In: Statistics and Probability
In a properly validated biotechnology process (i.e., it was proven that the process occurs in the same way each time it is run) three samples of the product are taken during a manufacturing stage to verify the percentage of the active ingredient. The researcher suspects that the first sample influences the result of the second and in turn the second influences the result of the third. For this reason, the researcher decided to compare lots to verify if the samples are independent. Table 1 presents the results of this experiment. Perform an analysis of variance to corroborate the independence between samples. Use a level of significance of 0.05. Carry out and present your calculations by hand (no Excel, no Minitab, etc.).
Table 1: Data Problem 1
Lot |
1st Sample |
2nd Sample |
3rd Sample |
A |
93.502 |
92.319 |
92.451 |
C |
91.177 |
92.230 |
92.097 |
D |
87.304 |
87.496 |
88.001 |
D2 |
81.275 |
80.564 |
80.386 |
G |
79.865 |
79.259 |
79.209 |
G2 |
81.722 |
80.931 |
81.351 |
H0: (The samples are not independent)
HA: (At least one mean is different from other 2 means)
From the given data, the following statistics are calculated:
1st sample | 2nd sample | 3rd sample | Total | |
N | 6 | 6 | 6 | 18 |
514.845 | 512.799 | 513.495 | 1541.139 | |
Mean | 514.845/6 = 85.8075 | 512.799/6= 85.4665 | 513.495/6 = 85.5825 | 1541.13918 = 85.619 |
44340.3869 | 44007.0946 | 44127.1807 | 132474.6622 | |
Std. Dev. | 5.7066 | 5.9993 | 6.0166 | 5.5526 |
From the above Table, ANOVA Table is calculated as follows:
Source of Variation | Sum of Squares | Degrees of Freedom | Mean Squares | F |
Between treatments | 0.3607 | 2 | 0.3607/2 = 0.1804 | 0.1804/34.9186 = 0.0052 |
Within treatments | 523.7783 | 15 | 523.7783/15 = 34.9186 | |
Total | 524.139 | 17 |
F = 0.1804/34.9186 = 0.0052
Degrees of Freedom for numerator = 2
Degrees of Freedom for denominator = 15
By Technology, p- value = 0.9949
Since p - value = 0.9949 is greater than = 0.05, the difference is not significant. Fail to reject null hypothesis.
Conclusion:
The data support the claim that the samples are not
independent.