Lost to Theft (decoy) Lost to Theft (control)

49,008 84,738

70,051 66,989

43,307 35,901

34,098 58,785

33,323 31,704

29,086 21,000

87,441 97,757

45,462 20,645

27,698 55,593

33,904 42,484

87,690 84,644

41,438 22,884

55,898 103,251

55,570 77,108

46,439 75,602

67,140 87,128

51,860 89,755

73,894 92,615

34,406 37,412

40,981 60,773

A retail chain is considering installing devices that resemble cameras to deter shoplifting. The devices only look like cameras, saving the expense of wiring and recording video. To test the benet of this decoy system, it picked 40 stores, with half to get the decoy and the other half to serve as comparison group (control group). Stores were matched based on typical levels of sales, local market size, and demographics. The comparison lasted for 3 months during the summer. At the end of the period, the retailer used its inventory system to compute the amounts lost to theft in the stores. Complete parts (a) through (c) below. a Click the icon to view the data. (a) Why is it important that all of the stores measure theft during the same time period? It is important to measure during the same time period because thieves tend to strike more than once in quick secession. B. It is important to measure during the same time period because theft patterns might change during non-operational hours. It is important to measure during the same time period because theft patterns might change during a major sale promotion at a given store. .9 D. It is important to measure during the same time period because theft patterns might change with the season, such as during specic holidays. (b) Compute separate 95% condence intervals for the amount lost to theft with and without the decoy cameras. Is there evidence of a statistically signicant difference? The 95% condence interval for the expected number of thefts with the decoy cameras is D to CI. The 95% condence interval for the expected number of thefts without the decoy cameras is D to D. (Round to two decimal places as needed.)