How to Use Statistics in HR to Drive Actionable Outcomes
Some of the common misconceptions around statistics are, 'Statistics can't tell you anything about the why, only the how much,' or 'Statistics needs normal distributions, and we don't have normal distributions with real data' and 'If something isn't statistically significant you can ignore it – it doesn't matter.'
It's time to bust these myths and misconceptions so that HR can harness the full power of statistics in their people analytics function — functions which – by the way – have grown by 43% since 2020.
Therefore, in this article, we're going to answer the top six common questions we hear about the basics of statistics in HR:
What's the difference between basic and advanced statistics?
How do I take care of ethics and privacy when doing statistical analysis?
How can data visualisation enhance communication in People Analytics leadership?
Bonus Question! Where can I learn more about statistics in HR?
How Do You Use Statistics in HR?
Statistics is analysing, making sense of and using information to make informed judgments and decisions. In an HR context, that means using relevant HR data to make less biased, more objective decisions and recommendations, as opposed to relying on gut feel or intuition alone.
As with any people analytics activity, starting with the problem your organisation is trying to solve, for example, a high attrition rate, is critical. Statistics can help HR professionals understand and validate potential causes for this issue, making data-driven recommendations to stakeholders, and positioning themselves as strategic partners for the future of work.
As Ben Teusch outlines, the key is to really understand the question you're trying to answer by spending time answering 'pre-analysis' questions. Therefore, it's important to always start with one key question you are trying to answer for your organisation. This can involve writing hypotheses and then testing them to ensure clarity in problem-solving.
What's the Difference Between Basic and Advanced Statistics?
A common misconception is that 'advanced analytics modelling' means 'better' – this is simply not the case! Basic statistics can be just as powerful in helping you answer questions and address challenges in an evidence-based way, providing valuable insights into HR data.
Basic statistics refer to descriptive statistics in HR, such as the mean, median, and mode. These are measures of central tendency, which can be considered measures of the 'middle'. We also use measures of dispersion (how spread out things are), such as the range and standard deviation (how spread out information is from the average).
Basic statistical analysis can also help understand things such as:
Correlations, which measure the relationship between two or more pieces of information (e.g. manager satisfaction and engagement).
Causation, which measures to what extent one thing leads to another (e.g. does low manager satisfaction explain low engagement? How many other factors could contribute?).
On the other hand, advanced analytics includes more complex techniques. For instance, you have data science, which analyses unstructured data such as employee feedback or communication platform data, using techniques such as natural language processing (NLP) to understand sentiments and themes in the feedback.
And if we look into the immense potential of AI in HR, we see that generative AI and people analytics techniques like regression analysis, predictive modelling and machine learning can provide more accurate insights and predictions.
Take for instance, Tomas Chamorro-Premuzic article which explains how AI is better at predicting job potential than humans. In his words,
"AI can accurately translate the words we use into a reliable estimate of our personality, values, and intelligence, as well as identifying language patterns that signal narcissism."
However, just because you have access to advanced statistical techniques doesn't necessarily mean you should use them for every HR problem. The focus should always be on addressing the specific business problem at hand, using whatever method is best suited for that particular challenge.
For example, conducting the basic statistical analysis may be more appropriate for understanding correlations between employee engagement and performance in a retail store (see case study in question 6). Whereas using predictive modelling through AI and organisational network analysis (ONA) could be an effective approach for identifying high-potential employees in a large organisation.
This is why it is ever important in upskilling HR in understanding the distinctions between data analytics, data science, and machine learning and know when each should be applied to make the most of HR data and drive actionable outcomes.
Who Can Benefit From Statistics in HR?
Simply put, anyone looking to make sense of HR data and use that information for better decision-making can benefit from statistics in HR.
HRBP's and connector roles
People analytics and statistics can be a great asset for business partners (HRBPs) – and anyone who acts as a bridge between HR and the business, such as connectors - because it allows them to move beyond simply presenting information to having strategic conversations with stakeholders. This enables them to take an evidence-based approach to HR strategy and align it more closely with overall business strategy.
The differences between people analytics and evidence-based HR?(Source: Evidence-based HR and people analytics are the same, right? Afraid not, by Rob Briner)
However, for HR to truly make evidence-based decisions, it's important in upskilling HR to consider multiple sources and types of evidence and information. As Rob Briner highlights in his People Management article on information integrations, this could include: stakeholders' views, perspectives and judgments; professional expertise of practitioners; data and evidence from the context or setting; and scientific findings.
Business leadership and people managers
Business leadership and people managers can also greatly benefit from statistics in HR. By providing a more objective and less biased understanding of their workforce, they can make better decisions to drive productivity and success.
As Lexy Martin, former research principal at Visier, shares on the Digital HR Leaders podcast,
"Providing data and insights to managers makes them more effective and more human. So, they're delivering cost savings and revenue improvement, they're preparing for the future, and they're balancing this need to have their organisations be both profitable and productive while also enabling their people and their teams to thrive."
Employees
And finally, the workforce itself can also benefit from statistics in HR. By using insights gathered through analytics modelling, organisations can improve employee experience and give a voice to employees by surfacing essential insights that were previously unseen. This can lead to more informed and influential people strategies that consider employee needs and well-being.
