Master of Science (MS)
First Committee Member
Number of Pages
Remediation decisions at contaminated sites are determined by comparing the Upper Confidence Limit (UCL) of the mean contaminant to a site-specific action level. The Environmental Protection Agency (EPA) has established methods for computing this UCL, also called the Exposure Point Concentration (EPC) term; The UCL formula depends upon the distribution of the contaminant concentrations, which are often positively skewed. The EPA frequently recommends the Lognormal distribution to model such data. However, that assumption often causes the UCL to be unreliable. What is the role of skewness in computing UCL statistics over different distributions? Computer simulation is used to test a range of skewness values over three similarly shaped distributions: Lognormal, Gamma, and Weibull. Three statistical tests will also be evaluated: Student's t-Statistic, Chebychev's Inequality, and Land's H-Statistic to compute the UCL. By modeling skewness effectively, we can determine the most appropriate distribution and statistical method to use in computing UCLs.
Comparison; Confidence; Distribution; Limits; Skewed; Upper
University of Nevada, Las Vegas
If you are the rightful copyright holder of this dissertation or thesis and wish to have the full text removed from Digital Scholarship@UNLV, please submit a request to firstname.lastname@example.org and include clear identification of the work, preferably with URL.
Theemling, Michael Gary, "A comparison of upper confidence limits for skewed distributions" (2002). UNLV Retrospective Theses & Dissertations. 1446.