摘要
根据大量土力学试验的经验，并参考统计学理论，提出了采用分位数法辨别对试验结果可靠程度负面影响的土样因素，以及减少土样因素的负面影响，提高试验结果可靠性的方法。实践证明，这种方法不但是可靠的，而且也便于应用。
A new method called cut - off value method was presented. With the new method, reliability of the result from laboratory test on soil can be increase significantly. The new method was proved to be reliable and useful in analyses on data from soil tests.
