In: Statistics and Probability
A laboratory manager wants to ensure that 95% of the carbon
analysis measurements made in the laboratory are within 0.1 ppm of
the true value. On the basis of experience, the manager is willing
to assume that the carbon content measurements made on identical
soil samples are approximately normally distributed with mean µ and
standard deviation ? . Recall that 95% of the measurements in a
normal population lie within 1.96 (about 2) standard deviations of
the mean. The requirement will be met if 2 0.10 or 0.05 ?? ?? .
Based on the ten data values below (ppm) use the chi-square test
for a single variance to see if the measurement precision can be
met with the instrumentation being used. Recall the chi-square test
is set up for the variance, not the standard deviation. Use a Type
I error of 0.05.
0.560 0.842 0.731 0.782 0.673 0.718 0.791 0.726 0.760 0.798
= 0.738, s^2 = 0.0062
H0: = 0.01
H1: 0.01
The test statistic = (n - 1)s^2/
= 9 * 0.0062/0.01 = 5.58
At alpha = 0.05, the critical values arre = 2.7004
= 19.0228
AS the test statistic value lies between the critical values (2.7004 < 5.58 < 19.0228), so the null hypothesis is not rejected.
So at alpha = 0.05, we can conclude that the measurement percision can be met with the instrumentation being used.