1. 1 Project One Carpoll

2. 2

3. 3 Excel Convert categorical data into dummy variables Type of vehicle: family, sporty, work Sort type Select some of family observations Expand selection in dialog box Code family as one, sporty and work as zero

4. 4

5. 5

6. 6

7. 7 Vehicle Type Family Yes: One, 155 No: zero, 148 Total: 303

8. 8

9. 9

10. 10

11. 11 Sort Gender and Code zero/one Male: one; Female: zero Select some observations in sex, expand selection, sort Ditto for marital status Married: one Single: zero

12. 12

13. 13

14. 14

15. 15 Cross-Classification of Type with Gender FamilySporty & Work Total Female7662138 Male7986165 Total155148303

16. 16 #-Way classification in one step

17. 17 Family: Yes MarriedSingleTotal Female581876 Male611879 Total11936155

18. 18

19. 19 Cut and Paste Fambern, age, gender, marital into EViews File Menu New workfile

20. 20 Workfile Dialog Box

21. 21 Generate x = 1

22. 22

23. 23

24. 24

25. 25 Estimate Linear Probability Model

26. 26

27. 27 Age and marital are significant, And have a positive effect on The probability of choosing a Family car. Gender barely adds to the explanation and is negative Given age and gender, being married increases the Probability of favoring a family car by 0.23

28. 28

29. 29 Married and Female Single and Male

30. 30

31. 31 Married and Female Single and Male

32. 32

33. 33

34. 34

35. 35 This probability model does a better job of explaining those Who favor a family car compared To those who don’t

36. 36 Logit Fit

37. 37 Married women Single men

38. 38 Linear Probability Model Car Size Car size: Small, medium, large Varies with age, gender, and marital status

39. 39

40. 40

41. 41 Sort by Size and then by Age

42. 42 Contingency Table Analysis: Observed SmallMediumLargeMargin 18-29705915149 30-39595821138 40- 8 7 621 Margin13712442303

43. 43 Contingency Table Analysis: SmallMediumLargeMargin 18-29149 30-39138 40- 21 Margin13712442303

44. 44 Contingency Table Analysis: Expected SmallMediumLargeMargin 18-2967.461.020.7149 30-3962.456.519.1138 40- 9.5 8.6 2.921 Margin13712442303

45. 45 Contingency Table Analysis: [Observed – Expected] 2 SmallMediumLarge 18-29 6.76 4.0 32.49 30-3911.56 2.25 3.61 40- 2.25 2.56 9.61    

46. 46

47. 47 9.5 5%  2 =75.09

48. 48 Contingency Table Analysis: SmallMediumLarge 18-29 Fewer than exp 30-39Fewer than exp 40- More than exp Margin

49. 49 Summary People in their 20’s prefer smaller cars People in their forties or older prefer large cars

50. 50 Country Vs. Size Country: American, European, Japanese Size: small, medium, large

51. 51 Contingency Table Analysis: Observed SmallMediumLargeMargin American 26 19 92137 European 53 17 54124 Japanese 36 4 2 42 Margin 115 40 148303

52. 52 Contingency Table Analysis: SmallMediumLargeMargin American 137 European 124 Japanese 42 Margin 115 40 148303

53. 53 Contingency Table Analysis: Expected SmallMediumLargeMargin American 52.0 18.1 66.9137 European 47.1 16.4 60.6124 Japanese 15.9 5.5 20.5 42 Margin 115 40 148303

54. 54 Contingency Table Analysis: [Observed – Expected] 2 SmallMediumLarge American676 0.81 630 European 34.8 0.36 43.6 Japanese 404 2.25 342    

55. 55 Contingency Table Analysis: SmallMediumLarge American fewer than exp more European Japanese more fewer Margin

56. 56 Summary Preference for large American cars and for small Japanese cars.