Question

1. The fifinal project of a class is assessed by a professor and two TAs. In section A, the

evaluators know a priori the current GPA of students. In section B, they do not.

(a) Which kind of experimental design is this? (5p)

(b) Which kind of bias may arise in project scores for section A? (5p)

(c) Which kind of treatment assignment procedure is applied for section B? (5p)

(d) Which section should have the most reliable project scores? Why? (5p)

Answer 1

a) This is a quasi-experimental design where the subjects (students in the 2 sections) aren't randomly assigned to groups (where group is one whose GPA is either known or not known).

b) For Section A, research bias/experimenter bias may arise since the experimenter knows about the subjects in the experiment which can impact the evaluation of the projects.

c) Random treatment assignment is applied for Section B, since the experimenters do not know about the subjects in Section B

d) Section B should have the most reliable project scores as there would not be any bias in the project evaluations given their grades are not known in advance by the Professor/TAs