Begin a meaningful career right here
Modeling and Statistics PhD Internship
It is of interest to identify statistical designs and analyses/methods that can increase learning while reducing cost. It is common in our industry to conduct expensive large scale studies. As alternatives, we would like to examine the potential of using sequential designs, Bayesian methods, adaptive designs, or other statistical techniques which enable us to combine learning from multiple smaller studies. Success is either being able to derive meaningful findings from the combined studies already run, or enable direction where we can investigate further with new focused studies. The successful intern will provide a literature review and identify approaches to implement now or to consider developing that will respond to a variety of scenarios to enable sequential and/or adaptive learning. The proposed ideas will be considered for implementation in consumer research, clinical studies, and manufacturing.
Key Priorities, Deliverables, or Outcomes of This Position:
1) Literature review of methods for sequential design and analysis that can properly respond to changes that need to be implemented between experiments
2) Select one or more of the methods explored in the literature review, preferably methods that (a) enable us to adjust design parameters after each stage of learning and (b) are compatible with our interest in lean, small-scale experiments
3) Determine what is needed to develop these methods: what type of data we can collect, what types of experimental designs we can use, how we expect to analyze the new data in combination with knowledge we already have. Determine what software needs to be developed and/or deployed to support these objectives.
4) Develop as much as is possible for at least one of the selected methods.
1) Knowledge of experimental design, especially sequential design and adaptive design
2) Background on statistical methods that Incorporate use of prior knowledge
3) Experience in real life data analysis
4) Ability to make recommendations on experimental designs based on outcomes of existing experiments
Join our Talent Community to receive via email information about P&G, upcoming events, and job alerts.
SIGN UP FOR: