Classics in the History of Psychology

An internet resource developed by
Christopher D. Green
York University, Toronto, Ontario
ISSN 1492-3713

(Return to index)

The Mental Traits of Sex

Helen Bradford Thompson (1903)
Posted July 2000

Classics Editor's note: The numbers appearing in parentheses throughout the text refer to the items
in the bibliography at the end of the volume.

[p. 93] CHAPTER VII.


THE investigation of the intellectual faculties covered four different fields:

A. Memory.
B. Association.
C. Ingenuity.
D. General information.

The memory factors investigated were the rate of memorizing, retentiveness, and the nature of the imagery employed. The material to be memorized consisted of two series of nonsense syllables, each series containing ten syllables. The first series ("mon, yit, zev, yer, zam, kig, sef, gav, cim, nis") was read aloud, and the second ("huc, cir, suv, nif, fom, mep, yom, fim, zok, seb") presented visually.

In both cases the syllables were given at the rate of one per second. To avoid rhythm and secure a constant rate of presentation for all subjects, they were timed by a metronome. The auditory series was read aloud as distinctly as possible, the visual series was placed on a stroboscopic drum and displayed syllable by syllable behind an opening in a black screen. The auditory series was learned first in each case, and the visual series immediately afterward at the same sitting.

The subject was given no directions about the way [p. 94] he should do the memorizing. He was merely told that the series of syllables would be given him as many times as he needed it to learn it correctly. A pause of from twelve to fifteen seconds intervened between successive presentations. As soon as the subject thought he knew the series, he began to say it aloud in the interval between presentations. If it were not correctly; said, the readings continued until it was. After it was learned the subject was to describe the type of imagery he had used in memorizing, to tell whether or not he had learned the series by means of associations, and to report any tendency to group the syllables in learning them.

Retentiveness was measured by the number of repetitions required to memorize the same series just one week after the first memorizing. The subject was not told that he would be required to memorize them a second time; the hours were merely arranged so that he came to the laboratory just a week after the first memorizing. In almost all cases the subjects said that they had made no effort to repeat the syllables since the first memorizing. A few had tried it simply out of curiosity, and one of the men suspected that he would be asked to memorize them a second time and had made a special effort to remember them. There were two women and two men who failed to appear at [p. 95] the hour when the second memorizing should have taken place. In these instances the interval was from one to three days over a week. The abnormally slow second learnings do not, however, coincide with these longer periods. According to the well-known psychological law of forgetting, there is very little difference, so far as memory is concerned, between a period of seven days and one of nine or ten. Aside from this consideration, however, the exceeding of the regular period by these four subjects may be disregarded in comparing the men's retentiveness with the women's, because two of them were men and two women.

Figs. 49 and 50 are the curves for the first memorizing of the two series. The auditory series has a total average considerably greater than that of the visual series. No subject learned the auditory series in less than eight repetitions, while a number learned the visual in six. Not more than thirty-five repetitions were required in any case for the visual series, while the auditory series has fifty-five as its upper limit. There are two factors which are adequate to explain this difference. The first, and by far the most important, is that the habit of memorizing by means of printed symbols is universal, while very few, if any, of the subjects had ever formed the habit of learning material that is read aloud. Moreover, the visual symbols are more easily grasped in the first instance [p. 96] than the auditory. The second factor which might tend to shorten the time of the visual series is that it was learned second in each case, and therefore had the benefit of the practice obtained in learning the auditory series. Since nonsense syllables were completely unfamiliar material for memorizing, the first series presented the additional difficulty of getting accustomed to a new subject-matter.

In both the auditory and the visual series the women show distinctly greater power of memorizing than the men. Here, again, we find both men and women at both extremes, but there are in both cases more women than men who memorized quickly.

Retentiveness was found to be practically the same for both the men and the women (Figs. 51 and 52). In the auditory series the curves for retentiveness correspond very closely. In the visual the women are slightly better, but the difference is too small to be of any significance.

Since the women are thus shown to have a greater power to memorize nonsense syllables than the men and an equal power to retain the memory, they may be said to have, on the whole, better memories for such syllables than have men. The question arises whether this fact justifies a general statement that women are superior to men in the faculty of memory.[p. 97] If what we wish to measure is mere power of memory, isolated as far as possible from the factors of reason and association, nonsense syllables are universally conceded to be the best material. The results would justify us, therefore, in the statement that memory in its purest form is better among women than among men.

A record was also made of the type of imagery used by each person in memorizing the syllables. The difficulty of making exact and adequate observations of imagery is great even for individuals with special training, and is still greater for the unpracticed. The results in the present case were made as trustworthy as possible by questioning each subject at once and carefully as to the exact nature of his mental procedure. The great variety of combinations of imagery used makes their tabulation in significant form difficult. The scheme which has been followed in the present case is to classify all the cases in which imagery of the same sort predominates together. For instance, all the cases in which auditory imagery is predominant are put together. Among these are some in which only auditory imagery was used, and some in which visual, or motor, or both, were secondary to the auditory. Where two or three types of imagery were used so equally that no one could be called predominant, [p. 98] the case is classified according to the components, as auditory-motor, visual-motor, etc. It is of course impossible to assert that the type of imagery used in this particular piece of memorizing is characteristic for the thinking of an individual, but habit must play a large part in determining the imagery to be used in dealing with novel material, such as nonsense syllables.

Table XII gives the classification of the types of imagery. It shows that there were more men than women in whom auditory imagery was predominant, while women were more numerous in the visual and motor types. This statement holds for both the auditory and the visual series. The same tendency is further illustrated by the fact that when the series is given auditorially there are far more men in the auditory division than in the visual. When the series is given visually the relation is not reversed, but the visual presentation tends to induce visual imagery, with the [p. 99] result that the men are about equally divided between the two classes. The same holds true with reference to visual imagery. In visual presentation there are far more women in the visual categories than in the auditory, while in auditory presentation they are more equally divided between the two.

No differences were discoverable in the use of associations in memorizing or in the formation of groups. The great majority of the subjects, both men and women, learned the series in groups. In most cases the first and last groups were learned before the middle. The syllables from the sixth to the eighth were particularly hard. About half of the subjects, both men and women, used associations of some sort in memorizing. These were of the most various kinds. Some were associations of a few syllables with words; a few subjects associated most of the syllables with words; some associated each syllable with a finger tip; and one subject was obliged to connect each syllable with a spot on the wall before he could remember them. In a few cases the first letters were associated, or the vowels of successive syllables. But none of these schemes were characteristic of either sex.

The results of the memory tests may be summarized as follows: Judged by the approved memory test, i. e., the memorizing of nonsense syllables, the women memorize more rapidly than the men. There is no difference between the men and the women in retentiveness. In memorizing nonsense syllables visual imagery is more common among the women and auditory among the men. No difference between the sexes was discovered in the use of associations in memorizing, or in the habit of memorizing in groups.[p. 100]


Experimental research in the field of association has been directed to two different problems: first, the nature of the relation between the idea furnished and the idea called up, or the quality of the association; and second, the rapidity with which the associated idea follows the given idea, which is a quantitative measure of the association process. Only the second of these two problems, that of the rapidity of the association process, was investigated in this series of tests. An attempt was made to deal also with the qualitative aspect of association, but it was abandoned for two reasons: first, because the classification of the associations, when obtained, was so difficult; and second, because it was so hard to obtain satisfactory results from untrained subjects. Until there is more unanimity in the psychological world about the best classification of associations, and the evaluation of the classified results, it seems useless to employ the test for a comparative study. Moreover, the test is a difficult one to apply. Many subjects when asked to write down the first word or phrase which comes in association with a given word find themselves quite at a loss. They insist that what comes naturally is not some other word or phrase, but first the image of the word itself, and second, some scene or train of ideas which it. starts. If a single word or phrase is required the process seems entirely forced and bizarre to them.

The quantitative test, on the other hand, proved to be much easier to make, and its results are easily formulated, though perhaps not easily interpreted. The method employed was that of requiring the subjects to write down as rapidly as possible for a fixed time [p. 101] the train of ideas started by a given word. In order to make the process as natural as possible no restrictions whatever were made upon the field of association. The subject was not required to come back to the word given for each fresh association. He was told to be just as natural as possible; to let his thoughts take their course, whether that consisted in clinging to the given word or in wandering away from it; to attempt to catch the ideas as they passed, and indicate them on paper by a word or phrase as rapidly as possible while making the chain of associations clear to the reader. The time allowed for each word was one minute and thirty seconds. The subject was allowed in each case to finish the word or phrase on which he was engaged at the end of the time.

In order that the results might be comparable for any two subjects it was of great importance to have words which would have approximately the same suggestiveness for both. Since the university life was the only field of experience which was sure to be common to all the subjects, a list of words connected with distinctively university institutions was selected. They were the following:

1. Registration.                    6. Faculty.
2. Convocation.                   7. Gymnasium.
3. Library.                           8. Football.
4. Flunk.                              9. Dean.
5. Matriculation.                 10. Degree.
The counting of the number of associations after they were written was by no means a simple task. They were frequently written down in detached words or phrases which afforded a convenient indication of what was to be regarded as a distinct association.[p. 102] But in many cases they were written in long compound phrases or sentences, and it was often difficult to decide whether a given phrase was to be regarded as one association or two. For instance, consider the phrase, "Groups of students on the campus." Is that to be regarded as one association or two? If it were written, "Groups of students -- campus," it would be evidently two associations, but what is to be done with the other form? Is the difference merely one of the method of recording, or does it represent a real difference in the association process? If the latter, then "Groups of students on the campus" would have to be regarded as one definite idea, while "Groups of Students -- campus," would represent one idea calling up another by a fresh association. But even if this latter interpretation were psychologically the more accurate, would it be fair to count the first example as one association and the second as two? The time factor would then play a very disturbing role. Those subjects who [p. 103] took time to write out their associations very definitely would appear in the result as having an extremely small number of ideas within the given time. It seemed fairer, therefore, to count each clearly analyzable idea as a separate association, even in cases where its correct psychological interpretation was not that of a distinct association, but rather that of one partly organized factor in an association. In the example cited, both records would be counted as two associations. The possibility for errors and inconsistencies in counting the associations is undeniable. It was avoided as far as possible by deciding on the above criterion for separating the associations, and by going over the records several times to make sure that similar cases were always counted alike.

