In: Statistics and Probability
The following data gives the number of hours 10 students spent studying and their corresponding grades on their midterm exams.
Hours Spent Studying | 0 | 0.5 | 1 | 2 | 2.5 | 3 | 4 | 4.5 | 5 | 5.5 |
---|---|---|---|---|---|---|---|---|---|---|
Midterm Grades | 60 | 63 | 75 | 81 | 84 | 87 | 90 | 93 | 96 | 99 |
Determine if r is statistically significant at the 0.01 level.
Solution:
n = 10
X | Y | XY | X^2 | Y^2 | |
0 | 60 | 0 | 0 | 3600 | |
0.5 | 63 | 31.5 | 0.25 | 3969 | |
1 | 75 | 75 | 1 | 5625 | |
2 | 81 | 162 | 4 | 6561 | |
2.5 | 84 | 210 | 6.25 | 7056 | |
3 | 87 | 261 | 9 | 7569 | |
4 | 90 | 360 | 16 | 8100 | |
4.5 | 93 | 418.5 | 20.25 | 8649 | |
5 | 96 | 480 | 25 | 9216 | |
5.5 | 99 | 544.5 | 30.25 | 9801 | |
Sum | 28 | 828 | 2542.50 | 112 | 70146 |
Putting values , we get
r = 0.9703
Now , we have to test the significance of correlation coefficient.
Use = 0.01
n = 10
df = n - 2 = 10 - 2 = 8
Using the critical value table for Pearson correlation coefficient, (two tailed )
Critical value are 0.765
r = 0.9703
| r | = | 0.9703| = 0.9703
r > 0.765
reject H0
significance correlation.
YES , there is sufficient evidence to conclude that there is significant linear correlation between the two variables.