Abstract
<p>There are many different data analytical procedures applicable to data obtained via single-case experimental designs, with mean differences, regression-based procedures, and nonoverlap indices being among the most frequently used ones. The possibilities are reduced when the data are ordinal and/or variable and cannot be meaningfully summarized by straight lines. For such scenarios, nonoverlap indices become prominent. The current text deals with the quantification of trend in the context of using nonoverlap indices, commenting some drawbacks of existing options and making a new proposal for assessing monotonic trend. This proposal is related to existing data analytical principles and is illustrated with several examples. The aim of the review of existing options and of the new proposal is to ignite further discussion regarding how to deal with trend without assuming that a straight line (or any specific nonlinear model) is a good representation of the data, while also taking into account the possibility that any trend may level off by reaching an upper or lower bound. A new graphical representation is proposed to accompany the quantification of monotonic trend.</p>