Interested in learning more about using statistics to analyse your HR data? Take a look at our online People Analytics certifications on myHRfuture
How Do I Take Care of Ethics and Privacy When Doing Statistical Analysis?
In our skill booster training course, 'Using Statistics in HR', instructor Heather Whiteman says,
"Just because you can measure something, doesn't mean you should."
Be conscious and considerate of how you use the data and what you are analysing. If you think it's a bit creepy, your employees will likely too. Be ready to explain what you're doing and why to every employee, and be as transparent as possible with the workforce and their data – whatever insights or results you can share.
As a People Analytics leader, you must lead by example when it comes to ethics and privacy. Advanced analytics modelling often involve handling sensitive employee data, therefore it is imperative that you ensure your team adheres to stringent ethical guidelines, respects privacy, and maintains transparency in all data-related activities.
This is why it is ever so important for organisations to have a strong ethics charter and council in place, which our research at Insight222 has found to be a significant characteristic of leading companies in people analytics. As such, we have amended and created a new model for people analytics, which you can discover her.
How Can Data Visualisation Enhance Communication in People Analytics Leadership?
Data visualisation is an essential tool to consider when communicating insights from HR data. How data is presented can significantly impact how effectively it is understood and acted upon by various stakeholders.
Designing for stakeholders
Effective data visualisation requires considering the audience who will be consuming the information. This means understanding their needs, preferences, and level of expertise with data. For example, if your audience is not data-savvy, consider using simple, easy-to-understand charts and infographics. If they are more familiar with data, you can use more complex visualisations.
Using colour sparingly
Colour in data visualisation can also play a significant role in drawing attention to specific insights or patterns. However, it's important not to go overboard with colour as it can be visually overwhelming and may distract from the main insights. It's best to use colour sparingly, strategically highlighting key data points or trends.
A clear language
In addition to design elements, consider the language used in data visualisations. Use clear and concise titles, labels, and descriptions to guide the audience through the information, and avoid using jargon or technical terms that may not be familiar to all stakeholders.
As Cole Nussbaumer Knaflic highlights on the Digital HR Leaders Podcast:
"If we do those three things, we think about our audience, we design with them in mind, we use colour sparingly to focus attention and words that tell our audience why we want them to look there and what the takeaway is. That is a successful scenario for communicating effectively with data."
And with generative AI, understanding how to apply data visualisation techniques effectively has become much easier - creating human-like visualisations that consider the audience and their needs, taking away much of the guesswork from designing impactful data representations.
What Are Some Real-life Examples of Statistics in HR?
Here we highlight two examples of statistics in HR in action.
Example 1: Correlating engagement and performance at Clarks
Clarks wanted to understand if there was a correlation between employee engagement and store performance. To understand performance in stores, they analysed 450 business performance indicators over a number of years, including store productivity, customer conversion, sales, profit and customer satisfaction.
To understand engagement in stores, they did a qualitative analysis by interviewing store managers. By performing basic statistical analysis on this information, Clarks were able to prove a significant correlation between engagement and business performance, proving that every 1% improvement in engagement was worth an extra 0.4% in terms of business performance. The recommendation attached to this insight was to create a store management development programme that focuses on improving store staff engagement.
Example 2: Bringing people analytics algorithms to life at Banco Santander
Another brilliant example that focuses on the value of bringing people analytics modelling algorithms to life is presented in Jonathan Ferrar and David Green's book 'Excellence in People Analytics: How to Use Workforce Data to Create Business Value'.
Banco Santander, one of the largest banking institutions in the world is known worldwide to have acquired many organisations across Latin America and North America. One of these acquisitions includes Santander Brazil, which has over 50,500 employees and nearly 4,000 branches.
Taking a unique approach to people analytics, Santander Brazil integrated multiple advanced statistical algorithms into one platform making it accessible to all business leaders, allowing them to make data-driven retention, compensation, and succession planning decisions.
By productising and democratising these insights Santander Brazil has driven a successful data-driven transformation across the organisation. As a result of their efforts, they have seen a significant return on investment and improved employee outcomes.
If you enjoy practical examples like these, we highly recommend reading "Excellence in People Analytics" and checking out our case study library in the myHRfuture Academy, which is packed full of real-life examples from a multitude of different organisations detailing their transformation journeys for the future of work.
Where Can I Learn More About Statistics in HR?
To learn more about how to use statistics to analyse your HR data, debunk common myths and misconceptions associated with analytics, and understand how you can leverage statistics in your people analytics projects to drive actionable outcomes and real business value, consider taking our learning pathway: "Understanding and Applying Statistics in HR".
This program is designed for upskilling HR and people analytics professionals looking to build a deep understanding of different types of statistical analysis and how they can be applied to HR data.
You'll learn about correlation, causation, statistical significance, dependent and independent variables, and how to use the Eight Steps Model of Purposeful Analytics to turn your analysis into actionable recommendations for business leaders.
Also consider taking our course on “How to Use Statistics in Your People Analytics Projects”, where you will learn how to apply statistical tests in both Excel and R, as well as understand when it's appropriate to use each tool.
Don't miss this opportunity to enhance your teams and HR colleagues’ skills in people analytics through statistics. The future of work is data-driven, and it's essential to stay ahead of the curve by upskilling in this area.