WORK PLANS_CASE 1 - Laserfiche WebLink

Home Browse Search

95% Non-Parametric(Chebyshev) UCL 0 Because the data do not appear to be normally distributed according to the goodness-of-fit test performed above, the non-parametric UCL (0) may be a more accurate upper confidence limit on the true mean. One-Sample t-Test A one-sample t-test was performed to compare the sample mean to the action level. The null hypothesis used is that the true mean equals or exceeds the action level (AL). The t-test was conducted at the 5% significance level. The sample value t was computed using the following equation: f — x—AL SE where X is the sample mean of the n=13 data, AL is the action level or threshold (10), SE is the standard error= (standard deviation) /(square root of n). This t was then compared with the critical value to 95, where to 95 is the value of the t distribution with n-1=12 degrees of freedom for which the proportion of the distribution to the left of to 95 is 0.95. The null hypothesis will be rejected if t< -to 95 ONE-SAMPLE t-TEST tstatistic Critical Value to.95 Null Hypothesis -1.#INF 1.7823 Cannot Reject The test did not reject the null hypothesis that the mean value at the site exceeds the threshold, therefore conclude the true mean exceeds the threshold. Because the data do not appear to be normally distributed, the MARSSIM Sign Test might be preferred over the One Sample t-Test. The following table represents the results of the MARSSIM Sign Test using the current data: MARSSIM Sign Test Test Statistic(S+) 95%Critical Value Null Hypothesis 13 9 Reject This report was automatically produced`by Visual Sample Plan(VSP)software version 5.000. Software and documentation available at http://dgo.pnl.gov/vsp Software copyright(c)2007 Battelle Memorial Institute. All rights reserved. The report contents may have been modified or reformatted by end-user of software.