Research and Evaluation

We design and deliver rigorous research and evaluation projects that help organisations understand what is happening in the civil society, what difference their programmes are making, and what the evidence suggests they should do next.

Our work ranges from large-scale analytical studies using administrative data — such as our study of Scottish Charitable Incorporated Organisations for the Scottish Charity Regulator, which analysed the entire historical charity register — to primary research combining surveys, interviews, and case studies, as in our mapping of digital youth work for YouthLink Scotland. We have delivered evaluation and research projects for the Department for Culture, Media and Sport, the Economic and Social Research Council, the Gradel Institute of Charity at the University of Oxford, Power to Change, and the Young Women's Trust.

What sets our research apart is the methods we use. Many consultancies offer surveys and descriptive reporting. We also offer econometric modelling, quasi-experimental evaluation designs, longitudinal analysis, and survival modelling — the same tools used in academic research and government analysis, applied to questions the charity sector needs answered.

Our evaluation work draws on these methods to move beyond asking "what happened?" to answering harder questions: what difference did this programme make? Would the outcomes have occurred anyway? For whom did it work, and why? We design evaluations using counterfactual approaches — difference-in-differences, regression discontinuity, instrumental variables — where the data and context allow, and are transparent about what methods can and cannot tell you when they don't.

All of our research is conducted to academic standards of rigour and transparency. Both directors are active researchers with peer-reviewed publications in journals including Nonprofit and Voluntary Sector Quarterly, the Journal of Social Policy, and Public Administration Review.

Data Science and Analytics

We work with data at a scale and complexity that most research consultancies in this sector cannot.

Our flagship project, the UK Third and Civil Society Sector Database, links records from ten regulatory bodies into a single longitudinal dataset covering over 770,000 organisations across all four UK nations. Building and maintaining that infrastructure — funded by the ESRC and DCMS — has given us deep experience in the practical challenges of working with large-scale administrative data: cleaning, linking, classifying, and extracting insight from messy, real-world records.

We apply that experience to client projects. Our capabilities include data linkage across registries and administrative systems, construction of analytical datasets from raw regulatory data, automated classification and coding of organisational records, and building reproducible data pipelines that clients can maintain after we leave. We led the quantitative data assessment for the Department of Culture, Media and Sport’s review of local civil society and community data, systematically evaluating 45 data sources for quality, coverage, and analytical usability.

We also use natural language processing (NLP) and text analysis to extract insight from documents at scale. Charity annual reports, social enterprise activity and beneficiary statements, regulatory filings, grant applications — these contain rich qualitative information that is impractical to read manually when you're working with thousands of organisations. We use topic modelling, text classification, and other NLP techniques to systematically analyse large collections of documents and surface patterns that would be invisible to manual review.

We use large language models as a research tool — for document summarisation, information extraction, and systematic coding — but we use them within a framework of research rigour, with human validation and transparent methodology. We are researchers who use AI tools to answer specific questions about the civil society sector, and we understand both the capabilities and the limitations of these tools.

Our data science work is conducted in Python and R, with code shared via GitHub where appropriate. We believe in reproducible, transparent analysis.

Training and Development

We deliver training in data analytics, research methods, and computational social science for organisations and individuals who want to build their own analytical capacity. Our training is distinctive because it is grounded in the same methods we use in our research and consultancy work — we teach what we practise.

We deliver courses for major UK and international research training providers including the National Centre for Research Methods (NCRM), the Scottish Graduate School of Social Sciences (SGSSS), the Wales Institute of Social and Economic Research and Data (WISERD), and Instats. We also deliver bespoke sessions for individual organisations including the Department for Culture, Media and Sport, and University of Dundee. Our current course portfolio includes:

• Data analysis for the applied research — foundational quantitative methods using R and Python, from data management through to statistical modelling.

• Collecting digital data — collecting data from websites and APIs using Python, with a focus on ethical and legal considerations.

• Text analysis and NLP — computational methods for analysing large collections of text, including topic modelling, classification, and working with large language models.

• Data visualisation — producing clear, effective, and accessible charts and graphics from analytical outputs.

• Causal estimation methods — quasi-experimental approaches including difference-in-differences, regression discontinuity, and instrumental variables.

All of our course materials are developed and shared using open source tools and made openly available on GitHub.

If your team needs to develop specific skills — whether that's getting started with data analysis, learning to work with your own administrative data, or building capacity in a particular method — we can design a session or programme tailored to your needs and data.