The ten curves, one for each word, showing the results of the tests, appear in Figs. 53-62. In words 5, 6, 9, and 10 the women have a distinctly greater number of associations than the men. The curve for their results is seen to be, on the whole,[p. 104] above that of the men in the higher ranges, i. e., from ten or twelve to twenty associations; in words 5 and 9 the women's curve also extends further than that of the men. In the lower ranges, i. e., 0-10, of these curves, on the contrary, the curve for the men is above that for the women. In word 4 the women have a slightly greater number of associations. In words 2, 3, 7, and 8 the two are too nearly equal to make any distinctions, while in word 1 the men have a somewhat greater number of associations. The total result out of the ten words is, therefore, as follows: one word -- associations by men more numerous; four words -- men and women equal; and five words -- associations by women more numerous. The only word which might be criticised as possessing more suggestive power for one sex than for the other -- football (No. 8) -- did not prove to call forth more associations from men than from women.

A combination of the ten curves into one shows very distinctly the advantage [p. 105] of the women on the side of number of associations. The curve was formed by a simple summation of the ten curves for women into one curve, and of the ten curves for men into another. All the cases of association series of less than two by women were grouped together, all those between two and four together, and so on for the entire series. The same proceeding was followed for the men. The resulting figure (63) shows that the curve for the men starts nearer the zero point than that of the women and keeps above it as far as eight. It crosses the women's curve at ten and falls far below it at twelve; at fourteen and sixteen it is slightly above, but falls below again at eighteen and twenty. In general, then, there are more cases of short association series belonging to the men than to the women, and more cases of long association series belonging to the women than to the men. We may conclude, therefore, that at least [p. 106] under the conditions of the experiment, women's minds form associations more rapidly than men's.

In counting the total number of associations the fact that the number of different topics touched upon within the single association series did not correspond with the total number of associations was very noticeable. Some short series covered several entirely distinct topics, while many long series consisted merely of many details about a single topic. The associations were accordingly counted a second time with reference to the number of separate topics touched upon within the series. Details about one event or one person, or reflections on one idea were counted as a single topic.

The results are given in Table XIII. They show that the men touched upon a smaller number of topics in the course of their associations than did the women. For every one of the ten words, there were more men than women who touched upon not more than two topics. It was also true that all the highest records were those of men, but these were few in num-[p. 107 (Table XIII)][p. 108]ber. The difference is shown more plainly by the curves plotted from the table (Fig. 64). These were formed like the curves for the total number of associations for the ten words (Fig. 63). The abscissas mean the number of associations, and the ordinates the number of cases in all ten words in which the series of associations touched upon the number of topics represented by the abscissas. The tendency on the part of the men to stick to one topic rather than to wander over several is shown by the far greater number of cases in which not more than two topics were touched upon in their series.

The outcome of the association test is, therefore, that women have a greater number of associations in a given length of time than men, and that they cover a greater number of topics. There are two factors which seem to be [p. 109] of equal importance in logical processes: one is the ability to concentrate the attention on one topic, and the other is the presence of a large number and great variety of associations. The solution of problems is usually effected through some association which is not at first seen to be relevant. Concentration of attention, if it means mere sticking to one idea, is of little value in intellectual processes, while mere variety of association, without concentration of attention is equally useless. The association test indicates that men have the advantage in one of these factors and women in the other. It furnishes no ground for any statement about comparative intellectual ability.


The mental processes included under the general head of ingenuity are very complex and difficult of analysis. The only attempt made to measure this: faculty consisted in determining the time required by different individuals for the solution of the same problem. Differences in method, important as they are for psychological analysis, were disregarded because of the extreme difficulty of determining them accurately and classifying them. The process was evaluated merely in terms of effectiveness in obtaining a solution quickly. In order to give different types of mind approximately equal advantages, five different tests of ingenuity were selected. They varied from one which required primarily perceptual quickness, to one whose solution depended chiefly on abstract reasoning. Each of the five experiments will be described and discussed separately.[p. 110]

The experimenter was led to regret that a more systematic attempt had not been made to record the methods of solving the problems. Although it seemed doubtful whether or not sexual differences would have been revealed by this procedure, yet it is probable that some valuable results in the technique of the solution of a problem, and in the relative effectiveness of various methods, might have been obtained. Wherever interesting differences of method were noticed they have been mentioned, not because of their bearing on the problem in hand, but because they seemed suggestive of further possible investigations of the more complex mental processes.

The chief source of error, both in the ingenuity tests and in the subsequent tests on general information, was that since the same problems were used for all, some individuals might have been told what they were by those who had already been subjected to the experiment. All the precautions possible were taken against this. The subjects were requested not to tell what the problems were, and were asked whether or not they had been told what they were before the problems were given. There were very few cases where there was any suspicion on the part of the experimenter of any previous knowledge. Even granting that there is an unknown error in the results due to this cause, they are fairer than they would have been if different problems had been used for different subjects, because of the impossibility of measuring the difficulty of a problem exactly.

