Numbers that Drive Test Results
Analytical test results are widely accepted as evidence that a material was tested for specified components and/or properties. Typically reported as Certificate of Analyses, consumers use these reports to make educated and informed decisions. Test results are almost always derived from critical data calculations and processing that are not always considered, given the ease of statistical processing software. This session will review several of the more significant background statistical processes that provide credibility to the decision rules driven by test results. First, the coveted “R2” correlation coefficient can be misleading and should not be used as the sole criterion to determine unknown concentrations. Second, measurement uncertainty, which is far different from experimental bias, provides the safety net around final test results. Confidence in experimental data relies upon these values which are often overlooked in third-party testing laboratories.
Dr Susan Audino, Chemistry Laboratory Consultant for S.A. Audino & Associates keynoted her fascinating talk back in April at the Analytical Cannabis Online Expo 2020.