skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Experimental uncertainty estimation and statistics for data having interval uncertainty.

Abstract

This report addresses the characterization of measurements that include epistemic uncertainties in the form of intervals. It reviews the application of basic descriptive statistics to data sets which contain intervals rather than exclusively point estimates. It describes algorithms to compute various means, the median and other percentiles, variance, interquartile range, moments, confidence limits, and other important statistics and summarizes the computability of these statistics as a function of sample size and characteristics of the intervals in the data (degree of overlap, size and regularity of widths, etc.). It also reviews the prospects for analyzing such data sets with the methods of inferential statistics such as outlier detection and regressions. The report explores the tradeoff between measurement precision and sample size in statistical results that are sensitive to both. It also argues that an approach based on interval statistics could be a reasonable alternative to current standard methods for evaluating, expressing and propagating measurement uncertainties.

Authors:
 [1];  [2];  [2];  [2];  [2]
  1. (Applied Biomathematics, Setauket, New York)
  2. (Applied Biomathematics, Setauket, New York)
Publication Date:
Research Org.:
Sandia National Laboratories
Sponsoring Org.:
USDOE
OSTI Identifier:
910198
Report Number(s):
SAND2007-0939
TRN: US200723%%374
DOE Contract Number:
AC04-94AL85000
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; DATA COVARIANCES; ALGORITHMS; STATISTICS; PROBABILISTIC ESTIMATION; DATA ANALYSIS; Statistics.; Uncertainty-Analysis.; Uncertainty-Mathematical models.

Citation Formats

Kreinovich, Vladik, Oberkampf, William Louis, Ginzburg, Lev, Ferson, Scott, and Hajagos, Janos. Experimental uncertainty estimation and statistics for data having interval uncertainty.. United States: N. p., 2007. Web. doi:10.2172/910198.
Kreinovich, Vladik, Oberkampf, William Louis, Ginzburg, Lev, Ferson, Scott, & Hajagos, Janos. Experimental uncertainty estimation and statistics for data having interval uncertainty.. United States. doi:10.2172/910198.
Kreinovich, Vladik, Oberkampf, William Louis, Ginzburg, Lev, Ferson, Scott, and Hajagos, Janos. Tue . "Experimental uncertainty estimation and statistics for data having interval uncertainty.". United States. doi:10.2172/910198. https://www.osti.gov/servlets/purl/910198.
@article{osti_910198,
title = {Experimental uncertainty estimation and statistics for data having interval uncertainty.},
author = {Kreinovich, Vladik and Oberkampf, William Louis and Ginzburg, Lev and Ferson, Scott and Hajagos, Janos},
abstractNote = {This report addresses the characterization of measurements that include epistemic uncertainties in the form of intervals. It reviews the application of basic descriptive statistics to data sets which contain intervals rather than exclusively point estimates. It describes algorithms to compute various means, the median and other percentiles, variance, interquartile range, moments, confidence limits, and other important statistics and summarizes the computability of these statistics as a function of sample size and characteristics of the intervals in the data (degree of overlap, size and regularity of widths, etc.). It also reviews the prospects for analyzing such data sets with the methods of inferential statistics such as outlier detection and regressions. The report explores the tradeoff between measurement precision and sample size in statistical results that are sensitive to both. It also argues that an approach based on interval statistics could be a reasonable alternative to current standard methods for evaluating, expressing and propagating measurement uncertainties.},
doi = {10.2172/910198},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Tue May 01 00:00:00 EDT 2007},
month = {Tue May 01 00:00:00 EDT 2007}
}

Technical Report:

Save / Share:
  • This report introduces a new perspective on the basic concept of dependent failures where the definition of dependency is based on clustering in failure times of similar components. This perspective has two significant implications: firstly, it relaxes the conventional assumption that dependent failures must be simultaneous and result from a severe shock; secondly, it allows the analyst to use all the failures in a time continuum to estimate the potential for multiple failures in a window of time (e.g., a test interval), therefore arriving at a more accurate value for system unavailability. In addition, the models developed here provide amore » method for plant-specific analysis of dependency, reflecting the plant-specific maintenance practices that reduce or increase the contribution of dependent failures to system unavailability. The proposed methodology can be used for screening analysis of failure data to estimate the fraction of dependent failures among the failures. In addition, the proposed method can evaluate the impact of the observed dependency on the system unavailability and plant risk. The formations derived in this report have undergone various levels of validations through computer simulation studies and pilot applications. The pilot applications of these methodologies showed that the contribution of dependent failures of diesel generators in one plant was negligible, while in another plant, it was quite significant. It also showed that in the plant with significant contribution of dependency to Emergency Power System (ESP) unavailability, the contribution changed with time. Similar findings were reported for the Containment Fan Cooler breakers. Drawing such conclusions about system performance would not have been possible with any other reported dependency methodologies.« less
  • This report introduces a new perspective on the basic concept of dependent failures where the definition of dependency is based on clustering in failure times of similar components. This perspective has two significant implications: first, it relaxes the conventional assumption that dependent failures must be simultaneous and result from a severe shock; second, it allows the analyst to use all the failures in a time continuum to estimate the potential for multiple failures in a window of time (e.g., a test interval), therefore arriving at a more accurate value for system unavailability. In addition, the models developed here provide amore » method for plant-specific analysis of dependency, reflecting the plant-specific maintenance practices that reduce or increase the contribution of dependent failures to system unavailability. The proposed methodology can be used for screening analysis of failure data to estimate the fraction of dependent failures among the failures. In addition, the proposed method can evaluate the impact of the observed dependency on system unavailability and plant risk. The formulations derived in this report have undergone various levels of validations through computer simulation studies and pilot applications. The pilot applications of these methodologies showed that the contribution of dependent failures of diesel generators in one plant was negligible, while in another plant was quite significant. It also showed that in the plant with significant contribution of dependency to Emergency Power System (EPS) unavailability, the contribution changed with time. Similar findings were reported for the Containment Fan Cooler breakers. Drawing such conclusions about system performance would not have been possible with any other reported dependency methodologies.« less
  • This report deals with the uncertainties associated with estimates of emissions from highway vehicles, which are one of the largest emission-source categories of hydrocarbons (HC) and nitrogen oxides (NO/sub x/) among all area-source categories. It presents the findings from an analysis of (1) uncertainties due to variabilities in the basic test or measurement data that are used in developing emissions-estimation parameters (EEPs) such as emission factors and (2) those due to potential biases that may be introduced in the highway-vehicle emission estimates of HC and NO/sub x/. Such EEPs and emission estimates are used in developing various emissions inventories formore » the National Acid Precipitation Assessment Program (NAPAP). The categories of emissions test data for which uncertainties due to data variability were analyzed include (1) the basic exhaust emissions test data for HC and its methane component, NO/sub x/, and temperature correction factors for HC and NO/sub x/ and (2) evaporative HC emissions test data. The sources of potential biases examined in the analysis include using annual mean ambient temperatures (instead of hourly temperatures) and using the NAPAP-selected values for vehicle speed and operating-mode distribution in computing the vehicle emission factors. Illustrative examples are presented for estimating the uncertainty that is a result of data variability in estimates of emissions from light-duty gasoline vehicles and for assessing the uncertainty that is due to systematic biases in estimates of emissions from light-duty vehicles. 20 refs., 8 figs., 6 tabs.« less
  • Probabilistic risk analyses are used to assess the risks inherent in the operation of existing and proposed nuclear power reactors. In performing such risk analyses the failure rates of various components which are used in a variety of reactor systems must be estimated. These failure rate estimates serve as input to fault trees and event trees used in the analyses. Component failure rate estimation is often based on relevant field failure data from different reliability data sources such as LERs, NPRDS, and the In-Plant Data Program. Various statistical data analysis and estimation methods have been proposed over the years tomore » provide the required estimates of the component failure rates. This report discusses the basis and extent to which statistical methods can be used to obtain component failure rate estimates. The report is expository in nature and focuses on the general philosophical basis for such statistical methods. Various terms and concepts are defined and illustrated by means of numerous simple examples.« less
  • This report presents the findings of a literature review on uncertainties associated with emissions-estimation parameters and emissions data, analyzes various types of errors and other factors that cause uncertainties in emissions-estimation parameters, and presents the interim results of an effort to compile and develop an uncertainty data base for several of these parameters. Illustrative, preliminary emissions uncertainty estimates are also provided, based on these parameter uncertainties and approximate mathematical procedures.