The first ingenuity test was one selected because its solution depended chiefly on skill in manipulating and transforming a visual perception, although it was [p. 111] not solved through perception by all subjects. Fifteen matches were laid on the table in such a way that they formed five squares in the relative position shown in Fig. 65.

The subject was then asked if he had ever seen the figure before or knew its purpose. One of the fifty -- a woman -- had seen it before, but had forgotten its purpose. She found the solution in ten seconds, but since she was doubtless assisted by her previous acquaintance with the figure, her record is not included in the curve. The others, upon stating that they had no previous knowledge of the figure or its purpose, were told that the problem was to remove three matches from it in such a way that three perfect squares only remained; in other words, to remove three matches in such a way that every match remaining on the table after the three were removed should be a part of a perfect square. No rearranging of the remaining matches was allowed. The subjects were all given exactly the same directions, and were left entirely free to use any method they chose. Removing matches on trial was permitted. Time was counted from the moment the conditions were understood.

Three different methods of solution were employed. The first consisted in trying, either actually or in imagination, the effect of removing various combinations of three matches; the second in attacking the problem from the standpoint of the solution and trying to discover what combination of three squares [p. 112] would leave three superfluous matches; the third in a logical process like the following: There are fifteen matches in the figure; removing three leaves twelve. The twelve remaining matches must form three squares, showing that the three squares must be detached, i. e., can have no side common to any two squares. There are only three squares in the figure which conform to these conditions. It is easy to select these three and to see that the removal of three matches leaves them alone on the table. In most cases more than one of these methods were tried before the solution was obtained. In general the second method was quickest, but in the case of the most rapid solutions, it was usually difficult for the subject to tell what method he had used; all he could say was that he saw, almost as soon as he looked at the figure, which were the required matches. The logical solution was used in only a few cases. It took from five to fifteen minutes. The long times were filled out by a more or less aimless trying of various combinations of three matches.

The curves showing the result of the first ingenuity test are given in Fig. 66. The women are, as a whole, quicker. The advantage of the women in this case is probably a little greater than is represented in the curves. One woman who was very quick at such problems was excluded from the tabulation of results because she was under the impression that she had seen the puzzle before. The woman who is recorded as "failed" was one of the first people tested. She gave up after working fifteen minutes. (In subsequent cases the subjects were required to work until the solution was obtained.) However, even as the curve [p. 113] stands, there is a distinct advantage on the side of the women. There are more of them in the range of short times, and there are five men who took longer than the slowest woman.

The second ingenuity test was designed to call a pure process of reasoning into play. It consisted of a puzzling mathematical problem, perfectly simple in the computations involved but demanding a somewhat complicated process of reasoning for its solution -- a problem in which it was easy to become confused unless all the factors were sharply separated and clearly

grasped. The problem was handed written to the subject. He was told that it involved no difficult computations. The process was timed from the moment the problem had been read through. A failure was recorded only in cases in which the subject had worked from forty-five minutes to an hour, and was completely hopeless of getting any solution. The problem was the following: "A man swimming in a river finds that he can swim three times as fast down stream as up stream. The river flows at the rate of a mile an hour. Find his rate of swimming in still [p. 114] water." accepted. A mere stumbling upon the correct answer was not called a solution.

The curves of results for the second test (Fig. 67) show no marked difference between the men and the women in quickness of calculation. On an average the men have somewhat the advantage. Two of the men with very good records had been teaching mathematics within a year, while none of the women were primarily concerned with it. Taking this fact into consideration, the difference between the two curves is insignificant.

The third problem involved to some extent both perceptual quickness and reasoning. The apparatus consisted of a checker-board, composed of red and yellow squares, and eight men. The board had eight squares, four of each color on each side, like an ordinary checker-board. The problem was to place the eight men on the board in such a way that no two were on the same straight line of squares, either perpendicularly, horizontally, or diagonally. Both red and yellow [p. 115] squares could be used. The problem is far too complicated to be solved by inspection, nor can it be reasoned out in detail. The process in solving it consisted in adopting some general method, and trying it, modifying it if necessary, until the proper combination was hit upon. There was one method which gave a logical certainty of some solution, but only one or two subjects discovered it. Most of them proceeded by starting at some part of the board and trying to work systematically from that point. If that failed, another portion was taken as starting-point and the trial made again. Some few placed all eight men in a row along one side of the board, and worked by moving seven of them out from this position, varying the placing until the conditions were fulfilled. In almost all cases the subject felt that the solution, when obtained, was largely a matter of chance. He had simply stumbled upon the right combination, rather than really solved a problem. Very few of them could have reproduced the solution after the men were removed from the board. However, the problem certainly required original method, quickness in seeing complicated forms, and perseverance.

In this test the men show themselves decidedly superior to the women (Fig. 68). There were two women who were quicker than any of the men, but there [p. 116] were also two more women than men who failed. The superiority of the men is shown in the great majority of them in the region from fifteen to thirty minutes. It is very difficult to evaluate this difference, because of the indefinite nature of the problem. Most of the women expressed a great distaste for all such problems, because they were uninteresting. Many of them were so uninterested that they did not really work at it. Whether the men found it equally uninteresting, but forced themselves to work in spite of lack of interest, or whether the problem appealed to them as more interesting, it is difficult to say. From the voluntary comments of the subjects the latter hypothesis seems more probable, but in this case we are confronted by the further question why such a problem should have more interest for men than for women. The test may point to a greater interest on the part of men in a problem, merely as a problem, regardless of any possible usefulness, or any further application.

