[Estimation of reproducibility and repeatability in microbiological ring trials--robust versus conservative methods].

Area of Science:

• Microbiology
• Statistical analysis
• Laboratory science

Background:

• Standardized methods for microbiological ring trials have advanced.
• Statistical strategies, including repeatability and reproducibility estimations, are crucial.
• Robust statistical methods are gaining traction alongside traditional variance analysis.

Purpose of the Study:

• To compare and discuss traditional and robust statistical methods for analyzing microbiological ring trial data.
• To evaluate the performance of robust estimators when data does not follow a normal distribution.
• To highlight the importance of data visualization for identifying critical laboratory results.

Main Methods:

• Comparison of variance analysis with robust statistical methods.
• Analysis of data from recent microbiological ring trials.
• Application of graphical data presentation techniques.

Main Results:

• Robust methods are increasingly discussed for parameter estimation in ring trials.
• The behavior of robust estimators is less understood when normal distribution assumptions are violated.
• Graphical data presentation aids in identifying outlier laboratories.

Conclusions:

• Robust statistical methods offer alternatives to variance analysis for ring trials.
• Critical evaluation of data and visualization are essential, particularly with non-normally distributed data.
• Identifying laboratories with critical results is facilitated by robust analysis and graphical tools.