Evaluation of uncertainty in measurements

Similar documents
EVALUATION OF UNCERTAINITY IN MEASUREMENTS

best estimate (mean) for X uncertainty or error in the measurement (systematic, random or statistical) best

Chapter 8. Inferences about More Than Two Population Central Values

Summary of the lecture in Biostatistics

CHAPTER VI Statistical Analysis of Experimental Data

ENGI 4421 Joint Probability Distributions Page Joint Probability Distributions [Navidi sections 2.5 and 2.6; Devore sections

Lecture 7. Confidence Intervals and Hypothesis Tests in the Simple CLR Model

Lecture 3 Probability review (cont d)

Functions of Random Variables

ENGI 4421 Propagation of Error Page 8-01

Chapter 5 Properties of a Random Sample

Chapter 13, Part A Analysis of Variance and Experimental Design. Introduction to Analysis of Variance. Introduction to Analysis of Variance

Quantitative analysis requires : sound knowledge of chemistry : possibility of interferences WHY do we need to use STATISTICS in Anal. Chem.?

A Study of the Reproducibility of Measurements with HUR Leg Extension/Curl Research Line

Lecture Notes Types of economic variables

STATISTICAL PROPERTIES OF LEAST SQUARES ESTIMATORS. x, where. = y - ˆ " 1

Estimation of Stress- Strength Reliability model using finite mixture of exponential distributions

ESS Line Fitting

Objectives of Multiple Regression

Lecture 1 Review of Fundamental Statistical Concepts

STA 108 Applied Linear Models: Regression Analysis Spring Solution for Homework #1

UNIVERSITY OF OSLO DEPARTMENT OF ECONOMICS

X ε ) = 0, or equivalently, lim

Continuous Distributions

Bayesian Classification. CS690L Data Mining: Classification(2) Bayesian Theorem: Basics. Bayesian Theorem. Training dataset. Naïve Bayes Classifier

Lecture 3. Sampling, sampling distributions, and parameter estimation

Chapter 14 Logistic Regression Models

Simulation Output Analysis

THE ROYAL STATISTICAL SOCIETY GRADUATE DIPLOMA

Lesson 3. Thermomechanical Measurements for Energy Systems (MENR) Measurements for Mechanical Systems and Production (MMER)

Chapter 13 Student Lecture Notes 13-1

1. A real number x is represented approximately by , and we are told that the relative error is 0.1 %. What is x? Note: There are two answers.

Lecture 12 APPROXIMATION OF FIRST ORDER DERIVATIVES

LECTURE - 4 SIMPLE RANDOM SAMPLING DR. SHALABH DEPARTMENT OF MATHEMATICS AND STATISTICS INDIAN INSTITUTE OF TECHNOLOGY KANPUR

Sampling Theory MODULE V LECTURE - 14 RATIO AND PRODUCT METHODS OF ESTIMATION

hp calculators HP 30S Statistics Averages and Standard Deviations Average and Standard Deviation Practice Finding Averages and Standard Deviations

ENGI 3423 Simple Linear Regression Page 12-01

Practical guide for the validation, quality control, and uncertainty assessment of an alternative oenological analysis method (Resolution 10/2005)

Outline. Point Pattern Analysis Part I. Revisit IRP/CSR

( ) = ( ) ( ) Chapter 13 Asymptotic Theory and Stochastic Regressors. Stochastic regressors model

BASICS ON DISTRIBUTIONS

Analysis of Variance with Weibull Data

Simple Linear Regression

MEASURES OF DISPERSION

Simple Linear Regression

4. Standard Regression Model and Spatial Dependence Tests

Econometric Methods. Review of Estimation

(Monte Carlo) Resampling Technique in Validity Testing and Reliability Testing

Point Estimation: definition of estimators

Some Applications of the Resampling Methods in Computational Physics

Linear Regression with One Regressor

Chapter 3 Sampling For Proportions and Percentages

STK4011 and STK9011 Autumn 2016

1. The weight of six Golden Retrievers is 66, 61, 70, 67, 92 and 66 pounds. The weight of six Labrador Retrievers is 54, 60, 72, 78, 84 and 67.

LINEAR REGRESSION ANALYSIS

STA302/1001-Fall 2008 Midterm Test October 21, 2008

Module 7. Lecture 7: Statistical parameter estimation

PROPERTIES OF GOOD ESTIMATORS

Module 7: Probability and Statistics

HOOKE'S LAW. THE RATE OR SPRING CONSTANT k.

Ordinary Least Squares Regression. Simple Regression. Algebra and Assumptions.

