A common quality assurance practice to assess the validity of experimental methods is to collect duplicate samples and calculate their precision. This process allows scientists to detect erroneous measurements by determining if the measurement precision is too low. Most methods to calculate precision, such as relative percent difference and root mean square error, are only valid under strong assumptions – eg. normality and homoscedasticity – that are often not true in practice. In this study, we developed a robust method to model the precision of duplicate datasets that imposes fewer assumptions and is therefore valid in a larger array of cases. Furthermore, we expanded this method so that it can be used to detect outliers in duplicate datasets and to create confidence intervals around both duplicate and singlicate samples. The end product of this study will be an application so users can easily detect outliers and create confidence intervals for their datasets.

Robust precision modeling in duplicate data sets: Applications for outlier detection and constructing confidence intervals

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.