Vol. 35, No.

1 January, 1938

THE

PSYCHOLOGICAL BULLETIN

MEMORY SPAN: A REVIEW OF THE LITERATURE

BY ALBERT B. BLANKENSHIP1
Psychological Corporation
Although many researches have been undertaken in the field of
memory span, both by experimental psychologists and by those in
the field of mental testing, there has been no attempt to summarize
or to systematize the findings of the investigators who have studied
the various aspects of this subject. This paper has been undertaken
with that end in view.

I. HISTORY OF MEMORY SPAN TESTS

(1) Early Anecdotes. Stories of exceptional individuals have probably-
been told since the beginning of time, and reports about memory spans of such
individuals are not lacking. There was a blind Swiss, for example, who was
reputed to have been able to repeat a series of ISO numbers, either forwards or
Tjackwards, after a single hearing (91) ! Other anecdotes are in the litera-
ture (8, 44, 47, 91, 130), but there was no real attempt at controlled observa-
tion, and none of the early writers realized the significance of what they had
Tieard of or observed.
(2) Nineteenth Century Studies of Memory Span. In 1870, Oliver Wendell
Holmes, in addressing the Phi Beta Kappa Society of Harvard University,
said, " . . . in uttering a series of unconnected words or letters before a
succession of careful listeners, I have been surprised to find how generally
they break down, in trying to repeat them, between seven and ten figures or
letters; though here and there an individual may be depended on for a larger
number . . ."(61). Holmes, however, made no formal experiments on the
phenomenon.
Sixteen years later, William James (70) wrote of "the present, . . .
merely a dividing line between the past and the future . . . ," but he did no
1
The writer wishes to express his indebtedness to Miles Murphy and
Francis Irwin of the Department of Psychology, University of Pennsylvania,
lor their aid in compiling a bibliography for this summary.
1
2 ALBERT B. BLANKENSHIP
experimental work on the problem. In the same year, Jacobs ( 6 8 ) , an English
philosopher, wrote, " There is . . . a certain number of syllables up to which
each person can repeat . . . after only once hearing; and it is probable that
this number varies with different persons." In 1887, he made the first formal
experiment on "prehension" (69), discovering that the ability increased with
age. Galton (48), in the same journal, noted that inmates of institutions for
the feebleminded possessed lower prehension ability than did the normal chil-
dren tested by Jacobs. In 1888, Burnham (21) summarized some of these
results in his comprehensive paper on memory.
From this time on, memory span research was more common. The most
important work in the field since 1890 will be reviewed here, not primarily from
the historical standpoint, but from the aspect of logical organization. If the
reader be interested in historical summaries including some of the work on
memory span, he is referred to the review by Burnham (21), and those by
Young (144), Kuhlmann (78), Wylie (142), and McGeoch (87).

II. WHAT IS MEMORY SPAN?

To define memory span, one must examine the question from
two viewpoints, the functional and the structural.2
(1) Functional Aspect. Functionally, Binet (8) has defined prehension as
"the maximum number of digits retained after a single hearing." But the defi-
nition, of course, need not be restricted to the use of digits. Humpstone (65)
broadened this definition when he described memory span as " the ability to
grasp a number of discrete units in a single moment of attention and to repro-
duce them immediately." Learning (80) elaborated upon this: " It [memory
span] appears to measure the number of discrete units over which the indi-
vidual can successively distribute his attention and still organize them into a
working unit." Watkins (133) stated that immediate memory " . . . is the
capacity to repeat impressions which have not entirely disappeared from con-
sciousness, the expression following immediately upon the impression."
Strong (125) defined it as " a line of successive presents."
To generalize, memory span refers to the ability of an individual to repro-
duce immediately, after one presentation, a series of discrete stimuli in their
original order. Practically any sort of material may be presented, such as
digits, letters, words, and sounds, and almost any sense organ or combination
of sense organs may be used to receive the impressions. Both of these variables
will be discussed under subsequent headings.
(2) Structural Aspect. A structural definition of memory span is difficult
to give, for one immediately is faced by the distinctions between the prerequi-
sites for memory span, and the actual processes involved. Although an intact
sense organ, an afferent tract, a central projection area, efficient association
fibers, and a certain degree of attention are all involved, as Smith (119) points
out, these terms do nofdescribe the processes actually involved in memory span.
Processes of attention are involved, as McCaulley (86), Gundlach, Roths-
child, and Young (56), Cattell (28). T. L. Bolton (16), Johnson (72), and
2
" Functional" in the sense of external or extrinsic behavior; " structural"
In the sense of the processes involved (the intrinsic aspect).
 MEMORY SPAN: A REVIEW OF THE LITERATURE 3

