Click on image for full size
Courtesy of Freeze Clip Art
Related links:
Correlations in Science
In conducting their research, scientists often want to know if two sets of data (variables) are related to each other. For instance, you might wonder if the amount of time a student spends reading the Windows to the Universe website is related to the grade that student gets in his or her science classes. How would you test this, and how would you express it in a way that would clearly tell other people what sort of relationship there is between these two variables?One of the most common ways a scientist does this is by using a concept called correlation. Correlation is basically a measurement of how independent two different variables are, and is usually calculated using a formula that results in a coefficient of correlation ranging from -1 to 1.
A correlation of -1 indicates that the two variables are inversely related, and that as one variable increases the other always decreases. For example, the total sales in a given day for an ice cream truck and the total snowfall for that same day might have a correlation close to -1. On days with lots of snow, not many people are buying ice cream from the truck, and on days where the ice cream truck’s sales are really high, it’s probably not snowing. A correlation of 1, on the other hand, indicates that the two variables are directly related, and that as one variable goes up the other does also. For example, the amount of time a basketball player spends practicing is usually closely related to the number of points he or she scores in games, and this relationship would probably have a correlation coefficient close to 1.
Many times a calculated correlation will be close to 0, and this indicates that there is no obvious relationship between the two variables (there may still be a relationship; in some rare cases two variables can be closely related but have a correlation coefficient of 0). It’s important to remember that even when two variables are correlated, this does not mean that a change in one variable causes the other one to change—it just means that they’re related. For instance, when it’s raining you can see people using umbrellas a lot more often, and you can see cars using their wipers a lot more often. So umbrella use and windshield wiper use are correlated, but neither causes the other—we don’t use umbrellas because other people are using wipers, or vice versa. We use both because it’s raining.
