How to Find the Variance of a Data Set in Statistics

Variance offers a measurement of reliability to a calculated average by describing how spread out the data is. As an example, an average of 50 could be calculated as the average of 49, 50 and 51, or 1, 2, 98 and 99. In the latter data set, the observed values have little similarity with the calculated average. Likewise, you cannot expect future observations to fall close to the average, because the previously observed values are so different than the average. Variance allows you to understand how representative the average is to the original data set, even without looking at the original data.

Instructions

    • 1

      Sum all the data values and divide by the number of observations to compute the average. In the introduction's examples, both data sets have an average of 50. The first is calculated as 49 plus 51 plus 50, and then dividing the 150 result by 3. The second is calculated as 1 plus 2 plus 98 plus 99, and then dividing the 200 result by 4.

    • 2

      Subtract each data point in the set from the average and then square the differences. In the first example, the differences between each data point is -1, 0 and 1. Squaring those differences gives you 1, 0 and 1. In the second example, the differences are -49, -48, 48 and 49. Squaring each difference gives you 2401, 2304, 2304 and 2401.

    • 3

      Sum the squared differences. In the first example, adding 1 plus 0 plus 1 gives you a total of 2. In the second example, the total is 9,410.

    • 4

      Divide the total by the number of data points to compute the variance. In the first example, dividing 2 by 3 gives you a variance of 0.667. In the second example, dividing 9,410 by 4 gives you a variance of 2,352.5. The considerably larger variance in the second example demonstrates that the data used when calculating the average was immensely diverse.

Learnify Hub © www.0685.com All Rights Reserved