out the law of a falling body, and none, save by chance, ever gets precisely V = gt2; but nobody suggests that these diverse values be accepted as founding a law different from the standard or 'normal' law of fall; instead we say that airresistance and mechanical defects and inaccurate readings have adulterated the data. The physicist is not concerned about the thousand rates of fall that one gets with diverse barometric readings and instruments and observers; rather is he concerned with the limiting value which appears more nearly as conditions are standardized. In like manner, scientific psychometry is primarily concerned with the limiting case about which many empirical curves cluster; and is secondarily concerned with the direction and degree wherein a particular curve departs from this limiting norm. To conclude: we have tried hereabove to show (a) that in quantitative psychology (as opposed to natural phenomena at large) we are very likely to obtain frequencies which can be represented with all needed precision by the normal formula, and (b) that in case the frequencies prove to be anormal, this formula (supplemented with the Pearson-constants) remains equally valuable for revealing the mode or kind of anormality, finding whether it is statistically significant, and thereby apprizing us whether and in what degree our experimental situation is inadequately stabilized and controlled. Quantitative psychology has, as we believe, no single concept more widely useful than the normal law, when correctly understood and applied. REFERENCES 1. ANTHROPOMETRIC COMMITTEE. Report of the British Association for the Advancement of Science, London, 1883. 2. BERTRAND, J. Calcul des probabilités. (2me éd.) Paris, 1907. 3. BORING, E. G. Logic of the normal law of error in mental measurement. Amer. J. of Psychol., 1920, 31, I. 4. BROWN, W., & THOMSON, G. H. Essentials of mental measurement. Cambridge, 1921. 5. CULLER, ELMER. Studies in psychometric theory, I-X. Psychol. Monog., 1926, 35, no. 2, 56-137. 6. CULLER, ELMER. Studies in psychometric theory, XI-XII. This JOURNAL, 1926, 9, 169-194. 7. Fechner, G. T. 8. HILLEGAS, M. B. Elemente der Psychophysik, Band I. Leipzig, 1907. Scale for the measurement of quality in English Composition by young people. Teachers Coll. Rec., 1912, 13, no. 4. 9. KELLER, H. Die Methode der mehrfachen Fälle im Gebiete der Schallempfindungen, usw. Psychol. Stud., 1907, 3, 49. 10. OTIS, A. S. II. PEARL, R. 12. PEARL, R. 13. PEARL, R., & of the egg. Statistical method in educational measurement. Yonkers, 1925. 14. PEARSON, K. Mag., 1900, SURFACE, F. M. Variation and correlation in the physical characters 15. PEARSON, K. Craniological notes. Biometrika, 1903, 2, 339. 17. PEARSON, K. 18. PEARSON, K., 2, 357. On the inheritance of the mental and moral characters in man, etc. Tables for statisticians and biometricians. Cambridge, 1914. 19. RIETZ, H. L. Handbook of mathematical statistics. Boston, 1924. 20. THOMSON, G. H. Criterion of goodness of fit of psychophysical curves. Biometrika, 1919, 12, 216. 21. THORNDIKE, E. L. Handwriting. Teachers Coll. Rec., 1910, 11, no. 2. 22. TOCHER, J. F. Anthropometric characteristics of the inmates of asylums in Scotland. Biometrika, 1907, 5, 298. 23. URBAN, F. M. Über die Methode der mehrfachen Fälle. Arch. f. d. ges. Psychol., 1910, 17, 367. 24. WATSON, J. B. Psychology from the standpoint of a behaviorist. (2d ed.) Philadelphia, 1924. SPECIFIC SERIAL LEARNING; A STUDY OF REMOTE FORWARD ASSOCIATION BY HULSEY CASON University of Rochester PRELIMINARY STATEMENT The theory of remote association assumes that, when a series is learned so that the items can be reproduced in the proper order, certain remote associations are formed between each item and all other items, in both the forward and backward directions. Ebbinghaus claimed that the remote associations are stronger in the forward than in the backward direction, and that the strength of the associations decreases with the number of intervening items.1 We have submitted this problem to an experimental investigation, and the results are summarized in two reports. The first 2 is concerned with the problem of backward association, while the present article describes experiments in which the problem of remote forward association is for the first time submitted to a systematic experimental investigation. The two papers should be considered together, since they are concerned with different aspects of the same problem of remote association.* PART I. TESTS WITH DIFFICULT PROSE MATERIAL If the series 1-2-3-4-5-6-7-8-9-10 is so learned that the items can be reproduced in the proper order, does this learning result in the formation of remote forward associations and facilitate the later learning of the items in the order 1-3-5-7-9, 1 See literature cited on page 195 supra (this JOURNAL, June, 1926). 