Psychology: Statistical Dos and Don'ts

| No Comments

Antonio Sundquist
Psych 1001
Blog 1

It is often found in research, newspapers, magazines, etc, that a certain event happens that causes the other to occur. This can indeed happen, however, in statistics the only time we are allowed to say that event A was caused by event B, is if we have conducted an experiment where there was a control group, and treatment was allocated amongst another group. Frequently, when conducting and interpreting studies where two variables appear to be associated with one another people make the mistake of claiming event A was a result of event B, when in reality, correlation is NOT causation - that is lesson number one in statistical inference.
A common example of a situation where one event appears to be a result of the other is the many studies that show an association between violence in the media and aggressive behavior. In essence, one would hypothesize that more exposure to violence through things such as: music (lyrics), television, movies, and the media, may suggest a larger probability of displaying aggressive behavior. Many studies and experiments have been conducted with similar hypotheses. Although the results vary, the majority of the data appears to suggest an association between the amount of violence exposure and the display of aggressive behavior.
The Adults & Children Together Against Violence Foundation (ACT) is an organization driven to help children and young adults with violent issues in society. On their website, they speak on behalf of
"the first longitudinal study of violent media on adolescents.... This study remains unique because it links exposures to violent media after childhood-i.e., among teenagers - and then follows them into young adulthood when aggressive behaviors actually take place" (Johnson, J.G., et al)
The data collected from the studies resulted in a positive correlation between the two variables. Now, based on the data and a positive association between our two variables, it would seem probable to restate our hypothesis and conclude that exposure to violence causes aggressive behavior. However, we must relate back to the golden rule of statistics, just because two variables appear to have an association or correlation, that doesn't permit us to say that one causes the other.

Here is a short video on how our culture is being shaped by violence in the media. Pretty interesting stuff!

"Special Topics: Media Violence and Aggression." Adults and Children Together Against Violence Web Site. Web. 02 Oct. 2011.

Find recent content on the main index or look in the archives to find all content.