WORK PLANS - Laserfiche WebLink

Home Browse Search

pointed out that for more accurate (reduced bias) estimates and reliable (increased precision) results,whenever possible, it is desirable to collect adequate amount of data,perhaps using DQOs with specified performance measures. A partial listing of the statistical and graphical methods as incorporated in ProUCL 4.0 is given as follows. The details of the various statistical and graphical procedures with illustrating examples can be found in the User Guide and the Technical Guide associated with ProUCL 4.0. ProUCL Version 4.0 Capabilities All of the capabilities of ProUCL 3.0 have been retained in ProUCL 4.0. It is anticipated that ProUCL 4.0 will serve as a companion software package for: 1) UCL Computation Guidance Document for Hazardous Waste Sites (EPA, 2002a), and 2) Background Guidance Document (currently under revision) for CERCLA Sites(EPA, 2002b). Several statistical and graphical methods for data sets with and without ND observations have been incorporated in the upgraded ProUCL 4.0 software package. Some of those capabilities are listed in the following paragraphs. Group Option ProUCL 4.0 provides a"Group" option. An appropriate Group-ID variable representing the various groups such as different site areas of concern(AOC) or monitoring wells (MWs) should be available in the data sheet. Using this option, graphical displays and statistical analyses can be performed separately for each of the group represented by the Group-ID variable. This group graph option is very useful to perform visual multiple comparison(multiple Q-Q plots, side-by- side box plots) of the various groups (e.g., AOCs,MWs) identified by the Group-ID variable. The details of this option are given in ProUCL 4.0 User Guide. Graphical Methods ProUCL 4.0 has several graphical methods including multiple quantile-quantile(Q-Q)plots, side-by-side box plots, and histograms. These graphical methods can be used on data sets with and without nondetect observations. A typical Q-Q plot (normal, gamma, lognormal)is often used to visually assess the data distribution of the COPCs. A Q-Q plot also provides important information about presence of potential outliers and multiple populations that may be contained in a data set. For data sets with NDs, ProUCL 4.0 can be used to generate Q-Q plots based upon regression on order statistics (ROS) methods including the robust ROS method. The graphical displays of multiple Q-Q plots and side-by-side box plots are useful to visually compare the concentrations of two or more populations, some of which are listed as follows: • Site versus background populations (areas) • Surface versus subsurface concentrations • Concentrations of two or more AOCs or MWs Goodness-of-Fit(GOF)Test Methods ProUCL 4.0 has GOF tests for normal, lognormal, and gamma distributions for data sets with and without nondetect observations. The following GOF tests to assess normality or lognormality of a data set are available in ProUCL 4.0.