详情

原文传递 Reivew of F and t Test Results 2000 Through 20034.

题名：	Reivew of F and t Test Results 2000 Through 20034.
作者：	CHAVEZ, E.
关键词：	COMPARATIVE EVALUATIONS
摘要：	The Colorado Department of Transportation (CDOT) uses the F test and t test to determine if two sets of test data are from the same population. The comparison is usually between the Contractor's quality control tests and the Department's verification tests. The F test and t test are calculated in accordance with standard statistical procedures to make the comparisons. The F test analyzes the differences in the standard deviations of the data sets. The t test analyzes differences in the means of the data sets. Starting in 1988 the first projects were awarded under pilot specifications which contained the requirement for F and t test analysis. Analysis was conducted between the Contractor's and the Department's flexural strength tests on these first projects. The specification was released as a standard specification on February 11, 2000 titled: Revision of Sections 105, 106, and 412, Quality of Portland Cement Concrete Pavement (Alternative Strength Criteria). This report analyzes the F and t test results from 2000 through 2004. Eight of the projects evaluated in this report contain the alternative strength specification. Two projects included in this report contained modified specifications: ten year warranty PCCP & ten year warranty HMA. Analysis on ten projects is included in this report. On two of the projects the F and t test analysis was designated as being for informational purposes only by the engineer. Irregardless of the outcome of the F and t tests, the Contractor's tests results were used on one of these projects. CDOT's test results were used on the other. On both of these projects the F and t test passed on all processes. Of the ten projects evaluated five passed both tests on all processes. Three projects had at least one failing test in the t test. Two projects failed both tests on all processes. Twenty-two processes were established for the material being evaluated. The F tests passed in all but two of the processes. The variation of the data between the Contractor and the Department was within allowable differences 91%of the time. The t test failed on 7 of the processes, 32%of the time. There was a consistent difference between the two sets of test results on these processes. CDOT's test results were used instead of the Contractor's for the calculation of I/DPs on four processes because of failing F and t test results.
报告类型：	科技报告