December 4, 2017 | Written by: Michael Stanka
Categorized: HR Analytics
Share this post:
Deciding to go on a journey
Recently, one of our long-time IBM clients asked us to investigate injury prevention and safety in their warehouses. By combining results of the annual employee survey with safety data, we found strong evidence that in warehouses where employees cooperate and share knowledge with each other, fewer injuries occur. While this was a great insight, we got another (maybe even more important one) for free: When you’re just getting started with People Analytics, sometimes the journey is the reward.
Determining the destination – A chance to reflect on your priorities
Employee safety and health is essential, and there is no question that injuries, especially those that require recovery time at home, negatively impact the business. That is one reason why companies, like our client, focus on reducing these so-called lost-time injuries (LTIs). LTIs also matter to People Analytics professionals, especially when presenting results back to the business. Often the ‘so what?’ in your story will need to be answered with cost savings, and quantifying the costs of lost working days is something your HR department has been doing for decades.
Planning the route – An opportunity to leave known paths
Workplace safety is a complex issue, especially when it comes to employees’ perceptions. There are a number of factors that act as antecedents for accidents and injuries, e.g. safety climate, leadership, personal characteristics and job attitudes.(1) With regards to employee engagement, we know that the more engaged your workforce is, the fewer injuries occur. We also know that a corporate focus on safety, and a company that truly cares about its employees, the community, and the environment, helps to builds up a positive reputation, which in turn can have a positive impact on employee engagement. So while LTIs are an important outcome measure for safety, there’s more to consider in the bigger picture of workplace safety, than just the bare numbers of injuries.
Starting the journey – Fully immerse yourself in the adventure
In our client’s case, the company not only tracked LTIs, but also Total Injuries (TIs). TIs include LTIs and minor injuries. TIs are reported as frequency rates by warehouse and that means two things can be accomplished. First, the occurrence of injuries can be compared across the hundreds of warehouses. And second, you create a tongue twister that has no rival in the HR controller’s handbook: Lost Time Injury Frequency Rate.
However, our analysis was missing a key metric – the frequency of minor injuries (separated from the TIs). We believed that separating out the different types of injuries could offer some further insights, not least because research suggests that minor accidents provide a more objective measure of behavioral safety, than lost time injuries.(2) As you might expect, minor injuries occur more frequently, so while a single relatively small incident may not have a huge impact on productivity, the effects of many small incidents can add up over time.
Navigate through unknown territory – learn how to use your new equipment
We moved forward creating the new variable mIFR (minor Injury Frequency Rate), which was as simple as subtracting Lost Time Injury Rates from Total Injury Rates. With IBM Watson Talent Insights, our People Analytics solution of choice throughout the project, creating a new variable was only a matter of seconds. In the course of the analysis, Watson Talent Insights helped us uncover relationships that were hidden deeply in noisy data, outliers and country differences.
Applying the full range of Watson Talent Insights’ analytics capabilities, we were ultimately able to show that in warehouses with great teamwork, minor injury rates were significantly lower than in those warehouses with low teamwork scores. At the same time, our analysis revealed no relationship at all between teamwork and LTIs. So, the message was this: Take action to increase teamwork, and there should be fewer minor injuries. Case closed? Well, not quite yet.
Arriving at your destination – orientation in new territory
Before starting the project, we reviewed prior research and found that communication and teamwork can act as enablers for a safety climate. It makes sense that when coworkers share information and cooperate well with each other, safety improves. As a result, the client’s safety culture would certainly benefit from actions that emphasize safety and teamwork, e.g. conducting training that embraces teamwork as an element for preventing injuries. But did we really have enough evidence for taking action?
Some might consider discovering causal relationships as the holy grail of People Analytics. But what if we cannot be confident in causal effects, e.g. if data sources are limited? Or, if investigating causality requires an effort that a company is not willing to take, like developing sophisticated experimental research designs? After all, in real world settings, definitive evidence of causality is elusive. In our case, we already had a solid basis of data, we formulated hypotheses based on prior research, and we were confident in our findings. Still, we’ve found ourselves confronted with some kind of a chicken-egg dilemma: There was a chance that in teams with great teamwork scores, minor incidents are just not reported. Perhaps, for example, injured people just continued to work, sacrificing their well-being for the team. Without further information, we couldn’t be sure about what causes the other.
So, what can be done without the holy grail of causality? We are confident (from our analysis) that there IS a relationship between minor injuries and teamwork. Discussing this insight with managers and employees can illicit valuable qualitative insights to add to the data-driven analysis. You could ask questions such as: “Do employees in your team speak about safety topics?”, “Would you consider teamwork to be an important factor for safety compliance?”, or “Do you believe your team reports minor injuries in a consistent manner?” The results of such discussions, together with the robust analysis, can add to your confidence that actions you take will directly impact safety at your organization.
Looking for your next destination
Working closely with stakeholders is an imperative for most People Analytics initiatives, but in large organizations interviewing all managers is not a practical approach. That’s why an alternative to a qualitative follow-up might as well be another survey. When accidents occur frequently at a given time and place, timely interventions and actions are essential. While conducting annual employee surveys remains important, a safety pulse survey provides the opportunity to act quickly when an area of concern is identified.
Find out how continuous listening can help your organization conduct pulse surveys and how cognitive analytics solutions can help with interpretation of data in our Amplifying Employee Voice homepage.
(1) Christian, Michael S.,Bradley, Jill C.,Wallace, J. Craig,Burke, Michael J. (2009). Workplace safety: A meta-analysis of the roles of person and situation factors. Journal of Applied Psychology, Vol 94(5), Sep 2009, 1103-1127.
(2) Zohar, Dov. (2000). A Group-Level Model of Safety Climate: Testing the Effect of Group Climate on Microaccidents in Manufacturing Jobs. Journal of Applied Psychology, Vol 85(4), Aug 2000, 587-596.