AS 2850-1986

1 - AS 2850-1986 CHEMICAL ANALYSIS - INTERLABORATORY TEST PROGRAMS - FOR DETERMINING PRECISION OF ANALYTICAL METHOD(S) - GUIDE TO
4 - PREFACE
5 - CONTENTS
6 - 1 SCOPE.
6 - 2 APPLICATION.
6 - 3 REFERENCED DOCUMENTS.
6 - 4 DEFINITIONS.
6 - 4.1 Accuracy
6 - 4.2 Bias
6 - 4.3 Cell
6 - 4.4 Critical difference
6 - 4.5 Critical value
6 - 4.6 Outlier
6 - 4.7 Precision
6 - 4.8 Range
6 - 4.9 Repeatability (r)
6 - 4.10 Replicate determinations
6 - 4.11 Reproducibility (R)
6 - 4.12 Single test result
7 - 4.13 Split-levels experiment
7 - 4.14 Straggler
7 - 4.15 Uniform-levels experiment
7 - 4.16 Within-laboratory reproducibility (R)
7 - 5 PRIOR DEVELOPMENT AND VALIDATION OF METHOD
7 - 6 SELECTION OF TEST MATERIALS.
7 - 7 LABORATORY / PERSONNEL REQUIREMENTS.
7 - 7.1 Qualifications and experience of personnel.
7 - 7.2 Laboratory-adherence to test schedule.
7 - 8 SETTING UP A TEST PROGRAM.
7 - 8.1 Choice of experiment.
7 - 8.2 Number of laboratories.
8 - 8.3 Number of concentration levels.
8 - 8.4 Number of replicate determinations.
8 - 8.5 Number of operators per laboratory.
8 - 8.6 Combining information.
8 - 9 EXECUTION OF TEST PROGRAMS.
8 - 9.1 Personnel and responsibilities.
8 - 9.2 Preliminaryconsiderations.
9 - 9.3 Preparation for the test.
9 - 9.4 Test instructions.
9 - 9.5 Reporting of results.
9 - 10 STATISTICAL ANALYSIS OF RESULTS.
9 - 10.1 General.
9 - 10.2 Personnel to analyze results.
9 - 10.3 Cells.
10 - 10.4 Excess data
10 - 10.5 Missing data.
10 - 10.6 Identification of outliers.
10 - 10.7 Calculation of repeatability and reproducibility.
10 - 11 EVALUATION OF PRECISION DATA.
10 - 12 EXPRESSION OF PRECISION DATA.
10 - 12.1 General.
11 - 12.2 Tabular format.
11 - 12.3 Functional relationships.
11 - 12.4 Statement on repeatability, within-laboratory reproducibility and reproducibility
11 - 12.5 Suspect results.
11 - 12.6 Reporting of experiment.
12 - APPENDIX A - STATISTICAL ANALYSIS AS A STEP BY STEP PROCEDURE
15 - APPENDIX B - STATISTICAL ANALYSIS FOR UNIFORM-LEVELS EXPERIMENTS
15 - B1 NOTATION, DEFINITIONS AND BASIC FORMULAS.
16 - B2 COCHRAN’S MAXIMUM VARIANCE TEST.
17 - B3 DIXON’S OUTLIER TEST.
18 - B4 CALCULATION OF REPEATABILITY AND REPRODUCIBILITY.
20 - B5 EXAMPLES.
24 - APPENDIX C - STATISTICAL ANALYSIS FOR SPLIT-LEVELS EXPERIMENTS
24 - C1 NOTATION, DEFINITIONS AND BASIC FORMULAS.
24 - C2 TESTING FOR OUTLIERS.
25 - C3 CALCULATION OF REPEATABILITY AND REPRODUCIBILITY.
25 - C4 EXAMPLE.
27 - APPENDIX D - CRITICAL DIFFERENCES DERIVABLE FROM REPEATABILITY AND REPRODUCIBILITY
27 - D1 CRITICAL DIFFERENCES FOR THE 95 PERCENT PROBABILITY LEVEL.
28 - D2 CRITICAL DIFFERENCES FOR PROBABILITY LEVELS OTHER THAN 95 PERCENT
29 - APPENDIX E - COMPUTER PROGRAM FOR CALCULATING REPEATABILITY AND REPRODUCIBILITY
29 - E1 INTRODUCTION.
40 - APPENDIX F - EXAMPLES OF INSTRUCTION/REPORTING SHEETS FOR ACQUIRING TEST DATA FROM PARTICIPATING LABORATORIES
40 - F1 EXAMPLE 1.
42 - F2 EXAMPLE 2.
43 - F3 EXAMPLE 3.
46 - F4 EXAMPLE 4.
48 - AOOENDIX G - ASSIGNMENT OF STATISTICAL SYMBOLS USED IN THIS STANDARD

Provides guidance in choosing procedures and determining parameters used to select and validate methods of chemical analysis through the determination of the precision of those methods. Precision is determined in terms of repeatability, within-laboratory reproducibility and reproducibility. Defines relevant statistical terms, specifies test material requirements and discusses the number of laboratories required as well as the functions of a 'steering' committee to coordinate activities. Includes computer program to calculate repeatability and reproducibility and four samples of results sheets. Applicable to laboratories reasonably experienced in the methods to be evaluated with respect to the materials being analyzed.

This standard sets out procedures for the planning and execution of interlaboratory test programs, including the evaluation of results, to determine the precision of methods of chemical analysis in terms of repeatability, within-laboratory reproducibility and reproducibility.NOTE: Depending upon the plan selected for the test program (see Clause 8.1), and with due consideration of industrial practice, it may be (more) appropriate, within the context of the statistical treatment of results, that the term 'repeatability (r)' be regarded as being interchangeable with or additional to the term 'within laboratory reproducibility (Rw)' as defined in this standard.