Diversity in Data – Why it Matters

Data is everywhere.  It influences everything from our shopping habits and how we interact on social media, to global marketing strategies and political campaigns. Throughout the pandemic, governments and health organisations across the globe have been using data to track progress and plan for recovery. Even prior to the pandemic, global data consumption was increasing exponentially. In late 2018, the ID predicted that the global datasphere would grow from 33 zetabytes (ZB) in 2018 to 175ZB in 2025.

Given that data influences so many aspects of our lives, ideally the data science and technology community should be reflective of the wider population.  However, a 2020 study by the Boston Consulting Group revealed that in the US, as few as 15% of data science related roles are occupied by women.  Statistics are even lower for women of colour.

The study highlights this lack of diversity as a serious issue.   With AI algorithms being susceptible to bias, it requires a team with a diverse range of experiences and backgrounds to create algorithms that work for all.

In her book, Invisible Women by Caroline Criado Perez, the author exposes how the ‘gender data gap’ has adversely affected women across many areas of life such as medical research, technology, workplaces, urban planning and the media, further highlighting the need for diversity across teams responsible for gathering and analysing data that is used to shape our everyday lives.

The Women in Data Science and AI research project at The Alan Turing Institute is looking to address both the issues of diversity in the tech world and the and bias in the design of AI systems.  The aims of the project are to increase the number of women in data science and AI, and therefore mitigate the risks caused by non-diverse teams.

Researchers have developed a ‘Diversity Dashboard’, which can be used by tech firms to monitor the different ways men and women are treated in the workplace, based on activity on commonly used tools, GitHub and Slack.  The aim is to allow firms to have insight into these differences and respond accordingly, as well as enabling the Alan Turing Institute to use the data to make a number of data-based policy recommendations.

There is much to be done at ‘grass roots’ levels also.  At Hamilton Forth we work with organisations such as DressCode, an organisation working to address the Computing Science gender gap in schools,  and are signed up to the Digital Technology Education Charter.  We are also fully committed to using recruitment methods that encourage diversity, and work closely with our clients to ensure their brand and recruitment methods reflect that to ensure the best outcomes for client and candidates.

 If you are interested in finding out more what Hamilton Forth is doing in data contact [email protected].