Increase the Autonomy & Confidence of Research Teams
Objective: Empowering Engineers with Statistical Knowledge and Optimize R&D Activities
For engineers active in research and development, statistics is an essential part of their toolbox. Beyond the data analysis methods it offers, statistics is above all a way of understanding science and decision-making in the context of uncertainty.
To this end, Creascience has been offering training programs for over 20 years focused on mastering statistical concepts and thinking in the context of applications in engineering and R&D. A recurring request from participants following these programs is the possibility of maintaining a longer-term connection and support in order to more deeply embed the approach taught in research practices. Consequently, we have implemented this coaching program to meet this need.
The Concept: Supporting a Small Group to Make it Autonomous in the Long Term
The proposed program is aimed at engineering teams, typically of 5 to 10 people, interested in improving their use of statistical tools and increasing their autonomy. While training remains the starting point and a central theme of the program, a variety of complementary activities are offered to strengthen understanding and mastery of the tools. The schedule and the aspects covered are adapted to the needs and availability of the participants. In all cases, the underlying principle is to maintain their interest by ensuring that they are regularly exposed to a variety of statistical issues.
While the coach may, in certain circumstances, handle the statistical aspect of an urgent or particularly complex project, engineers are encouraged to take charge of the interventions themselves under the coach’s supervision. The coach shares their experience, provides advice, and guides the team on the tools and methods to use..
The Method: Combining training, Supervised Study and Ad Hoc Support
Unlike our classic training sessions which focus on a few weeks, the coaching program extends over several months and combines different interventions :
- A long-term training program covering essential concepts as well as more advanced methods from an applied perspective.
- Case studies focused on analyzing previously collected data and implementing experiments.
- Applied workshops aimed at developing specific skills: use of software, interpretation of results, presentation of statistical aspects in a report, etc.
- An on-demand support service that can take several forms, the most popular being regular support sessions to discuss problems encountered and possible solutions with the team or individual members.
The Program: From the Integration of Basic Concepts to the Understanding of Advanced Tools
Duration
This program generally lasts one or two years, depending on the availability of the coached team and the variety of tools to be mastered. We nevertheless offer an initial trial period of 6 months, renewable, to allow the team to ensure the suitability of the program and the dynamics with their needs.
Recommended Elements
The following table provides an overview of the most frequently covered topics. In practice, the exact program is established based on the client’s specific context and identified needs.
Basic Concepts
- Descriptive statistics
- Notion of distribution
- Inferential statistics principle
- Sample size calculations
The Design of Experiments
- Base factorial design
- Analysis of variance
- Fractional designs
- Optimisation designs
- Algorithm-based designs
Data Modelling
- Visualisation tools
- Multiple Linear regression
- Goodness-of-fit measures
- Logistic models
Reliability Analysis
- Censoring, runouts, etc.
- Parametric models: Weibull and its alternatives
- Regression for reliability data
Machine Learning Principles
- Predictive models
- Cross-validation
- Algorithms based upon trees
- Use of models
Practical Aspects
Flexibility remains a core principle of the program. Preliminary meetings allow for broad outlines to be established while remaining open to the possibility of making rapid adjustments when new priorities arise. However, a typical sequence can be identified, at least for the beginning of the program :
- Months 1 and 2: training on the fundamentals of statistics and experiment planning (4-6 sessions of 3 hours).
- The actual coaching begins during the second month.
- Months 3 and following: at least one 3-hour training session per month.
Months 1 – 2
- Training on the fundamental of statistics and design of experiment
- 4 to 6 sessions (3h/session)
Month 2 – 6
- Coaching
- At least a 3h training session per month
Would you like to learn more?