Unimodality Tests for Global Optimization of Single Variable Functions Using Statistical Methods

Bootstrap Method for Testing of Equality of Several Coefficients of Variation

12.2 Estimating Model parameters Assumptions: ox and y are related according to the simple linear regression model

X X X E[ ] E X E X. is the ()m n where the ( i,)th. j element is the mean of the ( i,)th., then

Introduction to local (nonparametric) density estimation. methods

Chapter 11 The Analysis of Variance

The expected value of a sum of random variables,, is the sum of the expected values:

Special Instructions / Useful Data

Multiple Regression. More than 2 variables! Grade on Final. Multiple Regression 11/21/2012. Exam 2 Grades. Exam 2 Re-grades

Lesson 3. Group and individual indexes. Design and Data Analysis in Psychology I English group (A) School of Psychology Dpt. Experimental Psychology

Chapter 8: Statistical Analysis of Simulated Data

EXPERIMENTAL ERRORS. There are primarily two kinds of errors that affect experimental results:

UNIVERSITY OF OSLO DEPARTMENT OF ECONOMICS

22 Nonparametric Methods.

On Fuzzy Arithmetic, Possibility Theory and Theory of Evidence

PGE 310: Formulation and Solution in Geosystems Engineering. Dr. Balhoff. Interpolation

Multiple Linear Regression Analysis

BIOREPS Problem Set #11 The Evolution of DNA Strands

CODING & MODULATION Prof. Ing. Anton Čižmár, PhD.

Confidence Intervals for Double Exponential Distribution: A Simulation Approach

Determination of woven fabric impact permeability index

Chapter 4: Elements of Statistics

Median as a Weighted Arithmetic Mean of All Sample Observations

Statistics MINITAB - Lab 5

STK3100 and STK4100 Autumn 2017

ECONOMETRIC THEORY. MODULE VIII Lecture - 26 Heteroskedasticity

COV. Violation of constant variance of ε i s but they are still independent. The error term (ε) is said to be heteroscedastic.

We have already referred to a certain reaction, which takes place at high temperature after rich combustion.

Sequential Approach to Covariance Correction for P-Field Simulation

= 1. UCLA STAT 13 Introduction to Statistical Methods for the Life and Health Sciences. Parameters and Statistics. Measures of Centrality

1 0, x? x x. 1 Root finding. 1.1 Introduction. Solve[x^2-1 0,x] {{x -1},{x 1}} Plot[x^2-1,{x,-2,2}] 3

Johns Hopkins University Department of Biostatistics Math Review for Introductory Courses

Third handout: On the Gini Index

Application of Calibration Approach for Regression Coefficient Estimation under Two-stage Sampling Design

Midterm Exam 1, section 1 (Solution) Thursday, February hour, 15 minutes

Lecture 2: The Simple Regression Model

Johns Hopkins University Department of Biostatistics Math Review for Introductory Courses

Analysis of System Performance IN2072 Chapter 5 Analysis of Non Markov Systems

{ }{ ( )} (, ) = ( ) ( ) ( ) Chapter 14 Exercises in Sampling Theory. Exercise 1 (Simple random sampling): Solution:

Transcription:

Evaluato of ucertaty measuremets Laboratory of Physcs I Faculty of Physcs Warsaw Uversty of Techology Warszawa, 05

Itroducto The am of the measuremet s to determe the measured value. Thus, the measuremet begs wth specfyg the quatty to be measured, the method used for measuremet (e.g. comparatve, dfferetal, etc.) ad the measuremet procedure (set of steps descrbed detal ad appled whle measurg wth the selected measurg method). I geeral, the result of a measuremet s oly a appromato or estmate of the value of the specfc quatty subject to measuremet, that s, the measurad. Thus, the result of measuremet s complete oly whe accompaed by a quattatve statemet of ts ucertaty. Iteratoal Stadard Orgazato (ISO) prepared Gude to the Epresso of Ucertaty Measuremet, whch s deftve documet descrbg orms ad procedures the measuremets ucertaty evaluato. Based o the teratoal ISO stadard, Polsh orm Wyrażae epewośc pomaru. Przewodk was accepted the 999.

Sources of ucertaty a measuremet complete defto of the measurad; mperfect realzato of the defto of the measurad; orepresetatve samplg the sample measured may ot represet the defed measurad; adequate kowledge of the effects of evrometal codtos o the measuremet or mperfect measuremet of evrometal codtos; persoal bas readg aalogue strumets; fte strumet resoluto or dscrmato threshold; eact values of measuremet stadards ad referece materals; eact values of costats ad other parameters obtaed from eteral sources ad used the data-reducto algorthm; appromatos ad assumptos corporated the measuremet method ad procedure; varatos repeated observatos of the measurad uder apparetly detcal codtos.

