The Question at the end of the video describes a case where two groups of students(group A and B) are selected and undergo different teaching methods to then compare the performance from each group. The answer to the question suggests that a Paired t-Test should be used. However, according to the description, group A and B contain a different set of students and do not share students, therefore, the samples are independent and the suggested t-Test should be the two sample t-test and not the paired test.
Hello @dalapiz !
The Paired t-test is used when data is in the form of matched pairs (i.e. not independent).
If the researcher is interested in comparing the different teaching methods, then a two-sample t-test should be used. However, what is being compared in the question is the students’ performance before and after they undergo a particular teaching method, not the teaching methods themselves.
Hence,
The researcher is interested in comparing the individual changes in math skills within each group. Therefore, the appropriate statistical test to use is a paired t-test.
I’m hesitant to accept your explanation.
At the very least, the description could be improved to make it more clear that the goal is to separately test:
- If group A showed an improvement
- If group B showed an improvement
At least to me, and again, based on the description, it seems more intuitive to perform a two-sample t-test between Group A and Group B, both before and after the teaching method was changed. This would be a direct comparison of the control group vs the experimental group.
By performing a paired test, the way how you suggest it, you don’t really need the control group A, this because the group-B before the change serves as the control when compared to group B after the change.
Daniel
Hi @daniel,
The paired t-test is indeed the correct choice in this scenario because the comparison is not between independent groups A and B. Instead, the goal is to analyze within-group changes: how each student’s performance changes from “before” to “after” the teaching method. The key is that for each student, there is a before and after measurement, making the samples paired.
The difference here is that while you’re comparing the performance within each group (A and B), you don’t need to compare the two groups directly at the same time. Instead, the paired t-test focuses on the difference in scores for each individual (before - after), and this difference is what you’re testing across the students in both groups.
By using the paired t-test on the difference in scores for each group (A and B), you can then determine if there was a significant improvement within each group. If you were to use a two-sample t-test, you’d be comparing two independent groups (A vs B) without considering the before-and-after changes, which wouldn’t capture the improvements effectively within each group.
Hope this helps! Feel free to ask if you need further assistance.