In: Statistics and Probability
A programming team is in the process of testing a new software module. As part of the effort, they need to estimate the success rate of the module when used with a particular operating system. To do this, they plan to run the module on a randomly selected set of computers, record how many individual runs execute properly, and use that result to calculate the sample success rate (p-hat, the number of successes divided by the total number of tests). Assuming a confidence level of 99%, calculate n, the number of computers they need to use for the test in order to ensure a 0.03 margin of error in the success rate. Calculate n for the following two cases: (1) no assumption is made about the value of the sample success rate, and (2) in a recent test of a similar software module, that module ran successfully in 94% of the tests. Round your answers upward to the next higher integer.
(1) If no assumptions are made about the sample success rate, the sample size required to ensure a margin of error of 0.03 is n = .
(2) If it is assumed that the new module will run successfully roughly in 94% of the tests, the required sample size required to ensure a margin of error of 0.03 is n =
Solution,
Given that,
a) = 1 - = 0.5
margin of error = E = 0.03
At 99% confidence level
= 1 - 99%
= 1 - 0.99 =0.01
/2
= 0.005
Z/2
= Z0.005 = 2.576
sample size = n = (Z / 2 / E )2 * * (1 - )
= (2.576 /0.03 )2 * 0.5 *0.5
= 1843.27
sample size = n = 1844
b) = 0.94
1 - = 1 - 0.94 = 0.06
sample size = n = (Z / 2 / E )2 * * (1 - )
= (2.576 /0.03 )2 * 0.94 *0.06
= 415.84
sample size = n = 416