others indicate. Certainly the subject must be able to distribute his attention
over the series of stimuli, and concentration of attention is needed so that the
mental processes may continue in the direction started. But the range or span
of attention is distinct from memory span. Hunter (66) shows that attention
span and memory span are alike in involving only one presentation of the
stimulus, but that they differ in temporal duration of the stimulus. " If the
stimulus is presented for one-fiftieth of a second, the experiment is classified
as one on attention, whereas with longer exposure times, the behavior is
classified as . . . memory." Motor aspects of attending are evident in the
receipt of instructions and in the postural response of getting ready, as well as
in the receptive attitude during the presentation of the series. Much may be
said for the suggestion (56) that memory span be renamed " set."
" Associability " is also required in memory span. This term, originated by
Humpstone (63, 64, 65), refers to the ability of the subject to group the series
of elements together: to perceive relationships among the series in order to
better reproduce them.
Still another process involved in memory span is that of imagery (20, 86,97).
The subject, in order to be able to reproduce the series presented, must be able
to image the series. But memory span is not an after-image. Richet (110), as
early as 1886, compared the memory span with the sensory after-image. He
recognized that there was a difference, but believed the two to be comparable.
According to common psychological belief, it is generally held that the sensory
after-image depends upon activity not only in the brain, but also in the sense
organ. Humpstone (64) actually calls memory span an after-image.
The actual reproducing of the series of stimuli involves the process of
memory (28, 86). If the individual possessed no memory at all, reproduction
of the series would be impossible. But Binet (8), in 1894, was probably the
first specifically to point out that there was a difference between memory and
memory span. Fernberger maintains (41) that memory span and memory are
different in the length of time over which reproduction is possible. Memory
span is transitory; memory is fairly permanent. In addition, the amount of
material involved in memory span is ordinarily much less than the amount of
material involved in memory. W. G. Smith (119) showed that memory span
may be good and memory bad, or vice-versa, thus offering further proof that
the two are distinct. Reproduction of the series also involves certain other
" reproduction factors," such as language ability and arithmetical proficiency.
Now although memory span is dependent on all of the above functions, it
seems clear that it is not any one of them (20, 64, 123). The question of
whether the ability is dependent or independent is closely related to the present
discussion, and hence is the next topic.

III. Is MEMORY SPAN A GENERAL ABILITY?

(1) Introduction. A few workers (20, 63, 65, 97) regard

memory span as an independent ability. Such attention factors as
observation, distribution of attention, and description, and such
reproductive factors as language ability, memory, and arithmetical
efficiency may also be involved (20). Humpstone's independent
4 ALBERT B. BLANKENSHIP
ability tested is " associability," or, ". . . the ability to grasp and
associate a number of discrete units of perception in a definite
order" (65). Memory and imagination are involved, but memory
span itself is a specific ability.
Binet also holds that memory span is an independent ability (12),
but that the ability tested is the " capacity for effort." This view
assumes that memory span is a general 8 ability, and accordingly,
that the type of material used or the sense organ or organs through
which the material is received should have no effect either on the
number of discrete units reproduced by the individual, or on the
standing of the individual in relation to others tested for memory
span.
(2) The Type of Material. There is practically no limit to the
type of material that can be used in such a test.
Ideas, sentences, objects, pictures, noises, words, paragraphs, diagrams, and
syllables are only a few of the many types of material that have been used,
though as Bronner, Healy, Lowe, and Shimberg point out (19), the use of
digits has preempted the field. For a discussion of the various types of material
that have been used, the reader is referred to Jacobs (69), Travis (128),
•Calhoon (25), Lumley and Calhoon (82), Humpstone (65), Terman (126),
Terman and Merrill (127), Cattell (28), Squire (121), Whitley (136), and
Whipple (135).

