As Talent Development and L&D professionals, when you show the Correlation between metrics from your Talent Development initiatives and the Business Performance Measures, many of you might have got this question...."How do you know that the (Talent Development) Initiative or the Training Program was the key contributor to the Business or Individual Performance Improvement".
They would also argue that Correlation is not Causation. So, we try different Isolation Strategies (i.e., to isolate the impact of Learning Intervention from other potential factors that might have also influenced) like "Control Groups", "Pre & Post Intervention Measurements", etc. While Isolation strategies definitely helps in proving the point....but there are many practical challenges in actually doing these isolation studies in many organizations because of the very agile nature of business or mobile (moving) workforce, virtual teams, contingent staff, etc.
So, having said that, should we step back and think whether we need to go to this length to prove that our correlation is right? If there a strong correlation between A & B for whatever time period that we study, should we still spend time, effort and resources to then again validate B is caused by A?
So, with this context, now I would like you all read this post titled "When to Act on a Correlation, and When Not To" in Harvard Business Review HBR Blogs by David Ritter.
I could not agree more with David Ritter. I am also curious to know what you readers think?
They would also argue that Correlation is not Causation. So, we try different Isolation Strategies (i.e., to isolate the impact of Learning Intervention from other potential factors that might have also influenced) like "Control Groups", "Pre & Post Intervention Measurements", etc. While Isolation strategies definitely helps in proving the point....but there are many practical challenges in actually doing these isolation studies in many organizations because of the very agile nature of business or mobile (moving) workforce, virtual teams, contingent staff, etc.
So, having said that, should we step back and think whether we need to go to this length to prove that our correlation is right? If there a strong correlation between A & B for whatever time period that we study, should we still spend time, effort and resources to then again validate B is caused by A?
So, with this context, now I would like you all read this post titled "When to Act on a Correlation, and When Not To" in Harvard Business Review HBR Blogs by David Ritter.
I could not agree more with David Ritter. I am also curious to know what you readers think?
No comments:
Post a Comment