2 See this JOURNAL, 1926, 9, 195–227. 'I wish to make a grateful acknowledgment to my wife, Eloise Boeker Cason, for much valuable assistance in the preparation of the manuscript, and to the 33 subjects for their coöperation in a very tiresome procedure. 'We are not concerned in the present studies with mediate association. or 2-4-6-8-10, etc.? Where the subject actually practises the items in the order 1-3-5-7-9, while the principal task of learning the series in the order 1-2-3-4-5-6-7-8-9-10 is in progress, any associations formed in the order 1-3-5-7-9 are of course "direct" or "immediate," rather than "indirect" or "remote." The theory of remote association assumes that when associations are formed in the order 1-2-3-4-5-6-7-8-9-10, associations are at the same time established in the order 1-3-5-7-9. In the present experiment the subjects were required to learn difficult prose material, and appropriate tests were given to determine whether remote forward associations had also been formed. Material The prose passage selected is an adaptation from Dewey's "Human nature and conduct" (pp. 106-107). It is quoted in full as follows: SOME WRITERS HAVE MADE THE IDEA OF THE COMPLETE MALLEABILITY OF HUMAN NATURE THE BASIS FOR ASSERTING THE OMNIPOTENCE OF EDUCATION TO SHAPE HUMAN SOCIETY. OTHERS HAVE FOUND IN THESE PLANS FOR SOCIAL CHANGE EVIDENCES OF THE PRONENESS OF YOUTH TO ILLUSION, OR OF INCAPACITY ON THE PART OF THOSE WHO HAVE GROWN OLD TO LEARN ANYTHING FROM EXPERIENCE. CURIOUSLY ENOUGH BOTH PARTIES REST THEIR CASE UPON JUST THE FACTOR WHICH WHEN IT IS ANALYZED WEAKENS THEIR RESPECTIVE CONCLUSIONS. THAT IS TO SAY, THE RADICAL REFORMER RESTS HIS CONTENTION IN BEHALF OF EASY AND RAPID CHANGE UPON THE PSYCHOLOGY OF HABITS, OF INSTITUTIONS IN SHAPING RAW NATURE, AND THE CONSERVATIVE GROUNDS HIS COUNTER-ASSERTION UPON THE PSYCHOLOGY OF INSTINCTS. AS MATTER OF FACT, IT IS PRECISELY CUSTOM WHICH HAS GREATEST INERTIA, WHICH IS LEAST SUSCEPTIBLE OF ALTERATION; WHILE INSTINCTS ARE MOST READILY MODIFIABLE THROUGH USE, MOST SUBJECT TO EDUCATIVE DIRECTION. THE CONSERVATIVE WHO BEGS SCIENTIFIC SUPPORT FROM THE PSYCHOLOGY OF INSTINCTS IS THE VICTIM OF AN OUTGROWN PSYCHOLOGY WHICH DERIVED ITS NOTION OF INSTINCT FROM AN EXAGGERATION OF THE FIXITY AND CERTAINTY OF THE OPERATION OF INSTINCTS AMONG THE LOWER ANIMALS. HE IS A VICTIM OF A POPULAR ZOÖLOGY OF THE BIRD, BEE AND BEAVER. HE IS IGNORANT THAT INSTINCTS IN THE ANIMALS ARE LESS INFALLIBLE AND DEFINITE THAN IS SUPPOSED, AND ALSO THAT THE HUMAN BEING DIFFERS FROM THE LOWER ANIMALS IN PRECISELY THE FACT THAT HIS NATIVE ACTIVITIES LACK THE COMPLEX READY-MADE ORGANIZATION OF THE ANIMALS' ORIGINAL ABILITIES. This passage was selected because it is difficult to learn, and also because, at the time the experiment was carried out, most of the 17 subjects were taking the writer's course in social psychology, in which the question of biological and social heredity had just been discussed. For the method which was used in deriving II forward lists and II separated lists from the prose selection the reader is referred to the similar procedure for backward lists on page 214. The first 6 and the last 3 words of the prose selection were omitted, leaving 242 words. The first forward list was made up of the II alternate words in the F1 (forward) row, and each of the remaining F-lists, F2, F3, etc., was constructed in the same way. The first separated list was made up of the II words in the column above "S1." Each of the F- and S-lists was typewritten on a separate card. Procedure The writer acted as experimenter, and the subjects were taken individually. The procedure with each subject may be divided into 3 parts: (1) Learning the F- and S-lists, (2) Learning the prose passage, and (3) Relearning the F- and S-lists. Comparisons were made between the efficiency of learning these lists, both before and after learning the prose passage. Learning the F- and S-Lists.—After learning 3 practise lists, the subjects learned the 11 F-lists and the II S-lists in the order F1, S1, F2, S2, etc. He was given one list at a time, and was instructed to study it aloud by the whole method until he could make a correct oral recitation. A record was made of the time required to learn each list. The conditions of learning were kept as uniform as possible, and an effort was made to control the situation in such a way that the subjects would not know that the lists were made up of words taken from a logical prose passage and arranged as described above. No subject noticed the method used in constructing the lists. Learning the Prose Passage.-A printed copy of the prose passage was given the subject, and he was instructed to learn it in his own way. After the prose passage had been learned |