An article by Natalie Cramp, first posted by HRZone, 29 April.
Creating more diverse and inclusive organisations is a key priority for many HR and business leaders today. But without using data science to support decision making, bias will creep in and hamper any well-intentioned efforts.
The past 12 months have been challenging for everybody, not least HR professionals. Not only has the complete disruption of office life made supporting, monitoring and assessing teams much more difficult, the Black Lives Matter protests last year have placed issues around diversity and inclusion near the top of the corporate agenda.
Add to this the difficult economic situation that has necessitated rapid changes in hiring and retention policies as well as new furloughing and return-to-work processes, and the average HR function’s to-do list can be eye watering.
Despite this, every period of change offers an opportunity to challenge the status quo and tackle long-standing problems. In this instance, the increased prevalence of data science combined with remote working offers a way to revolutionise the HR function within most businesses.
If applied well to a data-driven HR function, data science has the capacity to help HR professionals and line managers identify unconscious bias.
What exactly is data science?
Let’s start by drilling down into what we mean by data science. For our purposes, it’s best to think about it as the extraction of knowledge and insights from data. It is not just about looking backwards at what has happened and analysing it, it’s also about making predictions about what will happen in the future.
The crucial factor is that data can be both structured and unstructured. Structured is the information you usually think of in the context of analysis – neatly arranged rows of numbers. In comparison, unstructured data is unorganised and often includes rows of text, videos and audio files. For HR, think about the structured data of an individual’s sales numbers contrasted with the ‘unstructured’ nature of written feedback, performances in meetings or on Zoom calls.
In practice, this means that data science can take into account nearly every stream of relevant information to not only provide a more complete picture of what is going on but also cast a more scientific and empirical light on factors that you might have considered ‘unquantifiable’. For example, how an individual contributes to morale, their potential or personal ambitions.
A simple example of data science in action is a recommendation engine. On sites like Netflix or Amazon your behaviour and preferences combined with demographic information is all used to generate predictions on what shows or products you will be interested in. The more you use the recommendation engine, the more data is collected and the accuracy of the algorithms that underpin it should improve.
What does data science mean for HR?
Well, as I’ve touched upon, people professionals use a lot of structured and unstructured data to make important decisions. Each decision is intertwined and has to be set in the wider context of the company they work in. The average HR professional is actually performing a series of incredibly complex calculations. Even with the best skills, intentions and processes, mistakes are going to happen.
The most obvious way that these mistakes have manifested themselves is in the dismal levels of diversity in UK businesses. Naturally, a lot of research has taken place to identify just what is going wrong, and one of the issues that comes up time and time again is unconscious bias.
Everyone, even with the best intentions, is being unduly influenced in their decision making in a way that can disadvantage underrepresented groups. To make matters worse, such is the nature of unconscious bias that these decisions are often self perpetuating and a direct contributor to systemic discrimination.
What I’m advocating is not the removal of the human from HR, but rather the application of the latest tools to empower HR professionals.
So how do we break this cycle? This is where data science can play a powerful role. If applied well, to a data-driven HR function, data science has the capacity to help HR professionals and line managers identify unconscious bias by providing more scientific rigor and insights on which decisions can be made.
Getting started on this approach does not mean tearing down your current HR processes and starting from scratch. What is required is first collecting and storing information in a systematic way that enables it to be continually and automatically updated. Then, data scientists can work with your HR team to build algorithms that will ingest this data and generate the insights that are needed to inform decisions.
For example, improving succession planning by identifying a broader pool of potential, or identifying a broader pool of top talent in an organisation to consider for progression.
Empowering, not removing, HR
It is very important to note that what I’m advocating is not the removal of the human from HR, but rather the application of the latest tools to empower HR professionals. Indeed, it is absolutely essential that the HR function works closely with data scientists to build these systems. This is because an algorithm is only as good as the data that is inputted and parameters within which it operates. This means that HR needs to ensure that the factors assessed provide a level playing field and do not actually perpetuate discrimination.
A natural consequence of this is the need for HR practitioners to upskill and educate themselves on the basics of data analysis. This will help to create safeguards that will ensure that HR and line managers are not blindly led by the data but can verify outputs and, crucially, recognise how certain data points, or lack thereof, can inadvertently discriminate against certain groups.
Make data science a key part of your diversity strategy
Data-driven HR will not solve all the problems linked to diversity and inclusion in the workplace. However, it will provide a whole host of new tools and techniques that will help to identify discriminatory or biased practices and ultimately enable fairer decisions to be made.
Our current period of upheaval, where new ideas and systems are needed to help businesses adapt, presents an unparalleled opportunity for companies to take their HR function down this route. My advice to every HR professional is to actively start learning about data science so they can see how data can help them do their jobs now and in the future.