Training in cancelling parts of speech does not improve ability to cancel words having the letters e and t. Training in memorizing "Paradise Lost" does not improve the memory for Hugo's verse. Training in memorizing nonsense syllables does not increase the memory-span for letters, numbers, nonsense syllables, disconnected words, Latin-English vocabularies, poetry, and prose; nor the ability to memorize completely meaningless visual characters, Latin-English vocabularies, and passages of poetry and of prose. Training in memorizing prose substance does not improve the ability to memorize dates, nonsense syllables, poetry, points on a map, dictation, letters, and names. Training in memorizing tables does not improve ability to memorize dates, poetry, prose, prose substance, dictation, letters, and names. Memorizing poetry does not improve ability to memorize dates, poetry of another sort, prose, prose substance, points on a map, dictation, letters, and names.

I have also examined the results of experiments on the transfer of training by Bagley and Squire, Briggs, Burnet, Coover, Dallenbach, W. F. Dearborn, Foster, Hewins, Judd, Ruger, Scholckow and Judd, Wallin, Whipple, and Winch, but have not been able to make them the basis of a theoretical discussion because they were too indefinite, irregular, or complicated to bring within a consistent rule. However, no theory that is at all specific can explain all cases of reported transfer. The psychological factors in the cases of positive transfer described above were pointed out. In the cases of negative or zero transfer it is difficult to find common sensory stimuli, and common conceptual stimuli that may exist are too general to be effective. It is also difficult to find common conceptual or associative responses. But some cases of positive transfer are equally baffling; for example, training in memorizing poetry has been found to improve the ability to locate points on a map and to memorize nonsense syllables, but interfered with the ability to learn poetry of another sort, prose, or prose substance. Memorizing tables improved the ability to locate points in a circle and to learn nonsense syllables, but interfered with the ability to learn dates, poetry, or prose. The difficulty with all these studies is that the associative processes were not investigated. If we knew what the common bonds of association had been in these cases of positive and negative transfer, we probably should have the cue to their explanation. The correlation between observable stimuli and observable responses is too irregular to make a consistent principle inferable with certainty, but such regular

correlations as there are point quite definitely to the solution of the problem in the laws of association. It is to be hoped that future investigators of this problem will more carefully examine the internal facts of transfer, i. e., the common associative bonds (pp. 281-283).

15. Transfer in School Subjects

[RUGG, H. O., The Experimental Determination of Mental Discipline in School Studies, pp. 114-115. Baltimore, Warwick and York, 1916.]

One of the most notable experiments made with school subjects to determine the effects of practice in one subject upon efficiency in other subjects was performed by Rugg. He made a study of the influence of training in descriptive geometry. He used a great variety of test material, some of it arithmetical, some geometrical. The experiments were given to 326 students in the University of Illinois, College of Engineering. As a control group he had 78 students in other colleges take the end tests but not the practice work. His conclusions follow:

The study of descriptive geometry (under ordinary classroom conditions throughout a semester of fifteen weeks) in which such natural and not undue consideration is given to practice in geometrical visualization as is necessary for the solution of descriptive geometry problems operates:

1. Substantially to increase the students' ability in solving problems requiring the mental manipulation of a geometrical nature, the content of which is distinctly different from the visual content of descriptive geometry itself.

2. Substantially to increase the students' ability in solving problems requiring the mental manipulation of spatial elements of a slightly geometrical character, i. e., problems utilizing the fundamental elements of geometry (the point, line, and plane), but apart from a geometrical setting and in such form as to offer no geometrical aids in solution.

3. Substantially to increase the students' ability in solving problems requiring the mental manipulation of spatial elements of a completely non-geometrical nature, i. e., problems in which the straight line and plane do not appear in any way whatsoever.

4. The training effect of such study in descriptive geometry operates more efficiently in those problems whose visual content more closely resembles that of the training course itself, i. e., in those

problems whose imagery content is composed of combinations of points, lines, and planes, and in which the continuity of the manipulating movements approaches the continuity of those in the training course.

16. Experimental Evidence of Transfer [THORNDIKE, E. L., Educational Psychology, Vol. II, pp. 397-398. New York, Teachers College, Columbia University, 1913.]

In 1901, Thorndike and Woodworth reported experiments made to determine the effects of practice in judging areas, lengths of lines, weights, and recognizing certain words (words containing certain letters), upon ability to do the same with other areas, lengths of lines, weights, and words (words containing other letters). Their results indicate transfer. There was some transfer which resulted in improvements; there was some negative transfer of an unde sirable sort; and there were results that indicated neither help nor hindrance, i. e., no transfer at all.

