In: Statistics and Probability
A species of bird has beaklengths which are normally distributed and believed to have an average of 4.2 centimeters. A simple random sample of 5 birds on one particular island has the following beaklengths: 4.25, 4.31, 4.18, 4.36, and 4.4. Do an appropriate hypothesis test to decide whether the birds on this island have an average beaklength which is different than 4.2 centimeters.
We make the computations here as:
X | X - Mean(X) | (X - Mean(X))^2 |
4.25 | -0.05 | 0.0025 |
4.31 | 0.01 | 1E-04 |
4.18 | -0.12 | 0.0144 |
4.36 | 0.06 | 0.0036 |
4.4 | 0.1 | 0.01 |
21.5 | 0.0306 |
The sample mean and sample standard deviation here are computed as:
As we are testing here whether the mean is different from 4.2, therefore this is a two tailed test. The test statistic here is computed as:
For n - 1 = 4 degrees of freedom, we get the p-value from the t distribution tables for two tailed test here as:
p = 2P( t4 > 2.5565) = 2*0.0314 = 0.0628
As the p-value here is 0.0628 > 0.05 but < 0.1, therefore the test is significant at 10% level of significance, but not at 5% level of significance. Therefore only at 10% level of significance, we can conclude here that the mean of the population is different from 4.2