Consultancy.

Every year I can help a small number of firms, non-profit organisations and individuals with their statistical or econometric problems. I can help with problems at most stages of the data pipeline, but my preferred areas of work are:

Statistical/Econometric Modelling and Machine Learning.

Design of experiments and causal Inference.

Survey design and data collection (including web scrapping).

Business Intelligence and visualisations.

Training.

Statistical/Econometric Modelling and Machine Learning.

I can use R, Stata, MatLab, Python and a variety of more niche statistical software (such as Julia or GAUSS) to estimate and interpret a wide range of statistical and econometric models. This includes model selection, prediction and classfication via Statistical/Machine learning models and methods. Below, you can find some examples of models/methods I can estimate and deploy:

Machine Learning

Survival/competing risk

Time Series

Regression

Structural Equations

Discrete Choice

Longitudinal and Multilevel

Social Networs

Textual analysis

Design of experiments and Causal Inference

If you want to test the appeal of a new product, the efficacy or effectiveness of a new technology or policy, or merely foresee people's reactions to a new innovation or policy, then experiments are arguably the most powerful methods to generate the insights you need. As an expert in Causal Inference, I can help you with all aspects of their design and analysis, from sample size calculations and randomization, all the way to addressing problems of non-compliance or cross-over. If your experiment will be online, I can advice on software and platforms to deploy your experiment and recruit participants. No experimental data? Then I can advice on quasi-experimental designs and other "conditional-on-observable" methods that can help you get closer to the answer you are looking for.

Online experiments

Sample size calculations

Randomization

Policy evaluation

Non-compliance

Matching methods

Marketing/Clinical trials

Instrumental Variables

Observational Studies

Survey design and data collection (including web scrapping).

If you need to collect data from users, clients or target populations, I can help you to design your sample and your data-collection instruments. This could include from the design of a frame, and the selection of clusters or strata, all the way to advance questionnaire designs and incentive schemes to minimise non-response, item response biases, and other deficiencies in your final dataset. If your data collection will proceed online, I can advise you on software, platforms and the design of scrappers and crawlers.

Business Intelligence and visualisations.

I can collaborate with firms and advice them in the design of their business intelligence plans, particularly in the areas of data preparation, modeling, visualisation and reporting in Tableau, Power BI, Shiny or other comparable applications. These tend to be larger bespoke projects, therefore just get in touch to discuss your needs and how I can help you or your firm.

Training.

This is a service I offer only occasionally, and only to people and organisations that need to adopt or learn new statistical methods for their everyday operations. Specifically, this service is not available to academics, students or educational firms. I can offer tailored training to teams and organisations on most aspects of statistical methods and software. This can involve face-to-face or online training or advice on the preparation of teaching materials that your team can use as a reference (although I will not be involved in the production of teaching materials).