The fourth trial of ingenuity was a mechanical problem. The subject was required to find out the method of operating the apparatus used for determining the light threshold described above on p. 76.

The test was always made shortly after the determination of the light threshold, when there was usually a spontaneous interest in the apparatus. The necessity for uniformity in the time of the stimulus was explained, and the subject was told that his problem was to find out, first, how a constant length for all exposures of light was secured automatically, and second, how the absolute intensity of the light: was varied. He was told that he might do anything he pleased with the apparatus, except take out screws, which would not be necessary to discover its workings.[p. 117]

The apparatus was particularly favorable for a comparative test, because it was so entirely unfamiliar to all the subjects. The mechanical principles involved in it were all very elementary. The difficulty was to find out how the various parts worked together. Since the apparatus was so unique, acquaintance with other mechanisms was of as little assistance as possible, although unquestionably a knowledge of mechanics in general is of assistance in solving any particular mechanical problem, however unlike previous ones it may be.

Two difficulties were experienced by all the subjects in this problem. The first was in making the connection between the metal ball and the inclined trough. f The ball was not found in the trough, but lying on the table by the apparatus, either in an open box or in the stop at the foot of the padded incline. The second difficulty, which was still more serious, was to discover that the ball could be released and the screen opened by a single movement. The fact that the ball would close the screen was found out very quickly. The method of regulating the absolute intensity of the light gave but little trouble.

Only twenty-one women and twenty-two men are recorded in the curves for this test, because it was not given to the first subjects on whom the experiments were performed. The apparatus was explained to them at their request. When later the decision to use it for an ingenuity test was made it was of course impossible to apply the test to these subjects.

In the fourth test the men have a somewhat better record on the whole than the women (Fig. 69), although the difference is not marked. There is a considerable majority of the men in the region under fifteen [p. 118] minutes, but they are also somewhat in excess at the other end of the curve. None of the women failed to get a solution in an hour, while two men worked from an hour and a quarter to an hour and a half and failed.

The fifth ingenuity experiment was also the solution of a mechanical problem. The subject was given a wooden model of a very simple combination lock. He was told that it was a model of an object with the use of which he was perfectly familiar, although the form was unusual, and that he was to find out what it was and how it worked. He was told whether or not he was correct when he thought he knew its use. Most of the subjects could tell what it was before they discovered how it worked, although in a few cases the reverse was true.

A diagram of the lock is given in Fig. 70. The inside of the lock is shown in its locked position, i. e., with the bolt A out. The bolt was held in place by a rod at B, which passed through a long, narrow opening, leaving the bolt free to move back and forth the distance of the opening, and also up and down about the rod as axis, from the top of the lock to the bottom. A spring, C, passed from the inner end of the bolt A to a rod fastened to the wall of the lock at D. E and F represent pegs which moved freely about their axes. Handles from these pegs projected about one inch through the wall of the lock. To unlock it, the peg [p. 119] E was first turned so as to raise the bolt toward F. When the bolt was at its highest point the catch, G, was opposite the peg F. When F was then turned so that its thin edge pointed to the back of the lock, it carried the bolt with it, and braced it inside of the lock so that none of it projected. The lock was then unlocked. To lock it, it was only necessary to turn the peg F back to the position shown in the figure, when the bolt, through the action of the spring C, at once flew back to its original position. The lock was always given to the subject in its locked position as shown in the sketch.

The chief difficulty experienced in this test was in discovering that the bolt would move in and out, as well as up and down. The up-and-down motion was apparent as soon as the pegs were moved, but the opening in the bolt on which it moved in and out was so far back that it could not be seen by looking in at the open end of the lock, and the subjects were not allowed to take it apart. In some cases the in-and-out movement was discovered by an accidental pressing on the bolt, sometimes it occurred to the subject to try that movement purposely, and sometimes it was found by experimenting with the peg F and its relation to the catch G. After the in-and-out motion was once discovered the solution usually came quickly. At first most of the subjects explained it on [p. 120] the basis of the up-and-down movement as a latch, and worked out the other solution only when they were told that it was not a latch.

The results show a very evident advantage on the side of the men. The majority of them solved the problem in less than twelve minutes, while the majority of the women took more than twelve minutes. Two women failed entirely, while all the men worked out the solution in forty-five minutes or less. The difference was no doubt partly due to the fact that most of the men were familiar with the construction of locks in general (although none of them knew exactly this form), while the women had had much less experience with locks of any sort.

