A statistics professor conducted a study on the effect of playing video games on a student’s test scores. On the first day of class, she randomly divided her classes into groups. Group one played video games 2 hours a day, group two played video games 4 hours a day, and group three did not play video games at all. After three weeks the students took their first test. The professor then compared the test scores of the three groups to see if the game playing made a difference in the students’ performance on the test.

Question

Identify at least one potential bias in the sampling technique.

Answer 1

The experiment was not 'double blind" because the same person who compared the test scores (the professor) was aware of the reason for the experiment.

If the groups were aware of the reason for them playing the games (possible affect on test scores) this would be another potential bias.

Answer 2

Since there are several classes, were the conditions randomly assigned within each class or by each class? If by each class, varying conditions between each class could be a confounding variable. Even teaching various sections of the same class can lead to different conditions.

If the test was objective and tests were machine scored, scoring would not be the need for double blind, but rather what happened in the classroom before testing would need double blind conditions.