii1, xii2, … , xiiii}ii=1 n . After fitting these data into this model, you obtained the linear system which isdenoted by ?? = ???? + ??. Which of the followings is NOT TRUE ?(A)(B) , then we cannot find the least square estimator for ??.(C) Follow the least square criterion, we have the residuals. The sum of residuals must be zero.(D) If you add new independent variable xm+1 into the original linear regression model,i.e. y = β0 + β1x1 + β2x2+. . . + + ε, then the coefficient of determination won’tdecrease.
(2) With p unknown and n the number of random people polled, let be the fraction of the polled people who support the referendum. What is the smallest sample size n in order to have a 90% confidence that is within 0.01 of the true value of p? Your answer to this problem should be an integer
2. Someone claims to have found a long-lost work by Jane Austen. She asks you to decide whetheror not the book was actually written by Austen. You buy a copy of Sense and Sensibility and count thefrequencies of certain common words on some randomly selected pages. You do the same thing for the‘long-lost work’. You get the following table of counts.Using this data, set up and evaluate a significance test of the claim that the long-lost book is by Jane Austen. Use a significance level of 0.1.
(3) Find (sum of squared residuals of the Model1).
(4) Find (sum of squared residuals of the Model 2).
(5) Find (coefficient of determination in Model 1 and in Model 2 respectively). (6) According to the value of R2, which Model (Model 1 or Model 2) is better?