Human resources (HR) is often criticized as a “soft” field where people talk more about ideas than data. It is contrasted to “hard” fields like finance, manufacturing or sales that make extensive use of numbers to guide decisions and track performance. In defense of HR, it is harder to measure abstract concepts like “human potential” or “employee value” than tangible things like profit, production rates, or sales volumes. Nevertheless, HR should do more to increase the use of data to guide workforce strategies. Sadly, many HR companies are going about this the wrong way. This is particularly true when it comes to numbers used in sales and marketing programs focused on HR solutions and services.
People like numbers because they carry an air of memorability and authority. It sounds better to say, “employees who completed this course are 146% more productive” than to say, “employees who completed this course perform better”. Numbers make arguments sound fact-based. Numbers also make it easier to tie HR programs to monetary outcomes. This is why HR vendors and consultants often put numbers in marketing materials and sales presentations. The problem is many of these numbers are highly questionable. Every time these questionable numbers are presented to business leaders to justify the value of HR methods, it erodes the credibility of the HR profession as a whole. HR might be justifiably criticized for lacking data. But this is better than being known for sharing misleading data.
Where does bad HR data come from?
The proliferation of bad data in HR comes from two general sources. First, companies may manipulate data to make it say what they want it to say even if it means sacrificing accuracy. This includes skewing statistics, sharing survey findings based on non-representative samples or misleading questions, or presenting research results in an inappropriate context. A lot of “empirical research studies” funded by HR vendors are far from objective. There is some truth to the joke comparing a research statistician and a marketing statistician. One collects appropriate data to determine the right answers. The other determines the appropriate answers and then collects the right data.
The second major source of bad data is what I call YOTA statistics. I first heard this reference when I asked a colleague for the source of a number he shared. He said it was from the YOTA research program, the most influential series of studies in the HR profession. YOTA stands for “yanked out of thin air”.[1] It is depressing how often people share numbers with no clear understanding of their source. Sometimes the numbers may be accompanied with a citation, but the citation is too vague to be a valid reference. Over time these numbers start to be treated as a truth even though no one knows where they came from or whether they are defensible.
What we can do about bad HR data?
I do not believe many people in HR intentionally share misleading data. But people in HR do engage in data practices that might be called “sloppy”. The following two things will help stem this tide of bad data that is flooding our field and threatening our credibility.
1) Never share statistics without providing a link to their source. Statistics shared in presentations without citations to their source should be considered as valid as statements made by politicians during an election speech. Only put numbers in presentations or articles if you know where the numbers came from. Ideally, learn how they were collected and calculated. And always provide adequate reference information so the audience can look up sources on their own.
2) Ask people to explain the source of numbers they share. When someone puts a number in front of you, they are trying to influence how you think about the world. Do not simply believe them, particularly if the number strikes you as surprisingly large or small. Use the number to trigger dialogue around whether it accurately describes the world. If a presenter feels something is important enough to warrant a statistic, then it should be something they think is worth discussing. Credible presenters welcome questions about sources of their data because it demonstrates an interest in their opinions.
Many things of interest to HR professionals are hard to measure (e.g., employee experience, workforce productivity). There is nothing wrong with sharing less than perfect HR data if one candidly notes its limitations. It is wrong to present bad data as though it were good data. HR already has a reputation for being soft when it comes to numbers. The field does need to further damage its reputation by being misleading and sloppy about data.
The post \"YOTA statistics: junk data is hurting the HR profession\" was first published by Steve Hunt here https://www.linkedin.com/pulse/yota-statistics-junk-data-hurting-hr-profession-steve-hunt/
About Steve Hunt
My career is focused on helping companies design and deploy technology-enabled processes to improve workforce adaptability, experience, and productivity. I believe that meaningful and fulfilling work is critical to human happiness. Creating high-quality work environments positively influences the lives of employees, their families, their managers, their customers, and their broader society. Better work environments create better world environments.
In the 20+ years, Ive been doing this sort of work, Ive had the fortune to play a role in the implementation of systems that have improved productivity and engagement of millions of employees working for hundreds of companies ranging from small start-up organizations to the largest employers in the world. My experiences have also led to authoring hundreds of articles and presentations on topics related to strategic HR and workforce productivity as well as two books on HR process design and implementation: “Commonsense talent management: using strategic human resources to increase company performance” (Wiley Press, 2014) and “Hiring success: the art and science of staffing assessment and employee selection” (Wiley Press, 2007). The link below will take you to a site where you can download most of my papers, blogs, and publications.