What We Do

Predictive Models — With over 30 years of experience, Z-Stats specializes in large predictive models — models that predict consumer responses and how consumers and products interact based on Sensory and Bench Test Results.  A variety of methods can be used based on the size of the data set and the desired outcomes.  These include Elastic Nets, LASSO, Stepwise Regression, Self-Validated Ensemble Models, and Machine Learning Techniques for large data sets.  We will guide you through the entire process of picking the best inputs for the data and designing outputs that generate actionable results.

Sensory Testing

Z-Stats can help from start to finish with your sensory project.

  • choosing the best methods,
  • determining sample size,
  • statistical analysis,
  • developing predictive models that link sensory and consumer

DOEs

Design of Experiments is a powerful tool to increase the ROI of your testing. Z-Stats can guide you in formulating and analyzing DOEs to maximize your testing money. The primary tool that we use is JMP’s Easy DOE. This is such a powerful tool that we can train your organization in how to use it for future studies. Z-Stats’s perspective is that DOEs should be a tool regularly used to guide product development. DOEs are also extremely useful in Sensory Panel Validation.