exponential regression excel data analysis

#Exponential regression excel data analysis how to#

What other regression analysis is out there? When creating salary benchmarks, you’re tasked with creating a set that works for the roles in your organisation. Perhaps you want to pay a lower salary at level 3 than what the regression analysis says, and that is totally fine. This type of analysis is there to support you, and you don’t have to fully adopt all of the output.You don’t have to align to the exponential data For many job families you may decide that the raw data behaves in the way it should there are no dips or holes and therefore it requires no regression.I always recommend checking the data with the hiring managers and any inhouse recruiters who will know the role and market best.This takes the raw data and models the relationship between the salary (Y) and number of levels (X), to fill in gaps and smooth out the dips.Ī few things to note about using this logic

This illustrates data dipping at level 5 in the Customer Success Team and in the Marketing Communications team, there is no data all for level 5. ( Click here to learn more about this spreadsheet).īelow are a few examples of how this works in practice on some data pulled from an external provider. I’m data agnostic, so it works with any external benchmarking providers’ data. I’ve built a handy spreadsheet to help model large data sets of salary benchmarks using exponential regression. The higher the R-squared, the better the model fits your data.

Why? Exponential regression consistently returns a higher R² number than linear regression. In exponential regression, the function is a power (polynomial) equation of the form or an exponential function in the form.Īfter a few years of using linear regression, I now use exponential regression and use the GROWTH function in google sheets/ Excel. In linear regression, the function is a linear (straight-line) equation. Typically, comp analysts will either use linear regression or exponential regression. So what regression analysis options are available? In our case, we want to understand how a level (dependent variable) influences salary (independent variable).

While there are many types of regression analysis, at their core they all examine the influence of one or more independent variables on a dependent variable. Simply put, regression analysis is a powerful statistical method that allows you to examine the relationship between two or more variables. It’s a very common tool that compensation analysts use to ‘clean’ salary data when it misbehaves. Rest assured you don’t have to be a data scientist or learn Python to understand it and crunch the numbers (I’m certainly not a data scientist and I don’t know Python).

Now, regression analysis may sound quite daunting, and it did make me shudder the first time I read about it. Solving dodgy salary data with regression analysis

#Exponential regression excel data analysis how to#

These remain among the most common problems people approach me with (after they’ve chosen a data source), so I wanted to shed some light on how to solve this.īefore reading any further, if you’re unfamiliar with levels then read this great post on this subject from Molly Graham (ex Facebook, Google, Quip). data going down instead of up as seniority increases I wrote an article in November 2020 about running a salary benchmarking process for the first time in which I highlighted two common challenges with salary data from external providers: