This includes the design of questionnaires, sampling techniques, modes of data collection, research designs, computation of survey weights, data integration, descriptive statistics, inferential statistics, nonresponse analysis and adjustments, data modelling, design of tables and figures summarising results that can help your business to understand markets, relationships and transactions.

Machine learning is a type of Artificial Intelligence which allows organisations to meet complex challenges, from creating new customised services to predicting new products. Using algorithms, the application of machine learning enables computer systems to carry out tasks based on patterns and inference rather than explicit instructions. It can be applied for across all sectors and industries, e-commerce, healthcare, charities, finance and more. It can be used for classification purposes, prediction and optimisation to reduce costs, increase sales, productivity and customer satisfaction.

Geographical Information System (GIS) is the study, manipulation, analysis and visualisation of data with a geographical component. Adding geographic locations to business data and mapping is an innovative business analytics solution facilitating the discovery of new insights for planning, organising and decision-making.

A group of statistical techniques aimed at developing estimates for areas (e.g. boroughs, counties, school catchment areas, National Health Service Trust areas) or domains (e.g. socio-demographic groups) with small sample sizes. These estimates can help with the formulation of social and economic policies, allocation of government funds, regional planning and business decision-making.

Open Data are data that are available to everyone to access, use and share. Code-Switch Research Consultants have expertise in using the main open-access datasets to measure social and economic issues in the UK and Europe, as well as data recorded by crowdsourced platforms and sources of Big Data. Our Researcher Consultants have developed path-breaking work aimed at reducing the potential biases that might affect some of those sets of data.

Time Series Analysis is used to analyse data that have been collected over time and is useful in developing an understanding the past and predicting the future. This technique can be used for things such as stock market analysis, economic forecasting, inventory studies, budgetary analysis, census analysis, yield projection, sales forecasting and identifying potential efficiency improvements.

State Space Modelling is a popular technique for forecasting and smoothing time series data. It is useful for making inferences about factors which are difficult, or impossible, to measure.

Bayesian statistics provide intuitive and meaningful inferences, answering complex questions cleanly and exactly by making use of all available information. Bayesian Statistics can identify causal relationships and can be usefully applied to improve decision-making, marketing and measuring the impact of business initiatives on performance.

Volatility models aim to capture the dynamics of volatility. In finance, it is well known that the volatility of returns is not constant over time. Specifically, there are periods of low volatility followed by periods of high volatility. This technique is useful for investing and forecasting purposes.

Quantile Regression is a comprehensive analysis of the relationship between variables. It’s practical applicability for businesses can range from calculating the percentage of buyers who access a business and make a purchase, identifying the relationship between an organisation’s performance and gender diversity of their boards, to the effect of innovation on growth for fast and slow growing businesses.

Quantile Treatment Effect compares two populations that differ in some way. A practical application of this technique could be to determine the difference in the growth of two groups of businesses which may be similar in terms of revenue, size, and maturity but where there is a variable that differs among the them, such as innovation.

Some time series are subject to breaks which change the underlying model. Using Gross Domestic Product (GDP) as an example, the process that drives GDP during recessions may not explain GDP during expansions. Regime switching models allow the model parameters to switch between two or more regimes.