In many scientific studies in the biological and behavioral sciences the response variables fall in the "borderland" between dichotomous classifications and refined measurement systems. Sometimes the response variable is a subjective scale (e.g. "minor", "moderate", "severe") and other times the response variable is numerical but the measurement system is heavily dependent upon details of protocol. These "borderland" response variables may not be adequately analyzed by either of the two traditional families of statistical techniques (i.e. the chi-square and t-test families). In this situation ridit analysis serves as a "missing link" between the two traditional families. In ridit analysis a specified class of individuals is chosen as the "identified distribution" and the other series are considered relative to this identified distribution. The ridit for a given category is simply the proportion of individuals in the lesser categories plus one half of the proportion of individuals in the category itself. Once the transformation has been made the data may be analyzed along the lines of the usual t-test families. The use of the ridit transformation is illustrated on data from a study of automotive crash injuries.