Skip to content
Methodnex

Data Science & Econometrics

Models whose assumptions are stated, tested and reported, so the conclusions hold.

Request a Scoped Quotation

Questions we help answer

For questions about cause, change over time or prediction, we build econometric and machine-learning models and test how far they can be trusted. Reports say what the model shows, what it does not, and how it was validated.

  • Does this policy or programme cause the change I observe, or is something else at work?
  • How do I analyse panel data on firms, countries or patients over time?
  • Can I forecast this series, and how wide is the uncertainty?
  • Is a machine-learning model appropriate here, and how do I explain it?

Scope of service

What you receive

  • Panel-data models with the diagnostic tests that justify them
  • Time-series analysis and forecasts with stated uncertainty
  • Efficiency analysis using data envelopment analysis (DEA)
  • Machine-learning models with validation and explainability reports
  • Reproducible R or Python notebooks and documented data pipelines

What is not included

  • We do not present correlations as causal effects.
  • We do not send confidential datasets to unapproved third-party tools or AI models.
  • We do not deliver a model without telling you where it fails.

Methods and tools, and when we use them

  • Fixed and random effects, dynamic panels

    When the same units are observed over several periods.

  • ARIMA, VAR and cointegration

    When the question concerns dynamics, shocks or long-run relationships in time series.

  • Difference-in-differences and related designs

    When a policy or intervention reached some units and not others.

  • Tree-based models with cross-validation

    When prediction matters more than a single coefficient and the data are large enough.

How the work runs

  1. 1

    Intake

    You describe the project through the request form or a call. No data are needed at this stage.

  2. 2

    Methodological scoping

    We review the design and agree the deliverables, exclusions, timeline and fee in writing.

  3. 3

    Execution

    The assigned consultant does the work and records methods, assumptions and changes.

  4. 4

    Quality assurance

    A second specialist reviews the methods and outputs before release.

  5. 5

    Handover

    You receive the deliverables, the code and a walkthrough session, with the agreed revision rounds.

Who this is for

  • Economics and finance researchers
  • Business schools and doctoral candidates
  • Public bodies evaluating policy
  • Companies with operational or customer data

Questions about this service

Do you work in Stata, R or Python?

All three. We deliver scripts or notebooks that rerun from the raw data.

How much data do I need?

It depends on the method. We check this before quoting and tell you plainly if the data cannot support the question.

Will you use AI tools on my data?

Only tools and environments approved for the project, and never without your knowledge. Confidential data are not sent to public AI services.

Can you review a model built by someone else?

Yes. An independent review covers the specification, the diagnostics and whether the conclusions follow from the results.

Other services

Request a scoped quotation

Describe the project in a few minutes. Complex research is quoted after scoping, never from a price list.

Request a ProposalAll services