Individuals practiced estimating the areas of rectangles from 10 to 100 sq. cm. in size until a very marked improvement was attained. The improvement in accuracy for areas of the same size, but of different shape, due to this training was only 44 per cent as great as that for areas of the same shape and size. For areas of the same shape, but from 140-300 sq. cm. in size, the improvement was 30 per cent as great. For areas of different shape and from 140-400 sq. cm. in size, the improvement was 52 per cent as great.

Training in estimating weights of from 40-120 grams resulted in only 39 per cent as much improvement in estimating weights from 120 to 1,800 grams. Training in estimating lines from .5 to 1.5 inches long (resulting in a reduction of error to 25 per cent of the initial amount) resulted in no improvement in the estimation of lines 6-12 inches long.

Training in perceiving words containing e and s gave a certain amount of improvement in speed and accuracy in that special ability. In the ability to perceive words containing i and t, s and p, c and a, e and r, a and n, e and o, misspelled words and A's, there was an improvement in speed of only 39 per cent as much as in the ability specially trained, and in accuracy of only 25 per cent as much. Training in perceiving English verbs gave a reduction in time of nearly 21 per cent and of omissions of 70 per cent. The ability to perceive other parts of speech

showed a reduction in time of 3 per cent, but an increase in omissions of over 100 per cent. . .

These experiments showed very clearly the influence of: The acquisition during the special training (1) of ideas of method and habits of procedure, and also (2) of facility with certain elements that appeared in many other complexes. Instances of (1) are learning in the 10 to 100 sq. cm. training series, that one has a tendency to overestimate all areas and consciously making a discount for this tendency, no matter what the size or shape of the surface may be; learning to look especially for the less common letter (e. g., s in the case of e-s words, p in the case of s-p words) in the training series and adopting the habit for all similar work; learning to estimate areas in comparison with a mental standard rather than the objective 1 sq. cm., 25 sq. cm., and 100 sq. cm. squares which each experimenter had before him (after one gets mental standards of the areas he judges more accurately if he pays no attention whatever to the objective standards). An instance of (2) is the increase of speed in all the perception tests through training in one, an increase often gained at the expense of accuracy.

In the opinion of the authors these experiments show that:

Improvement in any single mental function need not improve the ability in functions commonly called by the same name. It may injure it.

Improvement in any single mental function rarely brings about equal improvement in any other function, no matter how similar, for the working of every mental function-group is conditioned by the nature of the data in each particular case.

The very slight amount of variation in the nature of the data necessary to affect the efficiency of a function-group makes it fair to infer that no change in data, however slight, is without effect on the function. The loss in the efficiency of a function trained with certain data, as we pass to data more and more unlike the first, makes it fair to infer that there is always a point where the loss is complete, a point beyond which the influence of the training has not extended. The rapidity of this loss, that is, its amount in the case of data very similar to the data on which the function was trained, makes it fair to infer that this point is nearer than has been supposed.

The general consideration of the cases of retention or of loss of practice effect seems to make it likely that spread of practice

occurs only where identical elements are concerned in the influencing and influenced function.

A good statement of how Thorndike thinks transfer takes place is set forth in Educational Psychology, Vol. II, "The Psychology of Learning," pp. 358-359.

The answer which I shall try to defend is that a change in one function alters any other only in so far as the two functions have as factors identical elements. The change in the second function is in amount that due to the change in the elements common to it and the first. The change is simply the necessary result upon the second function of the alteration of those of its factors which were elements of the first function, and so were altered by its training. To take a concrete example, improvement in addition will alter one's ability in multiplication because addition is absolutely identical with a part of multiplication and because certain other processes,-e. g., eye movements and the inhibition of all save arithmetical impulses,-are in part common to the two functions.

Chief among such identical elements of practical importance in education are associations including ideas about aims and ideas of method and general principles, and associations involving elementary facts of experience such as length, color, number, which are repeated again and again in differing combinations.

By identical elements are meant mental processes which have the same cell action in the brain as their physical correlate. It is of course often not possible to tell just what features of two mental abilities are thus identical. But, as we shall see, there is rarely much trouble in reaching an approximate decision in those cases where training is of practical importance.

17. Mental Discipline in High School Subjects [THORNDIKE, E. L., "Mental Discipline in High School Studies," The Journal of Educational Psychology, 1924, Vol. 15, pp. 1-22, 8398.] (After Thomson.)

One of the best experimental studies of transfer of training from the point of view of extent and natural background is that conducted by Thorndike in 1922-1923. Over eight thousand children acted as subjects, and what is more, they had no knowledge of the fact that they were subjects. These children were practiced in their actual school work. The