Psychology experiment

Question

1. Explain why “a correlation does not prove causation.” 
2. Design an experiment that will examine the question, “Who will learn more about psychology, those students attend face- to-face lecture or those watch recorded lecture video?”

Answer 1

Explain why correlation does not prove causation. I wonder why this even gets asked. It should be the first thing taught in any statistics or science class. Sure, where there is causation, there will be correlation. The speed of an object is defined as the time it takes to move a set distance. Speed = distance / time. You find that an object covers a distance in a shorter time then the speed must have increased. Increasing speed will decrease time. (for same distance). Causation. But this is not always true. Sometimes we find a correlation between two measures, but we don't know which is a result of the other. The mere fact of correlation does not itself point the direction of the relationship.

Then there is the case where correlation has nothing to do with causation in either direction. Simple thought experiment:

Some plants are set up in greenhouse and oxygen levels and starting at 10 am the temperature are measured every 6 hours or so for 54 hours. It is found that when the temperature was the lowest, the plants gave off carbon dioxide and when the warmest , they gave off oxygen. Close correlation. Zero causation. The sun (not measured) caused warming and photosynthesis, no sun, cooler temperatures and cellular respiration but the temperature had nothing to do with photosynthesis. It was the sun.

Part B - Keep the experiment simple. Use a reasonably large group of diverse students who voluntarily enroll in a psychology course. Randomly select students from the incoming group to belong to one of two cohorts which will be kept separate from each other. Select two reasonably short issues to be taught that are relatively independent of one another (ex The Psychology of Bullying", and "The Psychology of Weird Beliefs".).

Phase I - have both cohorts study Psychology of Bullying, Cohort A by in person lecture with interaction and question and answer time. Cohort B will be presented with a video lecture. After the topic is 'taught', both cohorts get the same test and results are tabulated.

Phase II - have both cohorts study The Psychology of Weird Beliefs, Cohort B by in person lecture with interaction and question and answer time. Cohort A will be presented with a video lecture. After the topic is 'taught', both cohorts get the same test and results are tabulated.

You now have comparison information for all the students.

1. By teaching totally different topics in Phase I and Phase II you control for knowledge bleed over from previous training.
2. By not letting the students know before the class that the lecture method was going to be in person or video lecture or that the other cohort was going to be treated differently you reduce the effect of pre-disposition on the students parts.
3. By using both teaching methods on all the cohorts, you reduce the effects of un documented differences in the composition of the cohorts (which should have already been minimized by randomization)

Although there might be some interaction in the willingness of students who had face to face instruction first to go to video vs the students who did video first and then face to face, the design is such that this interaction will be able to be measured.