Question

1. I have hypothesized that eating a high-sugar breakfast will lead to low exam performance in the afternoon. What is the null hypothesis?

2. Explain when I would reject and retain a null hypothesis.

3. Give an example of when I would use an independent t-test and an example of when I would use a dependent t-test.

4. How do I know whether to use a t-test vs. a Chi-Square?

Answer 1

1. Null hypothesis: Eating a high-sugar breakfast will not affect exam performance in the afternoon.

2. We'll reject the null hypothesis if the test statistic or the average exam score falls below a certain critical value.

3. You can use an independent t-test when there are two different groups, one group is not given high-sugar breakfast while the other group is given.

You should use a dependent or paired t-test when the exam scores are taken from the same students, first without giving high-sugar breakfast and then after giving high-sugar breakfast.

4. But the t-test and Chi-square test give the same result when you're comparing two groups in a two-tailed test. But if the test is one-tailed like here, then we should we using the t-test. Also if the number of groups is more than two then, we should be using the chi-square test.