WORK PLANS - Laserfiche WebLink

Home Browse Search

also offers some useful graphical displays including histograms, multiple quantile-quantile (Q-Q) plots, and side-by-side box plots for data sets with and without ND observations. The graphical displays provide additional insight and information contained in data sets that cannot be revealed by the use of estimates (e.g., 95%UCLs) and test statistics such as goodness-of-fit(GOF)test statistics, t-test statistic, Rosner test, and various other statistics. In addition to providing information about the data distributions(e.g.,normal, lognormal, gamma),the graphical Q-Q plots are very useful to identify potential outliers and the presence of mixture samples (if any) in a data set. Side-by-side box plots and multiple Q-Q plots are quite useful to visually compare two or more data sets such as site versus background contaminant concentrations, monitoring well (MW) concentrations, and so on. Therefore, it is desirable and suggested that the conclusions derived using estimates (e.g., 95%UCL) and test statistics (e.g.,t-test) should always be supplemented with graphical displays. ProUCL 4.0 serves as a companion software package for the UCL Computation Guidance Document for Hazardous Waste Sites (EPA,2002a) and the Background Guidance Document (currently under revision) for CERCLA Sites (EPA, 2002b). Most of the statistical and graphical methods described and recommended in these two EPA guidance documents have been incorporated in ProUCL 4.0. It should be noted that ProUCL 4.0 also has some parametric and nonparametric single sample hypotheses approaches that may be used to compare site mean concentrations (or some site threshold value such as an upper percentile)with some average cleanup standards, C,(with a not-to-exceed limit, AO)to verify the attainment of cleanup levels (EPA, 1989, and EPA, 2006) after some remediation activities have been performed at potentially impacted site areas. Several of the statistical methods as incorporated in ProUCL 4.0 can be used in groundwater(GW) monitoring applications(EPA, 1992). Two reference guides: 1) ProUCL 4.0 User Guide and 2)ProUCL 4.0 Technical Guide have also been developed for ProUCL 4.0 software package. The User Guide describes and illustrates the uses of the various menu items and options as incorporated in ProUCL 4.0. The ProUCL 4.0 Technical Guide describes the theory (with references)behind the statistical methods as incorporated in ProUCL 4.0. These two documents can be downloaded from the EPA website fon ProUCL 4.0. ProUCL 4.0 also provides Online Help for the various methods available in ProUCL 4.0. Data Requirements Statistical methods (e.g., upper limits) as incorporated in ProUCL 4.0 (and also in other software packages such as SAS and Minitab) assume that the user has collected an adequate amount of data of good quality, perhaps using appropriate data quality objectives (DQOs) as described in EPA, 2006. However,many times (e.g.,using the available historical data, or due to budgetary and time constraints), it may not be possible to collect data sets based upon specified performance measures (e.g., decision errors) and other DQOs. It is noted that many times, administrators and decision makers do not want to collect many samples, especially background samples. Therefore,when it may not be possible to collect adequate amount of data using DQOs (EPA, 2006), Chapter 1 of the two ProUCL 4.0 reference guides can be used to determine the minimum sample size requirements associated with the various estimation and hypotheses testing approaches available in ProUCL 4.0. The suggested minimum sample size requirements as described in Chapter 1 are made based upon the practical applicability of the procedures incorporated in ProUCL 4.0. Those suggestions are particularly useful when the data are sparse and it may not be feasible to collect additional data based upon DQOs. However, it should be