Graphic Analysis of Data

Scientific data are important only insofar as they influence the behavior of the researcher and of other people. As Michael (1974, p. 647) points out: The observations [data] resulting from scientific experiments are stimuli that hopefully affect the scientist and his [or her] colleagues by producing better practical behavior, more sophisticated follow-up experiments, or better verbal behavior regarding the subject matter. These stimuli, however, may not result in any effective reaction, a fairly common reason being their complexity. Repeated observation of the same experimental condition, for example, may give rise to a set of numbers, all differing considerably from one another. This situation has occurred quite often and methods have been discovered for simplifying it to some degree. Some of the methods generate two-dimensional visual stimuli where the values of each dimension stand in a point-to-point relation to some feature of the data; a frequency polygon is such a stimulus. Another stimulus-simplifying technique results in a smaller set of numbers, each related to some important characteristic of the larger set, such as the mean and range of the raw data. Using the term “judgment” to refer to any of the various kinds of reactions that a scientist could make to the data of his [or her] experiment, it is useful to refer to these stimulus-simplifying techniques and their products as “judgement aids.”