The DJIA is a summary of data. Does the DJIA provide information about a popula-

Components of DJIA NYSE Stock Company Percent of DJIA Price 52704 Microsoft Corp. 1.8214 26.19 Pfizer Inc. 2.4619 35.4 Procter Gamble Co. 7.5511 108.58 SBC Communications Inc. 1.6586 23.85 United Technologies Corp. 5.848 84.09 Verizon Communications Inc. 2.4396 35.08 Wal-Mart Stores Inc. 3.8924 55.97 Walt Disney Co. 1.6489 23.71 H.R. 3.59 As one part of a review of middle-manager selection procedures, a study was made of the relation between hiring source promoted from within, hired from related business, hired from unrelated business and the 3-year job history additional promotion, same position, resigned, dismissed. The data for 120 middle managers follow. Source Job History Within Firm Related Business Unrelated Business Total Promoted 13 4 10 27 Same position 32 8 18 58 Resigned 9 6 10 25 Dismissed 3 3 4 10 Total 57 21 42 120 a. Calculate job-history percentages within each source. b. Would you say that there is a strong dependence between source and job history? Env. 3.60 A survey was taken of 150 residents of major coal-producing states, 200 residents of major oil- and natural-gas–producing states, and 450 residents of other states. Each resident chose a most preferred national energy policy. The results are shown in the following SPSS printout. STATE COUNT ROW PCT COAL OIL AND OTHER ROW COL PCT GAS TOTAL TOT PCT OPINION 62 25 102 189 COAL ENCOURAGED 32.8 13.2 54.0 23.6 41.3 12.5 22.7 7.8 3.1 12.8 3

12 26

41 FUSION DEVELOP 7.3 29.3 63.4 5.1 2.0 6.0 5.8 0.4 1.5 3.3 8 6

22 36

NUCLEAR DEVELOP 22.2 16.7 61.1 4.5 5.3 3.0 4.9 1.0 0.8 2.8 19 79 53 151 OIL DEREGULATION 12.6 52.3 35.1 18.9 12.7 39.5 11.8 2.4 9.9 6.6 58 78 247 383 SOLAR DEVELOP 15.1 20.4 64.5 47.9 38.7 39.0 54.9 7.3 9.8 30.9 COLUMN 150 200 450 800 TOTAL 18.8 25.0 56.3 100.0 CHI SQUARE 106.19406 WITH 8 DEGREES OF FREEDOM SIGNIFICANCE 0.0000 CRAMER’S V 0.25763 CONTINGENCY COEFFICIENT 0.34233 LAMBDA 0.01199 WITH OPINION DEPENDENT, 0.07429 WITH STATE DEPENDENT.

a. Interpret the values 62, 32.8, 41.3, and 7.8 in the upper left cell of the cross tabulation.

Note the labels COUNT, ROW PCT, COL PCT, and TOT PCT at the upper left corner. b. Which of the percentage calculations seems most meaningful to you? c. According to the percentage calculations you prefer, does there appear to be a strong dependence between state and opinion? Bus. 3.61 A municipal workers’ union that represents sanitation workers in many small midwestern cities studied the contracts that were signed in the previous years. The contracts were subdivided into those settled by negotiation without a strike, those settled by arbitration without a strike, and all those settled after a strike. For each contract, the first-year percentage wage increase was determined. Summary figures follow. Contract Type Negotation Arbitration Poststrike Mean percentage wage increase 8.20 9.42 8.40 Variance 0.87 1.04 1.47 Standard deviation 0.93 1.02 1.21 Sample size 38 16 6 Does there appear to be a relationship between contract type and mean percent wage increase? If you were management rather than union affiliated, which posture would you take in future contract negotiations? Med. 3.62 Refer to the epilepsy study data in Table 3.18. Examine the scatterplots of Y 1 , Y 2 , Y 3 , and Y 4 versus baseline counts and age given here.

a. Does there appear to be a difference in the relationships between the seizure counts

Y 1 ⫺ Y 4 and either the baseline counts or age when considering the two groups treatment and placebo?

b. Describe the type of apparent differences, if any, that you found in a.

40 80 Base 120 160 25 Y4 50 25 50 100 50 Trt 1 20 25 Seizure counts versus age and baseline counts 30 35 40 80 40 Y3 Y2 Y1 Age Med. 3.63 The correlations computed for the six variables in the epilepsy study are given here. Do the sizes of the correlation coefficients reflect the relationships displayed in the graphs given in Exercise 3.62? Explain your answer. Placebo Group Y 1 Y 2 Y 3 Y 4 Base Y 2 .782 Y 3 .507 .661 Y 4 .675 .780 .676 Base .744 .831 .493 .818 Age .326 .108 .113 .117 .033 Treatment Group Y 1 Y 2 Y 3 Y 4 Base Y 2 .907 Y 3 .912 .925 Y 4 .971 .947 .952 Base .854 .845 .834 .876 Age ⫺ .141 ⫺ .243 ⫺ .194 ⫺ .197 ⫺ .343 Med. 3.64 An examination of the scatterplots reveals one patient with a very large value for baseline count and all subsequent counts. The patient has ID 207.

a. Predict the effect of removing the patient with ID 207 from the data set on the size of

the correlations in the treatment group.

b. Using a computer program, compute the correlations with patient ID 207 removed

from the data. Do the values confirm your predictions? Med. 3.65 Refer to the research study concerning the effect of social factors on reading and math scores. We justified studying just the reading scores because there was a strong correlation be- tween reading and math scores. Construct the same plots for the math scores as were constructed for the reading scores.

a. Is there support for the same conclusions for the math scores as obtained for the read-

ing scores? b. If the conclusions are different, why do you suppose this has happened? Med. 3.66 In the research study concerning the effect of social factors on reading and math scores, we found a strong negative correlation between minority and poverty and reading scores.

a. Why is it not possible to conclude that large relative values for minority and

poverty in a school results in lower reading scores for children in these social classes?

b. List several variables related to the teachers and students in the schools which may be

important in explaining why low reading scores were strongly associated with schools having large values of minority and poverty. Soc. 3.67 In the January 2004 issue of Consumer Reports an article titled “Cut the fat” described some of the possible problems in the diets of the U.S. public. The following table gives data on the increase in daily calories in the food supply per person. Construct a time-series plot to display the increase in calorie intake. Year 1970 1975 1980 1985 1990 1995 2000 Calories 3,300 3,200 3,300 3,500 3,600 3,700 3,900 a. Describe the trend in calorie intake over the 30 years. b. What would you predict the calorie intake was in 2005? Justify your answer by explaining any assumptions you are making about calorie intake.