In: Statistics and Probability
A manufacturer has recorded its cost of electricity (cost) and the total number of hours of machine use time (time) each week for a total of 52 weeks. Use the Mod7-1Data (Links to an external site.) to answer the following questions. USE α = 5%!
Time | Cost |
7 | 886.44 |
14 | 1598.05 |
15 | 1385.82 |
8 | 594.93 |
8 | 792.09 |
15 | 1307.7 |
13 | 1334.7 |
12 | 900.14 |
11 | 1306.41 |
11 | 950.22 |
13 | 1249.98 |
14 | 1388.16 |
6 | 750.87 |
14 | 1185.85 |
8 | 683.89 |
12 | 1245.1 |
14 | 1342.55 |
15 | 1154.05 |
11 | 940.88 |
14 | 992.5 |
8 | 886.89 |
13 | 980.66 |
14 | 1171.71 |
10 | 1185.57 |
11 | 987.64 |
11 | 1035.9 |
11 | 902.04 |
11 | 1195.62 |
8 | 1001.72 |
13 | 1421.23 |
9 | 684.21 |
8 | 497.48 |
12 | 1043.28 |
15 | 1054.48 |
11 | 1043.03 |
9 | 876.73 |
6 | 833.68 |
8 | 815.6 |
15 | 1396.96 |
11 | 1103.17 |
7 | 464.26 |
12 | 1159.15 |
9 | 959.89 |
11 | 1171.85 |
8 | 740.38 |
7 | 1001.72 |
11 | 865.59 |
13 | 1237.42 |
9 | 1295.29 |
12 | 1527.08 |
10 | 962 |
7 | 688.7 |
sample covariance=
sample correlation=
null hypothesis=
computed test statistic=
alternative hypothesis=
statistical conclusion=
1. For the given data, the covariance of Time and cost can be computed using the formula:
Using excel,
We get Covariance between Time and Cost = 498.902
2. For the given data, the linear relation between of Time and cost (two continuous variable) can be computed using Pearson's Correlation (r) obtained using the formula:
Using excel,
We get r = 0.737
3. Now, we have to test whether this correlation coefficient is significant at 5% level. Le denote the population correlation coefficient.We have to test:
Null hypothesis: Vs
Alternative hypothesis: (Two sided test)
The appropriate statistical test to test the above hypothesis would be a t test, with test statistic given by:
Computing the test statistic:
We get t statistic = 7.718
Computing the p-value: (We would determine whether the p-value is less than the significance level, if yes, the null hypothesis is rejected)
Since, p-value = 0.000 < 0.05, we may reject H0 at 5% level. We may conclude that there exists a significant relationship between Time and Cost. Also, from the sample correlation coefficidnt, we find that the significant relationship is strong and positive.