When you run an Executive Regression go to "My Reports" where you can download your Executive Regression report. The report is an Excel spreadsheet that is very similar in look and feel to a Market Query report.
At the top of the report in columns R and S, users have the option to change the Revenue scope used for the report.
Below “Revenue ($)” is “Custom Percentile”. Users can adjust the "Custom Percentile" to reflect what market percentile a user employs for their existing pay strategy. The spreadsheet is interactive and will update.
Salary estimates will be given in columns W and X. Column W is the “Median” regressed value and will be constant- the 50th percentile. Column X is “Custom” and that will reflect what was entered into “Custom Percentile” (cell S5).
The other columns in this report are defined here:
Regression Scope: This refers to the Scope chosen when the report was generated, either “Revenue” or “Headcount”.
- Regression Scope Low Limit: For peer data used, this indicates the lowest Revenue value used for the analysis for this specific job code.
- Regression Scope High Limit: For peer data used, this indicates the highest Revenue value used for the analysis for this specific job code
The low limit and high limit indicate the minimum and maximum values for that role and are used to ensure that the salary estimate generated falls within that range. In general, you do not want to use a value greater or lesser than the minimum and maximum values for each role.
Warnings: This lets the user know what particular warning to look out for when they view the regression. There are two warning types:
- “The R Square value is below 0.2 for one or more pay components”
- For the regression model, the R Square value is letting users know what percentage of pay variance is due to the variance of revenue. The lower the R square value, the less the pay variance can be attributed to the variance in a company’s revenue/headcount. So the smaller the r squared value, the less relevant a company’s revenue is to the indicated pay. User Note: When R Square is .4 or higher, that is considered to be statistically valid.
- “ The slope is negative”
- The premise for the regression model is that as a company’s revenue size increases, pay for executives increases accordingly. That premise is represented by the slope of the regression line being positive. When the slope of the line is negative, it is saying that as the size of a company increases, the pay for that role is decreasing. When this occurs, it is recommended that users do not use the generated salary data, or use it with extreme caution.
Employee Count: This indicates the number of incumbents used for the regression
Company Count: This indicates the number of overall companies used for the regression
R Squared: This indicates what proportion of variance in the dependent variable (i.e., pay) can be explained by the independent variable (i.e., revenue or headcount). In other words, how much of a given pay level can be explained by a company’s overall revenue/headcount. For our purposes, the higher the percentage the better- anything over 40% is considered good.
Slope: This indicates how steep a line happens to be. In the case of regression, it indicates how quickly compensation increases as a company’s revenue/headcount increases. In a linear regression equation (y = ax + b) a is the slope where x is the independent variable (e.g., revenue/headcount) and y is the dependent variable (i.e., pay).
Intercept: The value of a regression equation when the independent variable is zero. In a linear regression equation (y = ax + b) b is the intercept where x is the independent variable (e.g., revenue/headcount) and y is the dependent variable (i.e., pay).
STD Error: A measure of the statistical accuracy of an estimate, equal to the standard deviation of the theoretical distribution of a large population of such estimates. Used to calculate other statistical elements such as a regression 75th percentile.