Types of measuremets Drect measuremet measured quatty ca be drectly compared wth the eteral stadard, or the measuremet s made usg a sgle strumet gvg result straghtaway seres of measuremets gross error Idrect measuremet measurg oe or more physcal quattes to determe quatty depedet o them

Basc deftos () Measuremet ucertaty - parameter assocated wth the result of measuremet characterzg dsperso of the values attrbuted to the measured quatty Stadard ucertaty u( the ucertaty of measuremet epressed as a stadard devato. Ucertaty ca be reported three dfferet ways: u, u( or u(accelerato), where quatty ca be epressed also words ( the eample s accelerato). Please ote, that u s a umber, ot a fucto.

Idrect measuremets Laboratorum Fzyk, Wydzał Fzyk, Poltechka Warszawska

Basc deftos () Type A evaluato of ucertaty the evaluato of ucertaty by the statstcal aalyss of seres of observatos. Result of a seres of measuremets: mea value Assumptos: Dstrbuto fucto s symmetrcal probablty for results smaller as well as bgger tha mea value are the same The bgger devato from the mea value the lower probablty Result: for bgger umber of measuremets observed dstrbuto of data pots s smlar to Gauss fucto Eample of a Type A evaluato of ucertaty: the stadard devato of a seres of depedet observatos ca be calculated, or least squares method ca be appled to ft the data wth a curve ad determe ts parameters ad ther stadard ucertates.

Gauss dstrbuto Dstrbuto for cotous varable : ( ep μ epected value σ stadard devato ( (d 3 3 ( d ( d ( d 0,683 0, 997 0, 954 Gauss dstrbuto for fte umber of pots: epected value s equal to mea value, stadard devato s equal to stadard devato of a mea value Type A stadard ucertaty for a seres of measuremets s equal to stadard devato of a mea value u( s ( ) (

Basc deftos (3) Type B evaluato of ucertaty the evaluato of ucertaty by meas other tha the statstcal aalyss of seres of observatos, thus usg method other tha type A. Type B evaluato of stadard ucertaty s usually based o scetfc judgmet based o eperece ad geeral kowledge, ad s a skll that ca be leared wth practce. Assumpto: uform dstrbuto probablty s costat the whole terval determed by measuremet ad calbrato ucertaty calbrato ucertaty (due to measuremet devce vestgator ucertaty (due to vestgator s epermetal sklls e ) u( 3 ( 3 Combato of ucertates u( s ( 3 ( ) 3 e

Uform dstrbuto Probablty desty the terval a to b s costat ad dfferet from zero ad equal to zero outsde ths terval Desty probablty fucto for uform dstrbuto: ( ( 3 3 3 ( 0 outsde ths rage Epected value: Varace: a b b a Type B stadard ucertaty s equal to stadard devato a = - b = u( ( 3 3

Type B stadard ucertaty () mechacal devces Rulers, mcrometers, calpers Thermometer, baromether Stopper calbrato ucertaty : half of the scale terval Aaloge devces u( 3

Type B stadard ucertaty () aalogue devces Measuremet rage mamal value to be measured for the set rage. Class of the strumet descrbes the precso of the measuremet devce covertg measured sgal to value preseted o a scale. Class descrbes ucertaty the percetage of the measuremet rage. Calbrato ucertaty: class rage 00 u( 3 Ivestgator ucertaty: e umber rage of scale tervals e u( 3

Type B stadard ucertaty (3) dgtal devces Measuremet ucertaty for dgtal devces: measured value z measuremet rage c cz c, c devce costats e.g. c = 0,%, c = 0,0% u( 3 Avalabe fuctos Measuremet rage

Ucertaty evaluato drect measuremets summary Perform measuremet (sgle or seres) Type A ucertaty Measuremet result mea value Stadard ucertaty stadard devato of the mea value u( s ( ) ( Type B ucertaty Calbrato ucertaty Ivestgator ucrtaty e u( 3 ( 3 Combato of ucertates u( s ( 3 ( ) 3 e

Idrect measuremets Laboratorum Fzyk, Wydzał Fzyk, Poltechka Warszawska

Basc deftos (4) Combed stadard ucertaty u c ( stadard ucertaty of the value calculated based o measuremets of other quattes ucertaty propagato rule Measuremets of correlated quattes Measuremets of ucorrelated quattes I the Physcs Laboratory all measuremets are ucorellated

