Understanding Compensation Data

The following terms are important to understanding salary surveys and other compensation tools:

Average Base Pay

Average base pay represents the simple average of reported pay rates, where each pay rate (and organization) is counted equally. It is found by first calculating the average of pay rates for each job reported by each organization, and then summing the calculated average pay rates and dividing by the number of reporting organizations for that job. Average base pay does not take into account the number of employees in the position. Given that all pay rates are counted equally, the simple average is influenced by extreme values.  

Weighted Average

Weighted average takes into account the number of employees in the position; it does not treat all pay rates equally. It is a truer measure, therefore, for determining the closest to market pay rate. Larger employee counts have more influence, or weight, than smaller employee counts. Weighted average takes into account the number of employees associated with each pay rate for a given job and the total number of employees. It is found by summing all individual pay rates provided for a job and dividing by the total number of pay rates, or employees.  

Median

Median indicates the middle marker in an ordered array of values. With an odd number of pay rates, the median is the true middle point. With an even number, the median is the average of the two middle responses. The median is unaffected by extreme values; therefore, it is a stable measure for small sample sizes and/or high variability across pay rates. Note for the purpose of this report, calculation of median takes into account the number of employees at each reported pay rate. Using this method and like the weighted average, the median is often accepted as a market rate measure.

Percentiles

The 10th, 25th, 75th, and 90th percentiles are provided with each displayed data cut. Percentiles represent positioning on the market spectrum. The 10th and 25th may represent paying below market rate, while the 75th and 90th may represent paying above market rate. A percentile denotes the numeric point at which X% of pay rates are below or equal, and X% of pay rates are above. For example, the 25th percentile is where 25% of reported pay rates are below or equal while 75% are above. Note for the purpose of this report, calculation of percentiles takes into account the number of employees at each reported pay rate. 

Percentiles rank the variability existing in the data, allowing them to be used as a set of market standards. Thus, percentiles are best used with larger sample sizes (at least 20 or more). The more pay rates provided, the more useful and reliable a range comparison.

Written by a Catapult Advisor.