There are two samples on this page.

Sample 1 describes a positive correlation in a study of student performance at a Texas community college.

Sample 2 describes a negative correlation in a study of the relations among different aspects of personality and life events.

Sample 1

Colleges and other large organizations need to make predictions about future trends. One way to do this is using correlational methods. A Texas community college analyzed student data, in order to predict the needs of future students. One question the college administration had was what is the relation between class registration from one semester to the next? In other words, could they predict how many credits students would take in the future from their current load?

They had a sample of 322 students from the prior year. There was a positive correlation (.64) between the number of credits students took in the fall and the number of credits students took in the spring. Students who took more credits in the fall also tended to take more credits in the spring (or, students who took few credits in the fall also tended to take few credits in the spring). 

This table shows a representative sample of 10 students, but keep in mind that they had 322 students in their sample.

The scatterplot below shows the data from all 322 students in the sample.

Sample 2

Personality researchers study relations among personality variables and various behaviors. In this example, psychologists studied a sample of 104 adolescents in a Midwest town. The researchers assessed a personality variable called “agreeableness” using a standard personality questionnaire. This is a measure of how nice that person is to be around. The questions ask about how cheerful, stubborn, polite, bossy, and cooperative the person is. Each adolescent’s average of those items is reported.

They also created a measure of behavior problems. The youths reported on various behaviors in the last six months, including cheating, swearing, stealing, and fighting. Each adolescent’s sum of behavior problems is reported.  

The researchers found a negative correlation (-.66) between agreeableness and behavior problems. Adolescents who were more agreeable tended to have fewer behavior problems (or, adolescents who were less agreeable tended to have more behavior problems).

This table shows a representative sample of 10 participants, but keep in mind that they had 104 adolescents in their sample.

The scatterplot below shows the data from all 104 adolescents in the sample.