By Berta Karaim (AI Analyst at Monash University & Volunteer @ GDI).
The data and analytics sector (D&A) is no poster child for diversity. For instance, only 17 percent of all roles are filled by women. Meanwhile, people of colour hold just 26% of leadership roles, in spite of making up 52% of the industry. Nonetheless, D&A can be utilised to further diversity and inclusion in organisations across all sectors, from non-profits to governments to tech giants.
Diversity and inclusion are not wish list items but business essentials. Deloitte's review reported that organisations with inclusive cultures are several times more likely to meet or exceed financial targets. So what is diversity and inclusion, and how can D&A be used to further it?
One simple way of defining diversity and inclusion (D&I) would be to say that diversity is:
while inclusion is about:
Now that we’ve defined D&I, let’s dive into how data and analytics can be used to further it – specifically, we’ll look at how to disclose, reflect on, discuss and exclude data.
Schematic: 4 stages to implement D&A for improved D&I
1) Disclose
The role of disclosure in D&I (publicly sharing workforce composition and targets) is curious, mainly because the power of disclosure rests with the one disclosing, not the one being disclosed to. More specifically, disclosure appears to drive change in the organisation disclosing the data, regardless of the response of the recipients. And yet, disclosure alone is not enough to further D&I. So how does one disclose D&I data effectively?
In their Harvard Business Review article How to Best Use Data to Meet Your DE&I Goals, Siri Chilazi and Iris Bohnet suggest sharing data to encourage action and focus on what has already been achieved:
Presenting data in a way that is “simple, salient, and comparable”
Empowering people to act using data. For example, highlighting to employees the connections between their daily actions and D&I outcomes.
“[Setting] diversity goals to create accountability and increase follow-through”
“[Leveraging] diversity data to shift social norms around DEI” (such as focusing on the overwhelming majority of companies who have at least one woman on their board instead of continuously emphasising the lack of women on boards)
2) Reflect
Image source: https://www.multitudes.co/blog/what-we-measure-and-why
Metrics signify what is important to a company, so being aware of bias is crucial when selecting what data to collect and how to analyse it.
Multitudes is an example of a company that is conscious of bias when gathering and analysing data. When explaining the reasoning behind two of their metrics, “Review Wait Time” and “PR [pull request] Feedback Received”, Multitudes refers to research on gender bias in the workplace. For example, research shows that people in marginalised groups get less feedback than others. That's why Multitudes looks at Feedback Received by each individual to make sure everyone is getting this opportunity to learn and grow.
3) Discuss
Image source: https://www.multitudes.co/blog/what-we-measure-and-why
One of the key powers of data is as a conversation starter, especially when put back in the hands of the people it is collected about. To continue the example from above about Multitudes and gender bias, let’s look at unconscious bias and feedback.
It has been observed that unconscious bias can result in women receiving less specific and less actionable feedback. Consequently, Multitudes measure the PR Feedback Received to serve as a starting point for discussions in teams about the feedback and support received by individuals within teams.
4) Exclude
Image Source: Author
Although it may feel counterintuitive in a discussion about inclusion, exclusion of data can be at times the key to a more inclusive, less biased workplace! This is especially the case, when demographic data (such as one’s address) can serve as a proxy for a protected characteristic (such as race). As Cathy O’Neill describes in her book Weapons of Math Destruction, a company examining its retention decides to exclude from its model of how long an employee will stay on with them once they realised that although a longer commute time could be correlated with a shorter retention, it was clearly associated with living in a poorer area. Therefore, excluding proxies for bias leads to not only a more accurate analysis, but also ensures that more people have the opportunity to break out of destructive cycles of poverty and the like.
If the research cited in this article is any indication, then the road to diversity and inclusion is a necessary one and one where D&A play a key role. So let’s disclose, reflect, discuss and even exclude as a way towards a world where D&I are the new normal!
Get Involved:
If you are a data & analytics professional looking to make an impact and become part of a global community of people like you, applications for our next (4th) GDI Data For Good cohort will open in early 2023 with the program running over the first half of 2023. This program serves as the gateway into the Good Data Institute community and consists of learning and development (L&D) events and a team-based hackathon with a NFP partner. Sign up for our newsletter here to stay in the loop.
About GDI:
The Good Data Institute (established 2019) is a registered not-for-profit organisation (ABN: 6664087941) that aims to give not-for-profits access to data analytics (D&A) support & tools. Our mission is to be the bridge between the not-for-profit world and the world of data analytics practitioners wishing to do social good. Using D&A, we identify, share, and help implement the most effective means for growing NFP people, organisations, and their impact.
