Você está na página 1de 6

Mateo Solano Stats 1510 Night Class Diet Soda Can Weight Analysis

Introduction In order to ensure that the manufacturer is not losing money by over filling their Product and at the same time not cheating their customers they must make sure that Each product has relatively the same weight once filled. The same holds true for filled cans of diet soda. Ideally the weights of any soda can should be relatively the same weight once filled with liquid. The purpose of this research was to show that no one brand of soda can when filled would on average weigh more or less than any other brand of soda can. A simple random sample was taken of the following brands of soda: diet coke, coke zero and diet Pepsi. Fifteen random numbers were generated using tcsats so that each population was represented equally. Each brand of soda had numbered cans where the population sample would be pulled from. The population of each of the brands was the entire number of cans of each brand. For Diet Coke the population was 85, for Diet Pepsi the population size was 96 and the population size for Coke Zero was 72. Once these numbers were generated each numbered can was then pulled from the population. Methods and Materials A simple random sample was taken of the following brands of soda: diet coke, coke zero and diet Pepsi. In addition to the cans of soda the materials used during this research a digital scale to weigh the cans and an Apple Ipad loaded with the program TCStats. Each can that was used was specifically chosen based on its assigned number which was randomly generated using the tcstats random data feature. In order to accomplish this 15 rows were set and then the upper bound number was set to 15 and the lower bound number was set to 1 and then the numbers were generated beginning in row 1 and ending in row 15. Once the cans were pulled

from the population each can was weighed on a digital scale and recorded in grams to the nearest hundredth. The summary statistics were then generated for each of the sample populations of soda cans. In addition to the summary statistics a KruskalWallis test was performed to verify if at least one of the weighted averages for the diet soda samples was different from the rest. Results The summary statistics were as follows for each sample population of soda can weights: Table 1.1 lists the five number summary for Coke Zero which was 368.28 grams for the minimum, 370.32 grams for Q1, 371.75 grams for the Mean, 373.27 grams for Q3 and 375.34 grams for the Max. For Diet Pepsi the five number summary in Table 1.1 lists the figures as follows: 365.71 grams for the minimum, 367.26 for Q1, 369.26 for the Mean, 371.17 grams for Q3 and 373.78 grams for the Max. Finally Table 1.1 lists the five number summary for Diet Coke with the Minimum being 367.98 grams, Q1 being 370.62 grams, 371.28 grams for the Mean, 373.94 grams for Q3, and 374.26 grams for the Max. The box and whisker plots for each of the diet soda samples are graphed individually in Figure 1.1. You can see by the graphs that the Coke Zero and Diet Pepsi have normal distributions but the graph shows that Diet coke does not have a normal distribution. This abnormal distribution is shown by the normal plot graph for the Diet Coke sample data. This graph is displayed as figure 2.1. Based on this plot a Kruskal-Wallis test was performed which resulted in a p value equal to .0104 and so Ho was rejected indicating that at least one of the diet soda sample means was different from the other two.

Table 1.1

Figure 1.1

Figure 2.2 Diet Coke Normal Plot

Appendix Raw Data


diet coke # 1 3 4 5 6 11 15 24 27 39 49 53 57 64 81 d. coke weight D.Pepsi C. C. Zero D.Pepsi Weight Zero Weight 370.89 15 368.69 6 372.14 371.21 21 369.13 10 370 374.26 45 366.44 12 370.37 367.99 52 373.78 14 368.28 370.8 55 367.72 18 372.11 373.94 56 370.42 28 372.52 368.83 60 371.17 33 373.61 374 62 367.26 34 373.4 371.26 73 370.22 46 373.27 371.09 74 371.64 48 375.34 371.43 78 372.1 50 370.81 370.8 80 366.81 56 369.95 370.62 81 369.56 58 371.99 367.98 89 368.25 68 370.32 374.06 92 365.71 70 372.11

Você também pode gostar