Consider data on the treatment response of 12 mice from strain A and 9 mice from strain B.
Strain A: | 55.2 | 58.1 | 41.7 | 44.9 | 44.8 | 48.9 | sample mean = | 48.15 | ||
47.5 | 48.1 | 48.4 | 51.6 | 40.6 | 48.0 | sample SD = | 5.06 | |||
Strain B: | 48.7 | 52.6 | 65.2 | 70.4 | 44.2 | 54.7 | sample mean = | 55.90 | ||
44.0 | 66.5 | 56.8 | sample SD = | 9.70 |
Assume that the measurements from strain A are independent draws from a normal distribution with mean muA and SD sigmaA, and that the the measurements from strain B are independent draws from a normal distribution with mean muB and SD sigmaB.
Test the hypothesis H0: muA = muB versus the alternative Ha: muA != muB. (By "!=", I mean "not equal to".).
Calculate the P-value for the test.
What do you conclude?
Repeat the above problem for the one-tailed test of H0: muA = muB versus the alternative Ha: muA < muB
Consider data for some measurement on 6 mice before and after some treatment.
Mouse | 1 | 2 | 3 | 4 | 5 | 6 |
Before | 81 | 101 | 76 | 67 | 125 | 144 |
After | 138 | 210 | 162 | 105 | 259 | 319 |
Difference | 57 | 109 | 86 | 38 | 134 | 175 |
Does the treatment have an effect? Assume that the differences are independent draws from a normal distribution with mean mu and SD sigma.
Test the hypothesis H0: mu = 0 versus the alternative Ha: mu != 0.
Calculate the P-value for the test.
What do you conclude?
Last modified: Sun Feb 26 00:14:08 EST 2006 |