Classics in the History of Psychology

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

(Return to index)

Functional Periodicity

Leta Stetter Hollingworth (1914)
Posted July 2000

[p. 15] III



This is a familiar test, which has been used to study a great variety of factors, such as individual differences, fatigue, effects of drugs, insanity, etc. For a complete list of investigations carried out by means of this test, see Whipple's "Manual of Mental and Physical Tests," p. 114.

G. S. Hall [1] says of this test, "The greatest number of taps that can be made in a given brief time interval is an important determination . . . for accessory muscular control. . . . This measurement is very important, and marks one of the factors of motor ability. How rapidly two like simple volitional contractions can follow each other is perhaps the best index we have of will time."

The apparatus used for this experiment was the very familiar motor-board employed in the Columbia laboratory. The individual sat before this board, which was placed on a table, and tapped 400 times at maximal speed, on a brass plate, with a small brass rod, each tap being registered by an electric counter. The subject began tapping at a signal from the operator, and the time of performance was recorded by the operator with a stop watch, in fifths of a second. Thus the amount of performance remained constant, while the time varied.

Whipple [2] points out that "The use of an electric counter such as some investigators have employed -- gives no indication of changes in speed during the trial." Also that the electric counter will miss a quick tap, and is therefore not absolutely reliable. In the present case no attempt was made to register anything except the gross speed of performing 400 taps, with the right [p. 16] hand. The apparatus seemed to be perfectly satisfactory for this purpose, and failure to record a tap was never observed.

If the will power of an individual were remitted during an affected period, if muscular tonus were lowered, or the general motor ability decreased, this test would be expected to reveal the condition. Hollingworth [3] found it to be sensitive to so small an influence as 1 to 2 grains of caffein.

The complete daily records of the seven subjects are given in Table I, and require no comment.

Table II gives the daily records converted into averages of five-day periods each. The five-day periods are determined in the case of the women by counting backward and forward from physiological periods. Thus a few "blocks" are formed by four or six days, instead of five, since the number of days between periods is not always exactly divisible by five. In the case of M1 and M2, the five-day blocks are determined by beginning to average parallel to the records of F1 and of F4, respectively. Thus are yielded two controls composed of five-day blocks arbitrarily determined, from records made by human beings not subject to the phenomenon in question.

Examining the contents of Table II, we find that the records marked (*) form a consistent part of an ordinary practice curve. They show no tendency to rise above the records preceding, and thus to cause "breaks" of inefficiency in the down-sweeping practice curve. The record made by F1 on January 12-16 is an exception, however, this being 53.1* as compared with 50.9 and 51.2 for the two five-day periods preceding. F1 shows no such tendency in the other two critical periods included in her record. and F2, F3, F4 and F6 show no such tendency whatever. The control records also reveal averages which revert to a level of efficiency attained ten and fifteen days previously.

The mean variation (M. V.) of records made during periods presumably affected shows no uniform tendency. It may be observed in passing that the M. V.'s for this test are very small, rendering the figures exceedingly reliable.

Table III gives the results of treating the figures as follows: The records are averaged in fives as in Table II, except that only three groups of periods are considered, i. e., the five days [p. 17] preceding, the five days during, and the five days following the critical period. Also, in Table III the women are treated separately, and the control records are averaged in periods exactly corresponding to those of the subject under special consideration, which was not done in Table II. The averages for the days before and after are then averaged, and the result compared with the average of the five days during, to see if the latter will yield a uniformly poorer or better average. The control records are treated in exactly the same way. In this manner is obtained a standard of efficiency by which to measure the ability of the critical period, eliminating the factor of practice.

This method of averaging preceding and following periods for purposes of comparison with a middle period to determine relative efficiency assumes a uniform or nearly uniform rate of improvement in the practice curve. And the frequency with which the average of the periods preceding and following corresponds exactly or nearly to that of the middle period seems to indicate that the rate of improvement, in this test at least, does tend to be uniform. Thus if the middle period were subject to impaired efficiency, it would appear that averages computed in the way described should reveal the fact.

Furthermore, Table III shows the average for four days of each critical period, excluding the first day, which has been said to be subject to greater impairment than the other days, and the record of the first day is given separately.

