Analysis

We work in collaboration with our customers to co-design analysis plans and projects to ensure that they meet their needs. We also work in partnership with colleagues across the National Centre for Social Research (NatCen) on mixed-methods projects, integrating findings from quantitative analysis into wider research projects.
Our team has extensive experience of sourcing, preparing, analysing and reporting on analysis for grant-funded projects and tendered work. We are regularly involved in all stages of the research process, including proposing, designing, delivering and reporting on quantitative analysis for different audiences. Project work can include:
- Scoping out existing datasets that can be used to address specific research
- Producing an analysis plan, independently or in collaboration with customers, to address a specific research question or questions given the data and resources available
- Carrying out secondary data analysis using data supplied by clients or collaborators
- Producing visualisations presenting findings from analysis in an accessible and engaging manner
- Writing lay summaries of methods and findings for non-expert audiences
- Writing technical reports, or appendices, to detail data and methods used
- Providing stand-alone syntax for data management and/or analysis to facilitate replicability
For each analysis project, initial scoping will be carried out to identify the most appropriate analytical approach to use. This may include:
- Linear or logistic regression – to estimate the relationships between explanatory variables and continuous, binary or categorical outcome variables while controlling for other relevant factors
- Longitudinal analysis – to investigate if or how factors of interest have changed over time
- Multi-level modelling – to take into account the nested structure of data (for example pupils in schools, or citizens in countries)
- Factor analysis (or other dimension reduction techniques) – to identify broader, latent concepts that underpin multiple survey items
- Latent class analysis – to estimate classes of people with similar profiles based on their responses to individual survey items
- Imputation – to overcome issues caused by missing data
- Data linkage – to combine and analyse data available from different sources