masthead image
 

The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.

John Tukey

Dustin Lennon Principal / Lead Statistician

Dustin combines an impressive academic pedigree with a breadth of experience across all levels of the data analysis pipeline. He delivers highly interpretable, actionable analyses that inform data driven decision making.

Dustin has several years of industry experience in statistics, data science, and machine learning. He recently left a Senior Data Scientist role at Microsoft where he worked as an internal consultant for product teams interested in adopting best practices for controlled experimentation.

Curriculum Vitae 

dustin's photo
email
dustin at inferentialist dot com
phone
206 291 8893

Data Science Consulting An Analysis Playbook

There are several questions worth asking before undertaking a statistical analysis. Here are a few, to set expectations for an initial engagement with Inferentialist Consulting.

First and foremost

  • What is the business problem to be solved?

From here, there are typically two broad classes of problems: either, a data driven decision is required, or, there is a need to characterize a data generating process.

Data driven decision making

  • What is the decision space?
  • What metrics will the decision impact?
  • What data do we need and under what conditions will that data be collected?
  • What is the consensus criteria for executing a decision strategy?

Intelligent data characterization

  • What exploratory data analyses have been done?
  • Has data quality been assessed?
  • Are interpretable results required?
    • Will confidence intervals be useful?
    • Is there a need for a statistical model?
  • Is a prediction engine sufficient?
  • Are there potential scaling issues?
    • Can the data be intelligently sampled?