To sum up the results of the ingenuity tests, they show that, on the whole, the men have a decided advantage. They were much superior to the women in two tests (the third and the fifth), somewhat superior in one (the fourth), equal in one (the second), and inferior in one (the first). There are several indications that special education plays some part in these results. Two of the problems in solving which the men proved superior to the women, viz., those of the lock and the visual apparatus, were in the realm of mechanics, with which men are by education more familiar than women, In the lock problem the men's superiority is marked, while in the visual-apparatus [p. 121] problem it is only slight. The latter problem, dealing as it did with a unique machine, was one in which previous experience with mechanical contrivances would be of comparatively little assistance. Experience with locks, however, would assist materially in solving the lock problem, though the form of the lock was unfamiliar. Of the three non-mechanical problems, the women were better in one, the men were better in one, and they were equal in one.


The questions to test general information were selected as a test for college students, not as a representative set of questions for intelligent people in general. The correct answers were facts that a college student of the third or fourth year could fairly be expected to know. The majority of them were facts that the average college student must have known at some time during his career. It was sought to make the questions perfectly fair and representative; there were no catch questions. In order to make the evaluation of the results as exact as possible, questions of fact only were asked. The answers are definitely either right or wrong; they can be marked with very little variation due to the personal equation. An exact evaluation of questions of theory or opinion is much more difficult.

The questions were handed to the subject written, and he was given all the time he wished to answer them. They were as follows:

1. Name two writers of English who wrote before Shakespeare; give the title of one work of each, and tell whether it was poetry or prose.[p. 121]

2. Give approximately the dates of the period during which Shakespeare wrote.

3. To what nation and what period does each of the following writers belong: Pope, Racine, Schiller, Coleridge, Balzac, Dryden, Petrarch, Heine, Browning, Ibsen?

4. Name one work of each of the following writers: Tolstoi, Charlotte Bronte, Macaulay, Victor Hugo, Nathaniel Hawthorne.

5. Who wrote the following works: Tom Jones, Cyrano de Bergerac, Two Gentlemen of Verona, The Excursion, Pride and Prejudice, Richard Feverel, Childe Harold, Adam Bede, The Vicar of Wakefield, The Newcomes?

6. Name the great subdivisions of the Aryan race.

7. Name the nations occupying the Tigris and Euphrates valley previous to the time of the Roman empire.

8. Name (a) two famous lawgivers of ancient Greece, and
    (b) three Grecian cities which, at different times, held supremacy over Greece.

9. (a) When did the French Revolution occur?
    (b) Name three men who were prominent in French politics during the five years subsequent to the beginning of the Revolution.

10. (a) When did the Roman republic cease?
      (b) What form of government followed the republic?
      (c) Who brought about the change?

11. (a) What was the Missouri Compromise?
      (b) What is its date, approximately?

12. Is hypnotism an established scientific fact, or is it fraud and superstition?

13. How does the binomial theorem lessen labor in mathematics?

14. Solve this equation for x: 5x²-3x = 2.

15. What is (a) a sine? (b) a tangent?

16. (a) What are the fundamental laws of motion?
      (b) Who first formulated them?

17. What does it mean to say that the specific gravity of a body is four?

18. What is the principle on which the telephone works?

19. Is the energy furnished by an electric battery created in the battery? If not, where does it come from?[p. 123]

20. Give the chemical formula for water, and explain its meaning.

21. What happens to the substance of a piece of wood when it is burned? Is any of it destroyed?

22. Are there any cases of spontaneous generation among living organisms? If so, where?

23. What is the nature of the simplest type of animal known?

24. Name the departments of biology, and the other branches which have contributed most to establish the truth of the theory of evolution.

25. For what were the following men noted, and in what century did they live: Weissmann, Socrates, Esterhazy, John Stuart Mill, Each, Charlemagne, Nebuchadnezzar, Kant, Pericles, Bacon, Rameses II., Goethe, Alfred the Great, Dante, Alexander, Kepler, Richelieu, Edmund Spenser, Galileo, Herbert Spencer?

Even in questions as definite as these some difficulties arise in grading. For instance, where approximate dates are asked for, how close an approximation shall be demanded? In each of these cases a more or less arbitrary standard was adopted. For instance, if the dates given for the period during which Shakespeare's plays were written included the greater part of the correct period, the answer was given full credit; if they included a small part only, part credit was allowed. If they fell entirely out of the correct period, no credit was given. Answers to the third question were considered correct as to date if the correct century was given for each of the writers named. The latter part of the eighteenth century was considered a correct answer as to the date of the French Revolution. Dates within fifty years, on either side, of the end of the Roman republic and within ten years of the Missouri Compromise were called correct answers as to those events. The last question was [p. 124] considered correctly answered as to dates if the century given for each man mentioned were within one hundred years, on either side, of the correct period, except in the case of men who lived in the eighteenth or nineteenth century, where the correct century was required in the answer. In the very early dates, still more leeway than a century was given.

Before giving the results of this test, one very evident source of error must be mentioned, which was also met with in the ingenuity tests. The same set of questions was used for the entire series of subjects, and there was always the possibility that later subjects had been told some of the questions by previous ones. Each subject was requested not to talk about the questions to other students, because the same set of questions was to be used for all. Each subject was asked before he was given the questions whether or not he had been told anything about them. Aside from these precautions, there was nothing to be done except trust to the honesty of the subject. Any accurate evaluation of the test would have been impossible if different sets of questions had been used, because no two sets of questions of exactly equal difficulty could be made out. Just how large a part a previous knowledge of the questions really played in the results, it is of course impossible to say. The impression of [p. 125] the experimenter was that it was very small. There were only one or two cases where there was even a suspicion of such knowledge.

