What is Social Data Science?

The UCL Social Data Institute is about training undergraduate students in social data science to give them skills to understand, interpret, evaluate - and importantly - produce research. Our starting point is interesting, socially relevant questions, such as:

• How can we reduce obesity?
• Why don't people vote?
• Which policies work best to reduce poverty in global south countries?

A key starting point is understanding quantitative methods (QM). There isn't a single definition of QM, but here are a few examples:

• Quantitative methods is about collecting and analysing numerical data
• Statistics is 'the science of organizing and analyzing information to make the information more easily understood' (Salkind 2004)
• Quantitative methods use real-world data to uncover differences or similarities in attitudes, behaviour or phenomena.

Tools and techniques

QM isn't just maths or statistics, though it does involve some of both-it's about using tools and techniques to answer important social, political and economic policy questions. To answer important questions we have to think about:

• Asking good research questions
• How we can design research plans to best answer our questions
• How we measure important concepts
• The right method to use, from simple techniques like descriptive statistics and linear regression to more complex approaches like cartography, causal inference, natural language processing or social network analysis
• Association and causation: educational attainment and income may be related, but is education the cause of higher incomes?
• Visualizing relationships: a picture really is worth a 1,000 words

By the end of the programme, students are familiar not just with the basics of statistics but also with a range of more advanced techniques that are often referred to as ‘data science’. This includes machine learning, text analysis, the design of randomised control trials and geographic methods of spatial analysis. Our emphasis on these advanced techniques, more commonly taught at the postgraduate level, is a unique feature of UCL’s Social Data Science Programme that really sets our graduates apart from their peers at other universities.

Coding

The social data science program also teaches students to code using the R programming language, as well as other languages such as python in some optional modules. Coding is an essential skill to learn for data scientists. It is through code that we implement our methods, produce results and visualise them. This leaves students with a lifelong skill that is in extremely high demand from employers.

The Demand

Jobs, jobs, jobs! Q-Step was established as a strategic response to the shortage of quantitatively-skilled social science graduates. Employers from public, private and the third-sector desperately need people who can draw out insights from quantitative data to inform policymaking, organizational strategy and resources.

The skills that our graduates learn are highly transferable across industry or sector. As a result, our students have gone on to postgraduate study at top universities around the world and are successfully working in data-intensive roles across the private, public and voluntary sectors.