Monte Carlo simulation of expert judgments on human errors in chemical analysis—A case study of ICP–MS

• The Monte Carlo method was applied for simulation of expert judgments on human errors.
• An expert behavior was modeled using plausible probability mass functions.
• Possible errors in analysis of geological samples by ICP–MS were treated as a case study.
• The error scenarios should be taken into account for quality risk management in a laboratory.
• Robustness of the error quantification scores is satisfactory for the quality system purposes.

Monte Carlo simulation of expert judgments on human errors in a chemical analysis was used for determination of distributions of the error quantification scores (scores of likelihood and severity, and scores of effectiveness of a laboratory quality system in prevention of the errors). The simulation was based on modeling of an expert behavior: confident, reasonably doubting and irresolute expert judgments were taken into account by means of different probability mass functions (pmfs). As a case study, 36 scenarios of human errors which may occur in elemental analysis of geological samples by ICP–MS were examined. Characteristics of the score distributions for three pmfs of an expert behavior were compared. Variability of the scores, as standard deviation of the simulated score values from the distribution mean, was used for assessment of the score robustness. A range of the score values, calculated directly from elicited data and simulated by a Monte Carlo method for different pmfs, was also discussed from the robustness point of view. It was shown that robustness of the scores, obtained in the case study, can be assessed as satisfactory for the quality risk management and improvement of a laboratory quality system against human errors.

Human errors in ICP–MS analysis of geological samples: a histogram of the error likelihood score P⁎ (%) corresponding to judgments of a reasonably doubting expert, simulated by the Monte Carlo method.Figure optionsDownload as PowerPoint slide