It has been found that the type of material used in the test does
definitely affect the results secured. In general, experimental results
indicate that the most difficult material to reproduce is nonsense
syllables, then letters, then digits, sentences and related words (25,
66, 82, 121, 132, 136). Bourdon (18) found that letters were easier
for children to repeat than other materials; all materials were found
to be of the same difficulty for subjects of from 14 to 20. The order
indicates that at least two factors are involved in making some
materials easier than others: familiarity with the material and " asso-
ciability."
If all of the material used produced the same results relatively,
the standing of the individuals in the group would not be affected
"by the type of material used. If the standing of the individuals in
the group is affected by the type of material used, other factors
remaining constant, we should expect a correlation of significantly
less than 1.00 between results secured by use of different materials.
3
That is, the same ability is said to be operating through the media of
•different sense organs and with different materials.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 5
Henmon (60) secured the highest correlation coefficients for memory spans
as ascertained for different types of materials when he reported a coefficient of
.77 between " memory" for syllables and " memory " for numbers, and found
the same correlation between " memory " for nouns and for syllables. Memory
span as tested by nouns and numbers correlated only .20. Abelson (2) secured
intercorrelations of from .34 to .66 for different types of material. Calkins (26)
found that concrete objects produced higher memory span scores than did verbal
stimuli, such as words. Fischler and Albert (42) were the first to treat this
problem statistically. These investigators secured intercorrelations of from
—.38 to .47 for different types of material. As a result of their experiment,
they concluded that immediate memory was apparently not a general ability.
They admit, however, that their results may be due to the fact that the same
subject is attentive in one test, inattentive in another.

