Unit 2 Regression and Correlation WEEK 3 - Practice Problems. Due: Wednesday February 10, 2016

Transcription

1 BIOSTATS 640 Spring Regression and Correlation Page 1 of 5 Unit 2 Regression and Correlation WEEK 3 - Practice Problems Due: Wednesday February 10, A psychiatrist wants to know whether the level of pathology (Y) in psychotic patients 6 months after treatment could be predicted with reasonable accuracy from knowledge of pretreatment symptom ratings of thinking disturbance (X 1 ) and hostile suspiciousness (X 2 ). (a) The least squares estimation equation involving both independent variables is given by Y = (X 1 ) 7.147(X 2 ) Using this equation, determine the predicted level of pathology (Y) for a patient with pretreatment scores of 2.80 on thinking disturbance and 7.0 on hostile suspiciousness. How does the predicted value obtained compare with the actual value of 25 observed for this patient? (b) Using the analysis of variance tables below, carry out the overall regression F tests for models containing both X 1 and X 2, X 1 alone, and X 2 alone. Regression on X Residual Regression on X Residual Regression on X 1, X Residual

2 BIOSTATS 640 Spring Regression and Correlation Page 2 of 5 (c) Based on your results in part (b), how would you rate the importance of the two variables in predicting Y? (d) What are the R 2 values for the three regressions referred to in part (b)? (e) What is the best model involving either one or both of the two independent variables? 2. In an experiment to describe the toxic action of a certain chemical on silkworm larvae, the relationship of log 10 (dose) and log 10 (larva weight) to log 10 (survival) was sought. The data, obtained by feeding each larva a precisely measured dose of the chemical in an aqueous solution and then recording the survival time (ie time until death) are given in the table. Also given are relevant computer results and the analysis of variance table. Larva Y = log 10 (survival time) X 1 =log 10 (dose) X 2 =log 10 (weight) Larva Y = log 10 (survival time) X 1 =log 10 (dose) X 2 =log 10 (weight) Y = (X 1 ) Y = (X 2 ) Y = (X 1 ) (X 2 ) Regression on X Residual Regression on X Residual Regression on X 1, X Residual

3 BIOSTATS 640 Spring Regression and Correlation Page 3 of 5 (a) Test for the significance of the overall regression involving both independent variables X 1 and X 2. (b) Test to see whether using X 1 alone significantly helps in predicting survival time. (c) Test to see whether using X 2 alone significantly helps in predicting survival time. (d) Compute R 2 for each of the three models. (e) Which independent predictor do you consider to be the best single predictor of survival time? (f) Which model involving one or both of the independent predictors do you prefer and why? 3. Using R or Stata (your choice), try your hand at reproducing the analysis of variance tables you worked with in problem #2. Stata Users How to access the data I ve created the STATA data set for you and uploaded it to our course website. It is larvae.dta. To access larvae.dta in STATA requires 2 steps: Step 1 Download larvae.dta to your computer From the course website page THIS WEEK, right click on larvae.dta Alternatively, from the course website page REGRESSSION, right click on larvae.dta Tip Make sure you know where larvae.dta is located on your computer. Step 2 In STATA, open larvae.dta Launch STATA. From the main menu bar: FILE > OPEN Browse to find larvae.dta on your computer. R Users How to access the data Issue the following commands library(foreign) stata <- larvae <- read.dta(file=stata)

4 BIOSTATS 640 Spring Regression and Correlation Page 4 of 5 4. An educator examined the relationship between number of hours devoted to reading each week (Y) and the independent variables social class (X 1 ), number of years school completed (X 2 ), and reading speed measured by pages read per hour (X 3 ). The analysis of variance table obtained from a stepwise regression analysis on data for a sample of 19 women over the age of 60 is shown. Regression (X 3 ) (X 2 X 3 ) (X 1 X 2,X 3 ) Residual Total, corrected IMPORTANT!! - Please read the following before you continue. The table above makes use of a new notation. Regression (X 3 ) Sum of Squares = : This is the regression sum of squares for the model containing the one predictor X 3. Regression (X 2 X 3 ) Sum of Squares = : The vertical line in this notation stands for conditional on This is the extra regression sum of squares obtained by adding the predictor X 2 to a model that already has the predictor X 3. You can also think of this as the change in the model sum of squares accompanying the addition of the predictor X 2, controlling for (conditional on X 3 ). Regression (X 1 X 2, X 3 ) Sum of Squares = : The vertical line in this notation stands for conditional on This is the extra regression sum of squares obtained by adding the predictor X 1 to a model that already has the predictors X 2 and X 3. You can also think of this as the change in the model sum of squares accompanying the addition of the predictor X 1, controlling for (conditional on X 2 and X 3 ). Tip! Since the total sum of squares is a fixed total, you can use the information provided in this table to obtain the analysis of variance tables for the following models. Model 1: Predictor = X 3 Model 2: Predictors are = X 3, X 2 Model 3: Predictors are = X 3,X 2, X 1 (a) Test the significance of each variable as it enters the model. (b) Test H O : β 1 = β 2 = 0 in the model Y = β 0 + β 1 X 1 + β 2 X 2 + β 3 X 3 + E. (c) Why can t we test H O : β 1 = β 3 = 0 using the ANOVA table given? What formula would you use for this test? (d) What is your overall evaluation concerning the appropriate model to use given the results in parts (a) and (b)?

5 BIOSTATS 640 Spring Regression and Correlation Page 5 of 5 5. Consider the following analysis of variance table. Regression (X 1 ) 1 18, (X 3 X 1 ) 1 7, (X 2 X 1,X 3 ) Residual 16 2, Total, corrected 19 28, Using a type I error of 0.05, (a) Provide a test to compare the following two models. In 1-2 sentences, interpret. Y = β 0 + β 1 X 1 + β 2 X 2 + β 3 X 3 + E. VERSUS Y = β 0 + β 1 X 1 + E. (b) Provide a test to compare the following two models. In 1-2 sentences, interpret. Y = β 0 + β 1 X 1 + β 3 X 3 + E. VERSUS Y = β 0 + E. (c) State which two models are being compared in computing: (18, , )/3 F = ( )/16