Tip of the week: medians (for Advanced Analytics)

Published October 21, 2013
Last modified October 21, 2013

As some of you may have seen, our partner GoodData just had a major release, which included the addition of a new aggregation function... Medians. This is a huge improvement for Advanced Analytics and Johann Carl Friedrich Gauss.

Until now, metrics to measure central tendency of a distribution have relied on averages or the AVG() function in GoodData. The average (or mean) is the result obtained by adding several values together, then dividing this total by the number of values.  Averages are the most common measure of central tendency, but are less robust if a distribution of values is non-normal or contains outliers.  In a case where the data has outliers, the average can be highly skewed by a small number of the total values.

Median on the other hand is also a metric to measure central tendency, but is less sensitive to non-normal distributions and outliers.  A median is the result obtained by finding the value that is at the mid-point of a distribution of values. The latest tip of the week give you a step by step breakdown of how to build a median first reply time metric in Advanced Analytics.

Head to the forums to learn more