Peak’s Women in Data ScienceBy Emma Bellamy on October 5, 2021
Historically, there has always been a lower proportion of women in data science and in the wider Science, Technology, Engineering and Maths (STEM) fields.
To delve deeper into why this might be the case, we wanted to get to know our own team a little better. We recently asked our data scientists to complete a survey in order to learn more about the backgrounds and career journeys so far of the women in Peak’s data science team!
How did they get into data science?
All of the data scientists currently at Peak continued their education after high school to university, once they had completed their A-levels or equivalent qualifications. Having graduated with a Bachelor’s degree, 57% of Peak’s female data scientists then continued into postgraduate studies and were awarded a Master’s degree as their highest level of education.
A further 36% completed a doctorate degree, such as a PhD. For further advice on whether you need a PhD to be a data scientist, check out this blog written by some of Peak’s data science management team.
The routes into our female data scientists’ careers were diverse, and covered a wide range of different subjects at A-level. The most popular subjects were Mathematics and Further Mathematics, Sciences (including Biology and Chemistry), then English and other Modern Languages. Interestingly, only a small percentage of women had studied Computer Science or Business Studies at A-level – even though these courses cover some of the key skills required for a data science position!
This solid background in mathematics continues to give Peak’s female data scientists high levels of confidence in their mathematics and statistics skill sets. It is perhaps a lack of computer science and programming opportunities in their high school years that led to fewer women choosing this field in their undergraduate degree, which continues to be an area with average levels of confidence.
What motivates them as data scientists?
An impressive range of different machine learning models, algorithms or domains were chosen by the female data scientists, potentially due to the diversity of the Peak platform that we work on and our varied data science backgrounds. The top three were Decision Trees/Random Forest/xgboost, Optimisation and KNN/K-means Clustering with 29%, 21% and 14% of the votes respectively.
This highlights the varied technical domains that data scientists can become experts in. As data science is a relatively new and rapidly expanding field, all data scientists have many opportunities to take their careers into a wide variety of different directions.
In order to be able to write machine learning models and algorithms, data scientists use various coding languages. The preferred programming language of the women in the Peak team is a closely fought battle between R and Python. SQL is another important programming language for data scientists at Peak, and is included in the ‘Mixture’ category.
43% of the female data scientists chose R as their favorite programming language, 36% chose Mixture and the minority at 21% chose Python. In contrast, 56% of the male data scientists chose Python as their favorite programming language. The data scientists’ favorite programming language is likely to play a factor in their confidence levels in programming skills.
From the density plot above, we can see that the levels of confidence between females and males are fairly similar when R is chosen as their favorite language. However, when Python and a mixture of languages category are chosen, men are likely to be more confident in their programming skills. A question that we can ask ourselves is whether this is a valid reflection of people’s actual skills and capabilities due to their background and experience, or whether women are portraying themselves as having less confidence, or limiting their belief in themselves.
Data scientists need to have many strings to their bow, with programming skills being only one of them. In order to be able to turn a real world problem into a data science problem and successfully deliver a Decision Intelligence solution, data scientists must also possess machine learning, mathematics and statistics skills and an understanding of business and commercial applications.
The overall diversity in a team, including gender and skillset, will give a broad range of ideas to the problems that we solve. Although men may be more confident in computer science and programming, the density plot below shows that women data scientists at Peak may have more confidence in business understanding and commercial applications. All data scientists have a curiosity to understand the customers that they work with so they can help shape the future of their businesses.
The company values that resonate the most with the women in the Peak data science team are collaborative, driven, approachable, open and responsible. These traits help to build effective work connections, increase productivity at work and ensure high-performing teams remain relevant.
Collaboration, drive and openness are also echoed in the responses from male data scientists at Peak. However, the company values that inspire male data scientists the most are smart and curious. Curiosity is clearly resonating with men as it helps to develop new ideas in creative brainstorming sessions and work through complex problem solving, which are all essential skills of a data scientist.
We have seen small differences in the responses from male and female data scientists at Peak. We continue to instill confidence in other women through a variety of ways including mentoring, our new graduate scheme and by reaching out to schools and universities.
However, we also embrace differences within the team, as diversity encourages more diverse thinking and problem solving. At Peak, the shared values of being collaborative, driven and open result in a positive working environment for all data scientists.
About the author
I work as a data scientist in Peak's Supply Intelligence team. Our team combines data from across the supply chain to give a unified view of demand to help businesses optimize stock levels or resource planning. I love working on projects that first add value to our customers by extracting valuable insights from their data – and then deliver practical and innovative solutions so that our customers can become more efficient!
I enjoy applying my previous experience in logistics, process improvements and project management to successfully drive business results with Decision Intelligence.
I hope you've found some of my research into the background of Peak's women in data science interesting! Feel free to reach out to me on LinkedIn if you have any questions.