Statistical Analysis

Statistical Analysis 1. Linear Calibration

The purposes of this procedure are to demonstrate how to correctly perform the technique of linear calibration, how to develop appropriate confidence intervals, how to reduce the interval estimates for more useful estimations, and the proper and improper uses of linear calibration.

View Method

Statistical Analysis 2. Limits of Detection and Determination

The purpose of this procedure is to define measures of low-level detection of analytes and to demonstrate their use. Two different definitions are given, detection limit and determination limit, each of which has a specific use.

View Method

Statistical Analysis 3. Ruggedness Testing

This procedure should be used by authors of assay methods to help them sufficiently specify such methods so that others may use them and obtain comparable results. A method developed in an author’s laboratory often may work well and appear to be reliable. Workers in other laboratories, however, may experience difficulties in using the method. In most instances, the author has assumed that certain laboratory practices would be followed that are not common to another worker’s laboratory. By specifying all relevant details, a method is made more rigorous. The procedure presented here is intended to help authors of methods to determine those sensitive features that may affect precision before collaborative study. Knowing sensitive method factors in advance greatly enhances the utility of an assay method.

View Method

Statistical Analysis 4. Youden Unit Block collaborative Testing Procedure

This procedure was developed to aid experimenters in collaborative studies. It is based on the Youden Unit Block method and reports precision using commonly accepted terminology. The procedure is intended to be self-contained; the experimenter need not consult statisticians for designing the study, computing the statistics, or analyzing the results. The procedure is not intended to be definitive, however. Many other possible designs or statistical approaches could be used. The design used here was chosen for its relative simplicity and practicality.

Calculator: Youden Unit Block Workbook (Excel)

View Method

Statistical Analysis 5. Comparison of Test Methods

This procedure was designed to aid experimenters conducting collaborative studies using two or more different analytical test methods. It evaluates test methods by determining whether significant differences in means and precisions exist. A comparison of methods within a single laboratory is also included. The procedure uses paired t tests for comparing means. F tests are used to compare precisions. The procedure was chosen for its relative simplicity and flexibility, making use of data that was collected according to a Youden unit block design (Statistical Analysis-4) for collaborative studies. Many other statistical approaches are available depending upon the experimental design.

View Method

Statistical Analysis 6. Comparison of Microbiological Data

Microbiological experiments often generate data that cannot be analyzed using traditional parametric statistical techniques, such as t tests and analysis of variance. Such data are often categorical, meaning that the data are grouped into categories such as growth, no growth, too numerous to count (TNTC), etc. The method discusses the contingency table approach, which can assist in the statistical analysis of such categorical data, and the Kruskal-Wallis signed rank test, a nonparametric one-way analysis of variance test used to determine whether two or more groups are different.

View Method

Appendix I. t-Distribution (Two-tailed Tests)

t Distribution (two-tailed tests)

View Appendix

Appendix II. F-Distribution, = 0.05

F Distribution, ( = 0.05

View Appendix

Appendix III. F-Distribution, = 0.025

F Distribution, ( = 0.025

View Appendix

Appendix IV. F-Distribution, = 0.01

F Distribution, ( = 0.01

View Appendix

Appendix V. F-Distribution, = 0.01

F Distribution, ( = 0.005

View Appendix

Appendix VI. F-Distribution, = 0.001

F Distribution, ( = 0.001

View Appendix

Appendix VII. Cumulative Distribution of Chi-Square

Cumulative distribution of chi square.

View Appendix

Appendix VIII. Definition of Statistical Terms

Glossary of statistical terms

View Appendix

I. Instrument Standardization

View Appendix

II. Nomenclature of Hop Resin Fractions and Components

View Appendix

III. Water Analysis

View Appendix