ECON 3100 Statistical investigation: This exercise again draws on the data set “grandfather clocks,” available on Canvas or under the Help menu, as a sample data set in JMP. 1. Develop an estimated regression equation with price as the dependent variable; age as the independent variable (yes, you did this for the previous problem set, use that output if you’d like, or run it again). Call this model A. 2. Develop an estimated regression equation with price as the dependent variable, age and number of bidders as the independent variables. Call this model B. 3. Which model do you prefer, model A or B? Why? 4. Interpret the coefficient on age in Model B. What precisely does it tell you about the relationship between age and price? 5. Interpret the coefficient on number of bidders in Model B. What precisely does it tell you about the relationship between the number of bidders and price. 6. In Model B, conduct an F-test as to whether there is a useful linear connection between the dependent variables and the independent variables in the population. Explain your results in non-technical terms. 7. In Model B, conduct t-tests as to whether the variables age or bidders individually have a statistically significant relationship to price, holding the other constant. Explain your findings fully. 8. Again, in Model B, include a plot of the residuals against y-bar and a histogram of the residuals (created in “Analyze Distribution”). 9. List and briefly explain the four assumptions of the regression model. Looking at the residual plot and histogram created in problem 8, do any of the assumptions appear to be violated? 10. Calculate the studentized residual of each observation. Are there any observations in the data set that you would consider outliers, based on a “studentized residual greater than 2 in absolute value” rule? Which observation has the largest studentized residual, in absolute value? 11. Are any observations unusually influential? Use Cook’s distance measure to determine the leverage of each observation. Which observation has the greatest leverage? 1 Everyday statistics Please read the New York Times article, “How Companies Learn Your Secrets.” You can find it in this folder on Canvas. You might also listen to or read the related radio story: http://www.mprnews.org/story/2012/03/07/target-data-mining-privacy 1. How does the process of habit formation make it more difficult to market new products? 2. Why does Target have an interest in predicting whether a customer (“guest”) is expecting a baby? What other life events might be of interest to retailers, in terms of selling new products? 3. What information (variables) might be relevant in a model used to predict whether a customer is expecting a baby? List at least five or six independent variables that data analysts might include in a model intended to predict whether a customer is expecting a baby. 4. Felix Wu, a scholar of privacy issues, describes three big data “threats,” that is, three ways in which the use of big data might be problematic. They are surveillance: giving people the uncomfortable feeling of being watched and studied, disclosure: the possibility that private information about you might be revealed to others inappropriately, and discrimination: the possibility that you will be treated differently than others based on information collected about you. Does Target’s use of customer data raise any of these threats (surveillance, disclosure, discrimination)? Which one(s) and why? 5. How did Target alter its approach to potentially pregnant customers to make it less threatening? Why were customers feeling threatened? 2 Grandfather Clocks - JMP Pro Edit Tables Rows Cols Analyze Graph Tools View Window Help 1 Grandfather Clocks srce Smyth, G. (2000), "Selling Prid 1 2 Bidders 13 12 7 3 4 5 6 7 8 9 6 11 12 10 9 10 11 12 13 14 Age 127 115 127 150 156 182 156 132 137 113 137 117 137 153 117 126 170 182 162 184 143 159 108 175 108 15 16 17 11 8 6 13 10 14 8 Price 1235 1080 845 1522 1047 1979 1822 1253 1297 946 1713 1024 1147 1092 1152 1336 2131 1550 1884 2041 854 1483 1055 1545 729 1792 1175 1593 785 744 1356 1262 18 19 11 20 10 21 6 Columns (3/1) Age Bidders Price 22 9 14 23 24 25 8 6 26 27 179 111 187 9 15 8 28 29 30 31 32 111 115 194 168 7 7 5 7 Rows All row Selected Excluded Hidden Labelled 32 0 0 Start X w
Purchase answer to see full attachment