Data Analysis & Research Methods Essay Example | Topics and Well Written Essays - 3750 words

Data Analysis & Research Methods - Essay Example 263). Apparently, ‘r’ is very close to zero, which means that the two variables have no relationship. In other words, support for the new waste policy has no statistical relationship with the volume of waste produced prior to the policy’s introduction. The other inference drawn from the analysis is that r is positive, meaning that â€Å"as one variable gets larger the other gets larger† (Conover, 1980, p. 263). However, since we have concluded that the two variables have no statistical relationships, then the positive relationship is spontaneous and hence has no scientific meaning (Croxton, Cowden & Klein, 1968). To better understand the correlation of coefficients, it is important to square them to get the proportion of the disparity in one variable that is associated with the variation in the other variable. In this case, the percentage is extremely small (0.00002%), which again adds evidence that the two variables have no relationship. These statistics r eveals that the volume of waste that the staff generated prior to introduction of the new policy does not have any influence in the manner in which they embrace the new policy. Embracing of the new policy, in this case, is determined by question six, which asks â€Å"How many times per week do you visit the central re-cycling bins to deposit waste of the central waste bins?† Question A (b) Assumption: 3 = yes, close enough, 1 = No, too far away (this means that a positive correlation should imply that support for new policy in positively related with convenience. respondent Question 4 Question 6 X Y X*X Y*Y X*Y 1 3 3 9 9 9 2 3 12 9 144 36 3 1 6 1 36 6 4 1 4 1 16 4 5 1 5 1 25 5 6 3 7 9 49 21 7 1 8 1 64 8 8 1 6 1 36 6 9 1 17 1 289 17 10 1 21 1 441 21 11 1 2 1 4 2 12 3 7 9 49 21 13 3 5 9 25 15 14 1 5 1 25 5 15 1 8 1 64 8 16 3 4 9 16 12 17 1 9 1 81 9 18 3 12 9 144 36 19 1 19 1 361 19 20 3 13 9 169 39 21 3 4 9 16 12 22 1 15 1 225 15 23 1 17 1 289 17 24 3 7 9 49 21 25 1 10 1 100 10 26 3 3 9 9 9 27 3 12 9 144 36 28 1 6 1 36 6 29 1 4 1 16 4 30 1 5 1 25 5 31 3 10 9 100 30 32 1 8 1 64 8 33 1 6 1 36 6 34 1 10 1 100 10 35 1 23 1 529 23 61 313 139 3785 511 N= 35 ?X = 61 ?Y= 313 ?XY =511 ?X2 =139 ?Y2=3785 Null hypothesis: there is enough evidence to show that support for the new waste policy was related to the convenience of the re-cycling bins’ location? Perform statistical tests to investigate these questions fully. Correlation(r) = [N?XY - (?X) (?Y) / v ([N?X2 (?X) 2] [N?Y2 - (?Y) 2])] r = [35*511-61*313/v ([35*139*3721] [35*3785-97,969)] r = -1208/790348.4 r = -0.002 r2=0.004 = 0.4% The coefficient of correlation between convenience of the re-cycling bins’ location and support for the new waste policy is -0.002, which means that when the location is more convenient, less people will support the policy. However, the coefficient is quite far from -1, therefore, the two variables are not closely related. The square of the coefficient is 0.4%, which mea ns that only 0.4% of the variance accounts for this relationship, while the rest 99.6% is due to sampling errors. In reality, it would be hard to find people failing to support the policy when it is more convenient, while they support it when it is less convenient. Therefore, we could conclude that the statistical relationship is largely because of sampling errors, and hence not a true representative of what could happen in the population (Rosner, 2000). Is there any evidence