## Question 345:

1

While there is no official list of five steps, as each statistics text will name or categorize them a bit differently they all generally look something like this:

1. Decide on the hypothesis to test (directional or non-directional)
2. Select the statistical test and alpha level (often .05)
3. Find the critical value based on alpha
4. Gather data and calculate the sample statistic and p-value
5. Draw a conclusion and interpret (reject or fail to reject null):

Here is an example:

Do students from private schools score better/worse on a standardized test than public school students?

1. We're testing to see if there is evidence that private schools have different scores than public schools.
2. The Null hypothesis is that there is no difference between the scores.
3. The Alternative Hypothesis is that there is a difference in scores (maybe higher, maybe lower).
4. We will reject the Null hypothesis if we have evidence that mean scores are different and the difference is less than 5% due to chance alone. This is alpha and the p-value.
5. To test the hypothesis we sample several scores from the same test that were taken by public school students and private school students, then we calculate the mean and standard deviation for each group.
6. We run a test (probably a t-test) on the means, depending on the evidence we either reject or fail to reject the Null Hypothesis.