This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
Has anyone used Market Basket analysis within the HR Analytics arena?
To be specific, I want to see what "basket" of goodies go with employees who quit, and which things go with an employee who stays? Which bonuses, what type of raises? Do they get mentoring, promotions, or lateral moves within the company?
A bit of background, I'm an economist by training. Everything fits in a regression equation, or I don't work on it. But in this case, the things that are colinear, if certain bonuses appear together, those really matter. (Colinear is a dirty work for regression equations.) I'm trying to say, Market Basket algorithms are not something I normally use.
I guess I'm asking, does this sound like a situation that a Market Basket would apply? If anyone has done anything like this, what things should I be aware of?
Interesting question! Typically market basket analysis looks to find relationships among items that co-occur in transactions. I suppose you could take a dataset of information about employees who have left the company and, for each employee, list the “items” that pertain to that person (mentorship program participation, receipt of bonus, promotion, etc.). Applying market basket analysis would reveal connections between the items in the “basket” of each employee (e.g., mentorship co-occurred with promotion). You could do the same for data on employees who have stayed, then see if there are interesting/comparable patterns in the “baskets” of coinciding items most frequently occurring among those who stay or quit.
I’ve written a couple of blog posts about doing market basket analysis in Designer: Market Basket 101 and 102.
Hope that helps a bit as you think about how you might use this. I’d also love to hear if folks have tried this kind of approach and if it brought useful insights.