The case of F1 is first considered in Table III. On the first month this subject, as already remarked, made a poor record at the critical period, the figures being 53.1* as compared with 49.9, the standard of efficiency for that period. On the other two months included, the averages compared are found to be identical, indicating no influence whatever.

Scrutinizing the record of F2, it is seen that the averages to be compared are identical for three out of the four months included, indicating no influence whatever. On the second month the average for the critical period is 38.5* as compared with 39.4, the standard of efficiency, but the difference is within the M. V., so no advantage is indicated for days presumably affected.

The records of F3, F4 and F6 confirm the records of the other subjects. No influence of periodicity is indicated by the [p. 18] average of performance. The very slight differences between the critical periods and the standards of efficiency fall in all cases well within the M. V. It must be said that these observations reveal no influence whatever on the average of performance.

Table III fails to demonstrate greater impairment on the first day of menstruation. In fact the figures suggest an increased briskness of performance on the first day in the records of all the women. In eleven instances out of the fifteen herein observed, the first day's record exceeds in speed the average, not only of the five days in which it is included, but of the period following. The amount of the apparently increased briskness is, however, very slight, and it is without doubt due solely to chance.

F2 experienced some physical pain on the first day of the first two periods, and none on tire last two. This fact seems to hare made no difference in the records.

As Hall remarks, "This process is always very rapidly fatiguing." Although no attempt was made to measure the difference in time between the first 200 and the last 200 taps, it might be supposed that if the tendency to fatigue were considerably greater during an affected period, the total time of performance would be measurably increased by the last 200 taps. This does not appear.

When curves are platted from Tables I and II. and are examined with a view to determining whether a regularly recurring period of maximum efficiency is discernible within each month, we find no clear result. There does seem to be in this test a tendency to a rhythm of efficiency, but this tendency is just as clearly present in the case of M1 and M2 as in the case of F1, F2, F3, F4 and F6, and on the basis of this experiment is, therefore, not attributable to the phenomenon of menstruation as a cause. It is not possible to say on the basis of these experiments whether the appearance of rhythm is accidental, especially since it is by no means clearly defined, but only vaguely suggested. The period of maximum efficiency (speediest performance) seems to fall just before and on the first day of the period in the case of the women subjects. At any rate the figures do not support the statement of expert opinion that fatiguability is much greater, will power weaker, and motor energy dimin-[p. 19]ished at this time. It will be well to point out in this connection that in any curve of work there will always be fluctuations which are of a more or less accidental nature, and which are not at all a matter of a fundamental "cycle" of efficiency. In the very nature of the case, regardless of concomitant phenomena, there will be high points and low points on the curve, within each month or within any limits of time we may wish to specify.[4][pp. 19-21, Table I]

[pp. 22-23, Table II]

[pp. 24-26, Table III]


In order to obtain a further and slightly different measure of voluntary speed, the time of performance for the first 300 taps in the experiment on fatiguability was computed and tabulated. (For description of method and apparatus see chapter on Motor Fatiguability.) Table IV gives the daily records for the three subjects (F2, F3, M1). Table V gives the records averaged in five-day "blocks," as in Table II under Part I of this test. Table VI shows the figures computed as in Table III, and with the same purpose, i.e., to compare the average of the critical period with a standard of efficiency; to examine the average performance of four critical days, excluding the first; and to determine whether the first day shows a remarkable impairment.

The results of this test are in agreement with those of Part I. In Table V the records marked (*) form part of an ordinary practice curve, revealing no inefficiency. Table VI shows the critical average to be always slightly better than the standard of efficiency, but the difference falls in all cases within the M. V., so no genuine advantage for critical periods is indicated. Four out of six first days show a performance superior not only to the average in which they are included, but to the average following.

The conclusion, therefore, is that in this test as in the other, no influence whatever on the average of performance is indicated for critical periods.[p. 28]

[p. 29]

[pp. 30-31]


[1] G. Stanley Hall, op. cit., p. 142.

[2] Whipple, Manual of Mental and Physical Tests, p. 106.

[3] H. L. Hollingworth, op. cit., p. 43.

[4] E. L. Thorndike, The Curve of Work, Psych. Rev., May, 1912, pp. 168-69.