I am reading this link [https://statistics.laerd.com/spss-tutorials/mann-whitney-u-test-using-spss-statistics.php][1] in which they describe the 4 assumptions of applying U-test.
I am not clear about the terms "dependent variable" vs "independent observations". If there are dependent variables, why the observations can be independent.
An example: I would like to see if two teaching methods A and B are different. I teach the class-A (of 80 students) using method-A and teach class-B (of 80 students) using method-B. Note, there is no student from class-A attend class-B and vice verse. Then, at the end of the term, I collect their exam scores into two groups. Can I apply U-test on the two groups of data in this example?
I am not sure, because the two groups of exam scores are still dependent: "I" as the teach is a dependent factor for the two group of scores, and the exam is the same for both groups.