CHAPTER IV
THE UTILITY OF THE AVERAGES OBTAINED
Section 17. Grouping of the Results of the Tests
The first question which awaits an answer from the investigations carried out in the manner described is, as explained in sections 7 and 8, that of the nature of the averages obtained. Are the lengths of time required for memorising series of a certain length, under conditions as nearly identical as possible, grouped in such a way that we may be justified in considering their average values as measures in the sense of physical science, or are they not?
If the tests are made in the way described above, namely, so that several series are always memorised in immediate succession, such a type of grouping of the time records could scarcely be expected. For, as the time devoted to learning at a given sitting becomes extended, certain variable conditions in the separate series come into play, the fluctuations of which we could not very well expect, from what we know of their nature, to be distributed symmetrically around a mean value. Accordingly the grouping of the results must be an asymmetrical one and cannot correspond to the “law of error.” Such conditions are the fluctuations of attention and the decreasing mental freshness, which, at first very quickly and then more and more slowly, gives way to a certain mental fatigue. There are no limits, so to speak, to the slowing down of the learning processes caused by unusual distractions; as a result of these the time for learning a series may occasionally be increased to double that of its average value or more. The opposite effect, that of an unusual exertion, cannot in the very nature of the case overstep a certain limit. It can never reduce the learning time to zero.
If, however, groups of series equal in number and learned in immediate succession are taken, these disturbing influences
