# What size group is required to do driver analysis?

To measure engagement drivers using multiple regression analysis, there need to be approximately 30 individuals for every dimension measured. In a typical TalentMap survey, we measure 12 dimensions of engagement.

Accurately measuring all of the drivers requires 360 respondents. That said, for smaller sample sizes, a driver analysis could be done using a linear regression model instead of a multiple regression models, but it would not be an apples to apples analysis, and it would be inconsistent with how the drivers are measured for the company as a whole.

TalentMap uses linear regression analysis for smaller organizations and we find that a minimum sample size of approximately 30 people is needed to effectively do a driver analysis. However, multiple regression analysis is the preferred method when possible as it takes into consideration all the variables in the relationship and not just one variable at a time, painting a more accurate picture.

Accurately measuring all of the drivers requires 360 respondents. That said, for smaller sample sizes, a driver analysis could be done using a linear regression model instead of a multiple regression models, but it would not be an apples to apples analysis, and it would be inconsistent with how the drivers are measured for the company as a whole.

TalentMap uses linear regression analysis for smaller organizations and we find that a minimum sample size of approximately 30 people is needed to effectively do a driver analysis. However, multiple regression analysis is the preferred method when possible as it takes into consideration all the variables in the relationship and not just one variable at a time, painting a more accurate picture.

Updated on: 02/18/2019

Thank you!