Making statistical calculations can get complicated. It's not just means and averages that get taken into consideration when doing a statistical calculation – it's the "weighted" means and variances that need to be considered. Weighted variances help take more data into account when doing a calculation so that you get the most accurate result possible.

## Understanding the Weighted Variance

In most statistical analysis exercises, each data point carries equal weight. However, some include data sets in which some data points carry more weight than others. These weights can vary due to various factors, such as the number, the dollar amounts or the frequency of the transactions. The weighted mean allows managers to calculate an accurate average for the data set, while the weighted variance gives an approximation of the spread among the data points.

## How to Calculate the Weighted Mean

The weighted mean measures the average of the weighted data points. Managers can find the weighted mean by taking the total of the weighted data set and dividing that amount by the total weights. For a weighted data set with three data points, the weighted mean formula would look like this:

[(W_{1})(D_{1}) + (W_{2})(D_{2}) + (W_{3})(D_{3})]/ (W_{1}+ W_{2}+ W_{3})

Where W_{i} = weight for data point i and D_{i} = amount of data point i

For instance, Generic Games sells 400 football games at $30 each, 450 baseball games at $20 each, and 600 basketball games at $15 each. The weighted mean for dollars per game would be:

[(400 x 30) + (450 x 20) + (600 x 15)]/[400+500+600] =

[12000 + 9000 + 9000]/1500

= 30000/1500 = $20 per game.

## How to Calculate the Weighted Sum of the Squares

The sum of the squares uses the difference between each data point and the mean to show the spread between those data points and the mean. Each difference between the data point and the mean is squared to give a positive value. The weighted sum of the squares shows the spread between the weighted data points and the weighted mean. The formula for the weighted sum of squares for three data points looks like this:

[(W_{1})(D_{1}-D_{m})^{2} + (W_{2})(D_{2} -D_{m})^{2} + (W_{3})(D_{3} -D_{m})^{2}]

Where D_{m} is the weighted mean.

In the example above, the weighted sum of the squares would be:

400(30-20)^{2} + 450(20-20)^{2} + 600 (15-20)^{2}

= 400(10)^{2} + 450(0)^{2} + 600(-5)^{2}

= 400(100) + 450(0) + 600(25)

= 400,000 + 0 + 15,000 = 415,000

## How to Calculate the Weighted Variance

The *weighted variance* is found by taking the weighted sum of the squares and dividing it by the sum of the weights. The formula for weighted variance for three data points looks like this:

[(W_{1})(D_{1}-D_{m})^{2} + (W_{2})(D_{2} -D_{m})^{2} + (W_{3})(D_{3} -D_{m})^{2}] / (W_{1}+ W_{2}+ W_{3})

In the Generic Games example, the weighted variance would be:

400(30-20)^{2} + 450(20-20)^{2} + 600 (15-20)^{2} / [400+500+600]

= 415,000/1,500 = 276.667

If that all seems too complicated, you can use a calculator or spreadsheet to help you calculate weighted variance. The calculation for weighted variance can help you get a more accurate picture of certain aspects of your business. It can be used to strengthen your sales pipeline, better diversify investments and know which parts of your business add more to profits.

References

Writer Bio

Living in Houston, Gerald Hanks has been a writer since 2008. He has contributed to several special-interest national publications. Before starting his writing career, Gerald was a web programmer and database developer for 12 years.