PSA2016: The 25th Biennial Meeting of the Philosophy of Science Association

Full Program »

The Evaluation and Interpretation of Systematic Uncertainty

Reports of experimental results often distinguish between statistical and systematic uncertainty to characterize measurement outcomes. This project examines two questions regarding the evaluation and interpretation of systematic uncertainty: (1) What do methods whereby systematic uncertainty is evaluated tell us about the nature of systematic uncertainty and its relationship to statistical uncertainty? (2) What role do systematic uncertainty assessments play in the use of measurement results (e.g., comparisons between different experimental results or tests of theoretical claims)?

This account builds on two recent contributions to the philosophical study of measurement: Eran Tal’s model-based account of measurement, according to which the evaluation of measurement accuracy is the outcome of a comparison amongst predictions drawn from a model of the measurement process (Tal 2016), and Hugo Beauchemin’s discussion of systematic uncertainty assessment as essential to determining the sensitivity of measurement results in High Energy Physics (HEP) (Beauchemin 2015).

The present account adds to these an argumentation perspective. The presentation of measurement results involves an argument, based on a model of the measurement process for a conclusion about the value of the measurand. The distinction between statistical and systematic uncertainty corresponds to a distinction between: (1) uncertainty about the value of the measurand claimed in the conclusion (statistical), the evaluation of which assumes the model of the measuring process to be adequate and the premises of the argument to be free of error; and (2) uncertainty about the adequacy of the model of the measuring process and the possible errors in premises assigning values to parameters of that model (systematic). Consideration of methods of evaluating systematic uncertainty then supports the claim that the evaluation of systematic uncertainty should be understood as a form of robustness analysis, performed on a model of the measurement process.

Turning to question (2), the impact of systematic uncertainty on scientific conclusions drawn from measurement results depends largely on the use to which a given result is put. Large statistical and systematic uncertainties alike reduce the sensitivity of a measurement, which is a function of the differences amongst theoretical predictions about the value of a measurand as well as the uncertainty of its measured value. But large systematic uncertainties carry the additional implication that the measurement process itself is not well-understood. The differential relevance of such judgments in cases of agreement and disagreement between theoretical predictions and measurement results are discussed.

These concepts are illustrated through a discussion of a recent measurement result in HEP: the measurement of the top-anti-top production cross section by the ATLAS collaboration at the Large Hadron Collider (Aad, Abbot, et al. 2012).

Aad, G., Abbott, B. et al. (2012). Measurement of the top quark pair production cross-section with ATLAS in the single lepton channel. Physics Letters B, 711, 244–263.
Beauchemin, P.-H. (2015). Autopsy of measurements with the ATLAS detector at the LHC. Synthese, 1–38. Retrieved from doi: 10.1007/s11229-015-0944-5
Tal, E. (2016). Making time: A study in the epistemology of measurement. The British Journal for the Philosophy of Science, 67, 297–335.

Author Information:

Kent Staley    
Saint Louis University


Powered by OpenConf®
Copyright©2002-2015 Zakon Group LLC