WORK PLANS - Laserfiche WebLink

Home Browse Search

Summary Statistics • For full data sets without NDs,ProUCL computes and lists all relevant descriptive summary statistics for raw and log-transformed data. • For data sets with NDs, ProUCL computes simple summary statistics using only detected data values for raw or log-transformed data. Note: Summary statistics option does not compute and lists the estimates of the population parameters. Those estimates are computed and listed by the `UCL' and `Background' options of ProUCL 4.0. Estimates of Population Parameters • Computes the maximum likelihood estimates(MLEs) and minimum variance unbiased estimates (MVUEs) of the various population parameters such as the mean, standard deviation, quantiles, coefficient of variation(CV), skewness, and also the MLEs of the shape parameter k and scale parameter 0 of a gamma distribution. These estimates (e.g., MLE, MVUE) are shown when the menu items Background and UCL are used to compute the upper limits. • For data sets with NDs, ProUCL 4.0 also computes parametric (e.g., normal MLE) and nonparametric (Kaplan Meier(KM), Bootstrap) estimates of population mean, variance, and standard error of the mean. These statistics do not represent simple summary statistics. Therefore, these estimates (e.g,MLE,KM) are shown when the menu items Background and UCL are used to compute the upper limits. Upper Confidence Limits (UCLs)to Estimate Exposure Point Concentration Terms A 95%UCL of the unknown population arithmetic mean,pi, of a COPC is used to estimate the EPC term and also to determine the attainment of cleanup standards. It should be noted that gamma distribution is often better suited to model positively skewed environmental data sets than the lognormal distribution. For positively skewed data sets,the default use of a lognormal distribution often results in impractically large UCLs, especially when the data sets are small (Singh, Singh, and Iaci, 2002). In order to obtain accurate and stable UCLs of practical merit, other distributions such as a gamma distribution should be used to model positively skewed data sets. ProUCL, Version 4.0 has procedures to perform the gamma goodness-of-fit tests and to compute UCLs of the population mean, and various other limits based upon gamma distributed data sets with and without nondetect observations. ProUCL 4.0 also has several bootstrap methods (e.g., percentile bootstrap, bias corrected bootstrap,bootstrap-t)to compute UCLs of the mean for data sets with and without ND observations. For full data sets without NDs and for left-censored data sets with ND observations, ProUCL 4.0 can compute several parametric and nonparametric UCLs with a confidence coefficient(CC) specified in the interval [0.5, 1.0) including the commonly used CC level 0.95. ProUCL 4.0 can compute parametric UCLs for normal, lognormal, and gamma distributions. It is noted that in environmental applications (e.g., estimation of EPC), a 95%UCL of mean is used,therefore, ProUCL makes recommendations only for an appropriate 95%UCL (s) which may be used to