Ucertaty evaluato drect measuremet summary Measure k quattes drectly (sgle or seres) z f (,,..., k ) Calculate mea value ad stadard ucertaty u( ) for every quatty usg Type A or Type B evaluato method,,..., k u( ), u( ),..., u( k ) Calculate fal value of studed quatty z z f (,,..., k ) Calculate combed ucertaty u c (z) (ucertaty propagato law) u c ( z) j k f ( j j ) u ( j ) Eample for two quattes u c ( z) f (, y) u ( f (, y) y u ( y)

Basc deftos (5) Epaded ucertaty U( or U c ( the measure of ucertaty that defes terval about the measuremet, that may be epected to ecompass a large fracto of the dstrbuto Stadard ucertaty u( defes terval about the measured value, where the true value est wth probablty: 68% for Type A ucertaty 58% for Type B ucertaty Epaded ucertaty: Allows to compare results from dfferet laboratores Allows to compare results wth referece database or theoretcal value Useful for commercal purposes Requred for dustry, health ad securty regulatos

Basc deftos (6) Coverage factor k umber used to multply stadard ucertaty to calculate epaded ucertaty Typcally k vares from to 3. I the most cases the Physcs Laboratory k = should be used. Epaded ucertaty U( defes terval about the measured value, where the true value est wth probablty for k = : 95% for Type A ucertaty 00% for Type B ucertaty (00% also for k=,73!) U( k u(

Reportg measuremet results Laboratorum Fzyk, Wydzał Fzyk, Poltechka Warszawska

Reportg measuremet results () Ucertaty s preseted wth accuracy (ouded) to two sgfcat dgts The measuremet result (the most probable value) s preseted wth a accuracy specfed by the ucertaty, whch meas that the last dgt of the measuremet result ad the measuremets ucertaty must be at the same decmal place. Roudg of ucertates ad measuremet results follows the mathematcal rules of roudg Stadard ucertaty t =,364 s, u(t) = 0,03 s t =,364(3) s, recommeded otato t =,364(0,03) s Epaded ucertaty t =,364 s, U(t) = 0,046 s (k = ) = t = (,364±0,046) s. recommeded otato ot requred

Reportg measuremet results () eamples Measuremet Proper Reportg a = 3,735 m/s; u(a) = 0,4678 m/s a = 3,74 m/s; u(a) = 0,5 m/s a = 3,74(0,5) m/s a = 3,74(5) m/s b = 3785 m; u(b) = 330 m C = 0,0000045 F; u c (C) = 0,00000056 F T = 373,43 K; u(t) =,3456 K b = 3800 m; u(b) = 300 m b = 3800(300) m b = 3,8(,3) 0 3 m b = 3,8(,3) km C=0,00000 F; u c (C)=0,00000056 F C =,00(0,56) 0-6 F C =,00(56) 0-6 F C =,00(56) μf T = 373,4 K; u(t) =,3 K T = 373,4(,3) K U(T) = 4,7 K T = (373,4 ± 4,7) K

Hypothess verfcato Laboratorum Fzyk, Wydzał Fzyk, Poltechka Warszawska

Lear fucto hypothess Graphcal Least squares method Statstcal tests

Graphcal test The most smple Plot theoretcal model fucto. I should cross ucertates bars for more the /3 of epermetal data pots If ot hypothess should be rejected

Least squares method () Goal: to verfy the theoretcal model depedece betwee measured quattes s vald Assumpto: every model ca be coverted to lear type fucto y = a + b Method: least squares to fd the le for whch the sum of squared devatos of epermetal pots from ths le s the smallest to fd le whch s the closest to all epermetal pots Results: a, b ad ucertaty u(a) ad ucertaty u(b) (Type A stadard ucertaty)

Least squares method () ~ a y b y a ~ ~ ~ ~ d y a y ~ ~ ~ s s d s a b a y = a + b

Least squares method(3)

Test () Test fucto Defto Statstcal weght Lear type fucto Sgfcace value probablty of hypothess rejecto Value the rage of to 0 Determed by the vestgator (typcally 0,05) Depeds o umber of freedom (umber of measuremet pots mus umber of calculated parameters) w ( y w ( y y( )) B( ) A) w [ u( y )] Crtcal value χ crtcal (lsted the table for every sgfcace value ad umber of freedom) Test fucto χ χ crtcal there s o argumets to reject hypothess χ > χ kcrtcal hypothess should be rejected