Premium
Communicating statistical conclusions of experiments to scientists
Author(s) -
Otava Martin,
Mylona Kalliopi
Publication year - 2020
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/qre.2697
Subject(s) - computer science , set (abstract data type) , visualization , statistical analysis , design of experiments , scaling , data science , machine learning , data mining , statistics , mathematics , geometry , programming language
Abstract The manuscript introduces a framework for presenting the results of the statistical analysis of experiments with multiple responses and multiple factors. We propose a utilisation of factors scaling to enable a transformation that combines main effects, quadratic effects and interactions into a meaningful summary that allows the scientist/experimenter to immediately recognise the most influential factors for a given response. The framework does not replace the thorough evaluation of the results but provides a clear high‐level summary of the relative importance of findings. The visualisation of such factor importance, using intensity heatmaps, allows the immediate understanding of the results across multiple responses that efficiently guides a following detailed analysis of certain responses and factors and contributes in designing subsequent experiments. The methodology is applied to a real industrial experiment and to a simulated data set with a larger number of responses and factors.