020 7650 1200

Computer Class

De-coding the gender conundrum

Where are the women in big data?

Posted on 08 March 2018


We are all used to turning to the internet to ask a burning question that we can’t answer. But have you ever wondered when you tap your enquiry into a search engine who decides what results are returned?
Most of us have heard of the algorithms that power search engines but increasingly questions are being asked about how neutral these algorithms are, and whether the lack of female IT engineers means that our search terms are being skewed before we even click through to the websites listed on the all- important search engine results page.
The internet was largely built by eager young white men and you could argue that many of the results you get from search engines tends to reflect a similar world view Even today, if you search for the term engineer, company director or lawyer in Google Images you get a screen full of pale males to choose from.
More disturbingly, Safiya Umoja Noble, Assistant Professor at the University of Southern California Annenberg School of Communications, found that when she ran a search on ‘black girls’, the search results page was full of links to porn sites and un-moderated forums, while a search for ‘white girls’ resulted in a very different set of results. This led her to write Algorithms of Oppression, in which she challenges the orthodoxy that search engines such as Google offer an equal playing field, and argues that data discrimination is a real issue.
The auto-complete function in search engines can also throw up some irritating suggestions, if you type ‘why women’ into google one prediction that is offered is the statement ‘why women can’t read maps’. Google is at least offering a way to report inappropriate predictions, although its success is yet to be measured.
How many women work in IT?
Can you remember the last time you phoned your IT support team to be answered by a woman? No, me neither. Figures from the Women’s Engineering Society  suggest that the picture isn’t going to change any time soon either.

  • The proportion of young women studying engineering and physics has remained virtually stactic since 2012.
  • The number of women in computing degree programmes is falling: 14% in 2010 and 13% in 2014.
  • In 2017 there were nearly 11,000 fewer women working as ICT professionals than in 2016
  • Only 17% of employees in the UK technology sector are women 

Encouragingly at the Women in Data conference 2017 more than 450 data female data professionals met to promote female talent, and to discuss the challenges they face. But while the number of female computer engineers is growing we need more.
In a blog last year, I wondered what we can do to encourage young women to study the all-important STEM (science, technology, engineering and mathematics) subjects which can lead to careers in science and technology. The work of organisations such as WISE and STEMettes can only help with reaching this goal.
Creating an encouraging learning environment for women studying computer science is one obvious step that might lead to more female IT professionals, but should the recruitment industry be doing more to avoid gender stereotyping when recruiting for these roles?

Research has been carried out in the US about the subconscious bias that can be found in the wording used in online job ads. By avoiding male-coded vocabulary when advertising for IT roles the number of women working in the IT sector could be increased. It must be a lonely place to be the only woman in an IT team and businesses must be alive to the challenges that this presents, and support their female IT staff as much as possible.

The theme of this year’s International Women’s Day is #pressforprogress, and one of the calls to action is to challenge stereotypes and bias. Why don’t you try doing your own google image research and report the findings using the #pressforprogress?

Data defines us, and we need a more diverse workforce in the IT sector to ensure the human biases that we all have don’t creep into the algorithms that power many of the services that we use, and choices that we make. The more women who are working in IT the more likely it is that when you call your IT team the call will be answered by Janet rather than John, and that when you search for a Stephanie Jones  the result won’t auto correct to Stephen Jones. And the quicker that day comes the better.



Helen Dewar
Library and information services

Helen Dewar