The twenty-five questions fall into the following classes:

English literature (five questions).
History (six questions).
Physics (four questions).
Mathematics (three questions).
Biology (three questions).
Chemistry (two questions).
Psychology (one question).
General (one question).
The results of the tests will be given both for the entire series and for the separate divisions. The answers were graded on a basis of 10 for each question. The highest possible grade for the entire series is therefore 250; that for each division is given below the appropriate diagram of results.

The two curves for the total examination (Fig. 72), although different, do not differ in such a way that we may call one better than the other. Both the extreme records, 100 and 220, are those of women; but on the other hand the men are more numerous than the women at both extremes. There are six men and only two women under 120, and also [p. 126] six men and two women over 180. The curves, on the whole, coincide remarkably. Both center around 160, where each stands at three. Each one has twelve below and ten above 160. The curves representing the grades in English literature (Fig. 73) show a decided advantage on the side of the women. From thirty-five down the curve for the men is above, while from thirty-five up, that of the women is above.

The results of the examination on history appear in Fig. 74. What difference there is between the two curves is in favor of the men, although it is not very great.

In physics (Fig. 75) the men have a decided advantage. The extremes of the two curves are the same, and the women are slightly more numerous in the region of the best records; but the general course of the men's curve is better than that of the women's. The majority of the women fall below fifteen, the majority of the men above.[p. 127]

The two curves for the examination on mathematics (Fig. 76) correspond closely. What difference there is, is in favor of the women. There are more men than women in the lower ranges, and more women than men in the upper.

In biology (Fig. 77) we find the men in excess at both the good and bad extremes. The general course of the curves, however, shows a somewhat higher average in the women's record. The questions in chemistry were both so simple that the answers were almost all correct. Eighteen women and seventeen men were graded at ten on both questions.

In the question on hypnotism the men have a better record than the women. The men all answered correctly. Four women failed on the question.

The results of the examination on question 25 appear in Fig. 78. Here again the men are in excess at both extremes. It is impossible to call either curve better, on the whole.

To assist further in the analysis of the results of [p. 128] the general-information tests, the grades in English literature and history, as the literary subjects, and those in physics, mathematics, biology, and chemistry, as the scientific subjects, were summed. The two summations are shown in Figs. 79 and 80. In the diagram for the literary subjects the men are more numerous in the middle ranges and the women in the higher, while in the diagram for the scientific subjects the reverse is the case.

The results of the series of tests on general information may be summed up as follows: In average grade on the entire series of questions there is no difference between the men and the women. There is, however, a difference in grouping. The men are more numerous at both good and bad extremes than the women, and the women more numerous than the men in the middle ranges. The women stand better than the men in the literary subjects, and not so well in the scientific. This does not mean that the women [p. 129] were superior in both the literary subjects nor that the men were superior in all the scientific. The relation of the sexes in the separate subjects was as follows:

English literature: women much superior.
History: men a little superior.
Physics: men much superior.
Mathematics: women very slightly superior.
Biology: women a little superior.
Chemistry: both sexes equal.
In the results of the tests on general information, as in those on ingenuity, special training is unquestionably a factor. As appears from Table XXVII, far more women than men were interested in English literature. Although the women were as interested in science as the men, probably the stress of their work had fallen more on literary than on scientific studies. Many of the women were preparing to be teachers, and had, therefore, from practical considerations devoted themselves primarily to those subjects in which the openings for women are most numerous, viz., literary subjects. Many of the men, on the other hand, intended to be physicians, and hence were laying the stress of their work on scientific studies. The slight superiority of the men in history is probably due to the presence of several students who were preparing for a law course.[p. 130]


There have been a number of researches on the comparative memory of the sexes, although none of the others have required memorizing or measured retentiveness. The other tests have all followed the method of making a single presentation of some series of stimuli and requiring the subject to reproduce it. The power of memory was then measured by the accuracy of the reproduction. Tests on university students have been made at Columbia University (82) and at the University of Wisconsin, the latter by Jastrow (35). At Columbia University visual, auditory, and logical memory were tested; the first two by presenting series of numerals through the eye and the ear, and the last by reading aloud a passage to be reproduced. The result was to show that women had a decided advantage in visual memory, men a doubtful advantage in auditory memory, while there was no difference in logical memory. Jastrow's method was to display a series of words one by one, requiring the subject to write the first association which occurred to him. Two days later the subject was asked to write the original list again from memory. In this test women made a better record than men. It was afterward performed on high-school students with the same result. Stern (76a) has made a few tests tending to show that the memory of women for the details of pictures is completer than that of men, but that they add in recollection more of imaginary material than do men. Bolton (9) tested auditory memory in school children, both in the high school and in the grade schools, by the method of reading [p. 131] aloud a series of numerals, which were then reproduced. He finds the girls decidedly better than the boys. Shaw (75) employed the method of requiring school children to reproduce a story which had been read to them, and again the girls made the better record. Netschajeff (60) experimented on school children of St. Petersburg, varying from nine to eighteen years in age. He tested memory for objects, sounds, numbers, and various kinds of words. His general result is that, with slight exceptions, girls excel boys in power to recall. Very much the same series of tests was performed by Lobsien (48) on school children of Kiel between the ages of nine and fourteen and a half. His results are formulated with reference both to the number of impressions reproduced and to the correctness of the order in which they were reproduced. He finds that the girls excel the boys in both respects. Ebbinghaus (22) used memory as one of his methods of testing mental ability. He stands alone in finding girls inferior to boys up to the age of eleven or twelve years. The fact that the boys and girls he tested were being educated in different kinds of schools (Gymnasium and Mädchenschule) may account for this discrepancy. The results of former experiments thus agree almost unanimously with the present series in showing better memory in females than in males.

