Who we chose as a participant, is more important than what we think!

One thing that almost all psychology students have experienced in McGill University is the class presentations on participation in psychological studies and earning 2% bonus. In my second year, I participated in a study where they grouped 12 psychology students into two groups of six where we had to do a group activity. At one point during the study, the examiner informed our group that, “I am going to photocopy these forms and will be back in few minutes”. As soon as she left the room, the atmosphere got awkward. We were all sitting in a circle, smiling and looking at each other as if we knew what is going on. After few seconds one of the students, while he was trying to hide his laugh, whispered, “This is so obvious though”. I looked over the circle and confirmed, “Totally”. “I wonder where they have hide it though!” said a student who was looking at the walls. This was followed by a short pause and a long loud group laugh.

The examiner had left us alone intentionally to see how we, as a group, interact with one another. However, as psychology students we were all suspicious from the moment she left. After we all confessed that we know that they have left us alone intentionally, we tried to guess where they have hided the hidden camera in the room. This behavior, which was the result of our basic knowledge of psychological testing, could affect the data and make them bias.

Sampling bias is a bias that happens when some members of intended population are less likely to be part of the sample. Many psychological studies in McGill University especially one year long honors projects, due to lack of time or funding, use convenient sampling and rely on psychology students as participants. This make their sample less representative of general population. Knowledge of psychology students regarding psychological testing and procedure is different than other people in the society. As a result, their responses and behaviors might be different from general population and thus reducing the external validity of data.

Therefore, one issue that researchers face in psychological testing is collecting a sample that is representative of their intended population. Low external validity of data really questions the conclusion reached by psychologists.

p