Abstract
In this article, a new test for increasing or decreasing trend in a set of observations is presented. Rather than making specific paired comparisons, as in the Cox and Stuart tests for trend, the user of the new test makes comparisons of all possible pairs of observations. The advantage of this procedure is that, for a given number of observations, many paired-comparisons are made. This allows the researcher to examine increasing or decreasing trend when there is quite a small number of observations although, admittedly, the power of the test is low in such cases. A normal approximation is available for use with large samples.
Get full access to this article
View all access options for this article.
