Home >> Science & Nature >> Science

What Is a Mesokurtic Curve?

Suppose you were graphing the performance of a large number of students on a standardized test. Most students will fall close to the middle. Only a few will do significantly better or worse than average. The number who do exceptionally better or worse will be smaller still. The result is a curve that starts off with a low value, indicating the small number who did very poorly, rises to a peak to indicate the majority of students with an average performance, then slopes off again to show the relatively small number who do above average, declining further to reflect the even smaller number who did exceedingly well. Statisticians describe such a curve as platykurtic, leptokurtic or mesokurtic depending on its shape.

Platykurtic Curves
- If a curve is relatively flat, it is termed platykurtic. In a platykurtic curve describing test scores, the number of students who performed significantly better or worse than average is relatively large. If the average score were 50, there would still be a large number of students who scored less than 40 or more than 60, for example.
Leptokurtic Curves
- If the curve rises extremely sharply and slopes off just as sharply, that's a leptokurtic curve. If a leptokurtic curve described test scores, then the majority of students would score close to average. If the average score were 50, the number of students scoring less than 40 or more than 60 would be very small. The vast majority of scores are close to the average in a leptokurtic curve.
Mesokurtic Curves
- A mesokurtic curve is between a leptokutic and a platykurtic curve. It is neither exceptionally flat, nor exceptionally peaked. A mesokurtic curve describing test scores with an average of 50 will have about the same number of students scoring below 40 and over 60. This amount would be greater than for a leptokurtic curve, but would be less than would be seen in a platykurtic curve.
The Bell Curve
- People sometimes use the term "bell curve" to describe a mesokurtic curve, because that's what it looks like. You may also see the terms "normal curve" and "Gaussian distribution" which mean essentially the same thing. When you gather a set of data, such as scores on a test, you can establish a variable called standard deviation, which quantifies how much the individual scores deviate from the mean, on average. In a normal distribution, about two-thirds of the scores fall between +1 and -1 standard deviations from the mean. In a leptokurtic curve, more than two thirds of scores would be within one standard deviation. In a platykutic curve, less than two thirds would be within one standard deviation from the mean.

Science