Discussion Post: Choosing the Right T-Test
Research Scenario
I am investigating the effectiveness of the Grammarly application on the writing quality of undergraduate ...
Discussion Post: Choosing the Right T-Test
Research Scenario
I am investigating the effectiveness of the Grammarly application on the writing quality of undergraduate students at HUFLIT university. Specifically, I want to know if there is a significant difference in the number of grammatical errors students make in an essay before using the software compared to the number of errors they make in a second essay after using the software for one month.
Test Choice: Paired Samples T-Test
-
Number and Type of Groups: There is one group of participants (undergraduate students), but they are being measured at two different points in time.
-
Data Source: The data comes from the same participants. Each student provides two scores: a "Pre-test" error count and a "Post-test" error count.
-
Reasoning: Since I am comparing the means of two sets of scores that are "linked" to the same individuals, the paired samples t-test (also known as a dependent t-test) is the most appropriate. This test accounts for individual differences in writing ability; it focuses specifically on the change within each student rather than comparing two entirely different groups of people.