Titled ‘Data: A New Hope’, Data Summit 2022 had a decidedly aspirational tone – informed in no small way by the formidable presence of space-related keynotes and panels that encouraged big-picture thinking.

That’s not to say that more sensitive data subjects weren’t tackled. A more grounded major theme across the event was addressing data bias, diversity and inclusion across the sector. 

Brian Hills, the CEO of the Data Lab, noted two main areas of ‘hope’ for the Scottish tech sector. 

More transformational projects need to be the goal, Hills remarked, as well as focusing on inclusion and diversity. 

How Can Data Help Society’s Most Vulnerable?

Tying in values with data was the main conclusion of a panel focusing on using data to help vulnerable people. 

With a mix of public and social sector data scientists, the panel on Using Data to Improve Outcomes for Vulnerable People touched on the importance of an open-minded approach. 

Starting conversations with values first to drive the appropriate use of data to inform decision making was a general theme. 

Giselle Cory of DataKind, a charity helping other charities use data to make informed decisions, noted how data can help manage decision through even the most tumultuous of socio-polictical climates.

Reflecting on the pandemic, Giselle found that charities that weren’t using data “struggled” as the world dramatically changed for them and their target demographics. 

Sharing data responsibly was also a theme of the talk, and of the day in general. 

Forming a culture of trust was a hope of Alex Fassio, who worked for the Ministry of Justice, using data to inform policy and track change. 

The panel ruminated over how powerful data resources in the private sector – represented by organisations who are very willing to aid the third-sector – are not being used properly as there isn’t a good framework in place where private sectors firms can “share insights without risking data sharing or breaching data protection,” said Fassio.

How Key is Data to the Scottish Government’s Vision?

Tom Arthur MSP, Minister for Public Finance, Planning and Community Wealth, the special guest and Scottish Government representative for the day, affirmed that the government most do more to gain the trust of people to use data, and must do this ethically.

Unable to attend in person this year, First Minister Nicola Sturgeon still weighed in on the topic of data’s place within Scotland. 

She said: “Data is central to the Scottish Government’s vision for our economic future. It is an area where Scotland already excels, where we are ambitious to do better, and which has potential for significant economic opportunities and important social benefits.

“Data is absolutely crucial to tackling social and economic challenges, to creating jobs and wealth, and to improve people’s health and wellbeing. The talent, innovation and energy on display at an event like Data Summit demonstrates that the data sector here in Scotland is in good health.”

Does Data have an Unearned Aura of Objectivity?

The final talk of the day, keynote speaker Helen Fry, Professor of Mathematics, took a deep dive into data bias, a major underlying theme of the entire day. 

Fry challenged the audience, using different case studies to question the formation of data and algorithms when both are often built from a bias foundation at their inception.

If you Google ‘maths professor’, for instance, you’ll find that only two pictures out of 20 on image results are women. 

94% of maths professors in the UK are men, so therefore, it follows that this is a somewhat accurate representation of that particular vocational landscape, right? 

Not really, according to Fry. As she puts it, “Sometimes we don’t want technology to be a mirror of the world we live in, it can be more aspirational.”

Most bias, Fry said, occurred as a result of “thoughtless omission” but resulted in a message of “You do not belong here,” especially to those who were not involved in making the technology or databases. 

Fry provided an array of examples, from Nikon’s camera AI detecting smiling eyes on an Easter Asian person as a ‘blink,’ automatic soap dispensers not recognising dark skin, and letting agencies having no genderless selection options. 

Importantly, Fry noted, “bias is not just a technical problem.”

Nothing more succinctly expressed this than data she shared about the US healthcare system selecting high risk patients for care. 

The algorithm’s creators attempted to remove race as a proxy in an attempt of fairness. The proxy they did use to identify the ‘sickest’ patients was the amount of money already spent on treatment. 

By removing the proxy on race, they also eliminated considerations of racism as an institution. When using a proxy of number of chronic illnesses, it was found that Black people had to be even sicker than white people in order to access the same amount of healthcare when cost expenditure was used as a proxy.

Bias in data cannot be fixed with statistical tricks – bias exists outside of data and will be exacerbated by data if not dealt with properly. 

Recommended

• St Andrews biotech takes top prize at 2022 Converge Awards
• TikTok staff in China can access UK user data, firm reveals
• Energy analytics giant Wood Mackenzie acquired for $3.1bn

Tying in earlier talking points, Fry lamented that even in education, bias was steering certain demographics away from careers in STEM. 

When maths gets hard, as it does, Fry said, women are less likely to stick with it because they were not encouraged based on their intelligence, where as men of the same ability were already self-assured based on conversations at an early age focusing on their aptitude. 

Bias in data is the “nearest shadow of deeper problems” the world needs to address, Fry said. 

This is by increasing the diversity of people making databases, algorithms, and technology, while taking bias into account rather than ignoring it in vain attempts to eliminate it.