Two investigations other than the present are on record in which the attempt has been made to obtain a statement of the comparative rapidity of the association process in men and women. Jastrow (34) took the time required to write one hundred words as rapidly as possible. He found no difference in the [p. 132] time required by men and women; but as he himself says, his method of measurement was rough. The second association test referred to was made at Columbia University (82). It consisted in requiring the subject to write the first association to each of nine words. The words were given the subject printed on a card. The entire time of the process was taken. The men proved to be more rapid than the women. Neither of these results accords with those of the present test, which show the women to be somewhat more rapid. If what is sought is a measurement of the normal rapidity in passing from one idea to the next when the process is made as natural as possible, the form of test employed in this series seems better fitted to give the required value than the two just described. The process of merely observing and noting down the thought sequence as it occurs from a given starting-point seems less artificial than that of writing down one hundred different words, or of writing the first association to a given word. The writer is therefore inclined to put more faith in the results of this test than in those of the others.

Qualitative, as distinguished from quantitative, comparisons of the association faculty in men and women have been made experimentally by Jastrow (34, 35, 36, 37) and by Miss Calkins (13) and her students (61). Two methods were employed by each. The first was to require the subject to write one hundred different words as rapidly as possible. Jastrow finds that men furnish a greater variety of words and a greater number of unique ones than women. Miss Calkins, on the contrary, finds the variety of words furnished by women about equal to [p. 133] that of men, and their number of unique words greater than that of men. The second method consisted in requiring the first association to each of a list of given words. Neither investigator discovered any difference by this method. They agree in finding some classes of words mentioned more frequently by one sex than by the other, such as food-stuffs by women. But as both Miss Calkins (13) and Miss Tanner (79) have pointed out, this fact points not to original sexual difference in type of mental activity, but to difference of training and surroundings from childhood on.

There are no previous tests on ingenuity and general information with which those above recorded may be compared. It is well known that in school work girls have better records on the whole than boys (55, p. 1045). But the general average of school work is not comparable to the results of our test on general information. The former takes account of the way in which the lessons assigned are learned, the latter of the amount of definite information which the individual has at hand when it is suddenly called for.

There are a few other investigations which have a more or less remote bearing on the intellectual tests. Lindley (47), investigating puzzle interests, says that he discovered no difference in the age at which the various puzzle interests develop in the two sexes. He attributes this failure to the small number of individuals investigated. Dearborn (19), studying the imagination by means of ink blots, found no difference between the sexes. Minot (58), found greater uniformity in women's diagrams than in men's. Miss Calkins (12), investigating the mathematical consciousness,[p. 134] found from answers to a questionnaire that men are more likely to reason out a mathematical demonstration, and less likely to memorize it, than are women. This is not in agreement with the present tests, which show that the women as a whole have an equal capacity with the men for furnishing an original solution of a mathematical problem when it is called for unexpectedly. Ebbinghaus (22), and Bellei (8), both made tests on school children which were intended to measure intellectual ability. The former used the methods of mental arithmetic, memory, and what he calls a combination method, which consisted in requiring the child to fill in the omissions in a text which had been prepared with some syllables or letters omitted. The rapidity and accuracy with which this could be done was regarded as a measure of mental ability. The latter used the first and third of the methods just described. Ebbinghaus found the boys superior to the girls up to the age of fifteen, when the girls were somewhat superior. Bellei's results do not agree with Ebbinghaus's. His experiments were confined to children of a single class in school having an average age of eleven years. He finds the girls superior to the boys.


It is well established that women have better memories than men; they memorize more quickly and retain as well. The results of the various experiments on association do not agree as to either quantitative or qualitative differences of sex. The most trustworthy evidence goes to show that the process of association [p. 135] is somewhat more rapid in women than in men. As to qualitative differences, none of the methods employed seems to have thrown, or to be capable of throwing, any real light on this question. The experiments which have been performed to determine comparative ingenuity show the men superior to the women. There are indications, however, that mechanical training, which boys unquestionably receive to a greater extent than do girls, is an important factor in this result. The question whether the more extensive mechanical education of boys is not to be accounted for by their greater natural ingenuity will be discussed later. In total amount of general information there is no difference between men and women who have taken the same course of education. The women are somewhat the better informed in literary and the men in scientific subjects, but this is probably due to selection of studies and not to sex.