Additional Common Topics (Listed Alphabetically)

Percentiles are a function of several Assessor Series applications.

Definition of Percentiles

The general definition of a percentile is a point on a rank-ordered scale, found by dividing a group of observations into parts in order of magnitude from lowest to highest. The first percentile approximates the very lowest/bottom number found, while the 100th percentile is the very highest reported. The nth percentile is the point exceeding n percent of the observations. For example, in test taking, a score equal to or greater than 55 percent of those earned on an exam is said to be at the 55th percentile.

Percentile wage amounts shown in the Assessor Series applications are a measurement of how much variation there is in the wages paid to employees in comparison to the mean or median wage for the job. The more variation in the wages paid to employees, the larger the dollar difference will be between the mean/median and the 25th, 75th, 10th, & 90th percentile values. The more uniform the wages are for employees in a job, the smaller the dollar difference between the percentile measurements and the mean/median.

"Quartile" is a statistical term that describes three points, or intervals, that divide a group of ordered data points, or observations, into four parts. Each quartile includes 25% of the total number of data points. Depending on how the data points are distributed, the quartile intervals could be close together or very far apart.

The salaries displayed for the percentiles are statistically determined from the wage data of salaries paid for the job in question. Wage data is non-parametric in nature, and therefore does not follow a normal distribution, a bell shaped curve, above and below the mean. The percentiles shown are as close as can be statistically determined given the nature of non-parametric data.

Range Minimum and Maximum

Range minimums and range maximums refer to actual salaries paid, rather than to salary structures, and are initially set to approximately the 10th and 90th percentiles respectively of salaries or bonuses paid.

ERI's default percentiles are the 10th and 90th percentile of the salary paid. ERI does not report data for the 1st percentile and the 99th percentile for the following reasons:

1.	Most surveys do not report data on the 1st percentile and 99th percentile, so ERI has little or no data upon which to base a study for these percentiles.

The 1st percentile and 99th percentile do not offer a good measure as to the distribution of pay to employees in relationship to the measures of central tendency (mean/median). The 1st percentile and 99th percentile are more easily distorted by a few employees in a wage survey and as such do not offer a good measure as to the typical or normal amount of variation in pay for employees in a specific job.