Median and mean are useful measures of distribution that identify central values and tendencies of a data set. In a WageWatch compensation survey, median is the middle wage in a set of ranked market wages, which separates the data set in half. When an even number of wages are ranked, with no true middle value, the average of the two middle data points is the median wage. Median is also called the 50th percentile.
The arithmetic mean which is also called the simple average or mean is one of the most commonly used statistics in business. One of the reasons is that it is very easy to calculate. It is the sum of all values of a data set divided by the number of values in that set.
In a normal distribution of data, the median and mean wages will be within a few cents of each other. When the mean is greater than the median, this indicates the data set is skewed towards higher wages. Similarly, when the mean is less than the median, this indicates the data set is skewed towards lower wages. Skewing typically occurs for one of two reasons: either the data set is not normally distributed, or the mean is affected by outliers or errors in the data set. The first situation is not uncommon and occurs when there is either wage compression or the data is bimodal, meaning that the job description may be too broad and need to be bifurcated into two job titles. The second reason for the mean and the median being significantly different, due to an outlier or data error, is much more of a concern.
Here is an example of how data affects the median and mean differently when an outlier is added to the data set.
Wages in Data Set A Wages in Data Set B
Set A Median Wage $11.27 Set B Median Wage $11.98
Set A Mean Wage $11.21 Set B Mean Wage $27.32
In this example, it is clear that once the $125.00 value is added as an outlier, the median wage becomes the better indication of the central value. Comparing the median with the mean from Set B tells us that while the middle wage or median is $11.98, there appear to be employers paying at the top of the market, skewing the mean up sharply. The only way to know for certain if the difference between the median and mean is due to an error is a more thorough analysis of the data. If more data analysis is not possible, then the median is your better statistic to use.
WageWatch, Inc. is the leading compensation survey provider for over 5,000 hotels, non-profits organizations, healthcare organizations and management companies. The PeerMark™ Wage Survey is the only Web-based custom survey tool that allows individual survey participants to select their competitive set for comparison purposes.
At WageWatch, our expert evaluators provide businesses in a large range of industries with accurate and beneficial benefits survey data, compensation surveys and salary reports to ensure that payment and benefits plans are on par with those in the industry. For more information on market compensation data, please call WageWatch at 888-330-9243 or contact us online (https://www.wagewatch.com/Contact/ContactUs.aspx).