In none of these studies was there careful control of all experi-

mental conditions. Thus, additional and more careful work is needed
before conclusions can be drawn.
(3) The Sense or Senses Through Which the Impression is
Received. The sense or senses through which the impression is
received also appear to affect the memory span score as secured by
the clinical or experimental test. The material may be presented
through almost any sense organ or combination of sense organs.
It has been found that in general, for example, the adult will have a
greater memory span with the visual than with the audito-vocal method, but
that the child's score is higher with audito-vocal than with visual presenta-
tion (1, 59, 83, 103). Kirkpatrick (75) reported that visual presentation pro-
duced higher scores for immediate reproduction than did auditory or actual
presentation of objects before the eyes.
The sense organ which receives the impression does, according to
experimental results, make a definite difference in memory span
score, at least in terms of the number of units reproduced. If the
standing of the individual in relation to others is changed by results
on memory span as attained through different sense organs, other
factors remaining constant, correlations between the results should
be significantly less than 1.00. But again, the experiments were not
carefully controlled. Davis (37) found a correlation coefficient of
only .49 between visual and audito-vocal presentation results.
Hao (57) reported the coefficient to be .39 between visual and audi-
tory presentation.
If the material is presented visually, successive or simultaneous presentation
may be used. Miinsterberg and Bigham (93), and Gates (53), found that
adults profited more through simultaneous presentation, although Warden (132)
reported that his college students profited more through successive presenta-
6 ALBERT B. BLANKENSHIP
tion, and when movement was involved in the presentation, scores were even
higher. Hawkins (59) verified the results of Miinsterberg and Bigham (93)
and Gates (S3), and further reported that children secured higher scores
through the use of successive rather than simultaneous presentation.
The memory span apparently increases as the number of sense
organs through which impressions are received increases, except
where distraction may occur, as in the results of Smedley (116).
Jones (74) found that a method of combining as many sense organs as
possible was superior to any other method. This was said to be due to the fact
that there are certain " visual," " auditory," and " motor " types of individuals.
When all the possible sense organs are stimulated, each subject has the
fullest possible advantage in the method of presentation. Miinsterberg and
Bigham (93) found that a series presented to 2 senses at the same time is
much more easily reproduced than if given only to sight or to hearing.
Smedley (116) concluded that the audito-visual-articulatory and the audito-
visual-hand-motor memory were superior to visual presentation, which was
superior to the audito method, but Chambers (31) could not substantiate this
order.
Nichols (96) early demonstrated the possibility of using the tactual recep-
tors for testing memory span, though Nichols had little idea of such an appli-
cation of his work. The fact that memory span can be tested through the
medium of any sense modality makes the Knox Cube Test a particular type of
memory span test. In the Knox Cube Test, 4 blocks in a row are tapped by
the examiner in a given order; the subject is then asked to tap the blocks in
the same order. Obviously this comes under the functional definition of memory
span, being simply a new method of presenting the stimuli. Davis (37) recog-
nized this test as a type of memory span when he ran correlations with the
results of the Knox Cube Test and results of memory span for digits.
(4) Summary. Results indicated in the 2 preceding sections
show that substantially different spans are secured depending upon
the type of material used, and the sense or senses through which the
impressions are received. From all the evidence available, however,
it would appear that memory spans for different types of material
may be specific spans, rather than different aspects of a general span.
Likewise, it appears quite probable that the memory span for each
sense organ or combination of sense organs is a specific span. But
the whole question is far from settled and is open for a real experi-
mental attack, since it is more than possible that differences are due
to inaccuracies in the methods rather than intrinsic differences in
the mental processes involved.
For example, if the results obtained by methods using different
materials could be freed of differences in the subjects' acquaintance
with the materials, there might be no differences in relative results.
This possibility is to be discussed further under a subsequent heading.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 7
And again, if subjects who had practiced equally with the differ-
ent sense organs could be obtained, results secured through different
sense organs might give the same relative rankings for individuals.
But there is a strong possibility that the imaginal endowment, as well
as the imaginal type, would still affect the memory span, as Jones
(74) suggests.
IV. FACTORS WHICH AFFECT MEMORY SPAN
There are a number of factors which definitely affect memory
span; the effects of practically all of these factors have been investi-
gated in statistical and experimental studies. Some of the factors
are extrinsic, or present in the testing situation itself. These factors,
if not carefully controlled, cause the memory span test to be statis-
tically unreliable. Other factors are intrinsic in the individual, and
it is these factors which are the basis of " true " memory span.
Though numerous factors affect memory span, the test is one
that shows surprisingly high reliability. Results obtained by different
investigators show that the reliability coefficients for memory span
may be as low as .28 (15), or as high as .93 (22). Table I sum-
marizes the reliability coefficients secured by different investigators
for the more common methods of testing memory span.
TABLE I
RELIABILITY COEFFICIENTS REPORTED FOR MEMORY SPAN
Audito-Vocal Visual for Audito-Visual
Investigator for Digits Digits for Digits
E. B. Bolton(lS) 28
D. Mitchell (90) .44, .47 *
Hao(57) 52 .83
Garrett (49) .80 .68
Davis (37) 74 .84
Wyatt(141) .76
Burt (22) .70, .93 5
Abelson (2) .73, .70 c
Reliability coefficients for other types of memory span reported in
the literature vary from .70 (2) to .81 (141). Both the figures sum-
marized in the table, and the data indicated in the preceding para-
graph lead to an important conclusion regarding the use of memory
span tests. The range of figures indicates that the extrinsic factors
can probably be controlled carefully enough to make the test a
reliable one.
4
Two figures secured by use of two methods of scoring.
6
Burt (22) secured the first figure with elementary school students, the
latter with preparatory school students.
6
Abelson's first figure was secured with girls, the second with boys.
8 ALBERT B. BLANKENSHIP
A. Extrinsic Factors
(1) Characteristics of the Material That is Used. The char-
acteristics of the material used will definitely affect the memory
span score.
If, for example, the material is all closely related, it will be much more
easily reproduced. This relationship of the material is called by Calhoon (23)
the " coefficient of associability." In the use of digits, the figures must be so
placed that none are in their natural or reverse order. There must be an avoid-
ance of rotation, or of any numbers suggesting addition, subtraction, division,
or multiplication (Brotemarkle, 20). Binet and Simon (11) state that no
numbers which follow one another must be used beside each other. T. L.
Bolton (16) and Bourdon (18), in their experimentation, were sure to ascertain
that no digit came in its accustomed order, and that no digit was repeated.
Terman (126) was not so careful in this respect in his 1915 revision of the
Binet Test, but has taken greater care in his 1937 revision of the scale (127).
Xilliez (143) analyzed the effect of the relation of the digits to one another,
and noted that a negative interval (the interval is the difference between 2 digits
which follow one another; it is positive if the second is larger, negative if the
first is larger), is inferior, in terms of recall, to a positive interval. To sum-
marize, the units of the series must not be presented in a manner that would
facilitate groupings through the apperceptive background of the subject.

In addition, the units of the material must not be presented in

groups. If the visual method is used, the material should be pre-
sented either one unit at a time (successive presentation), or all units
at the same time (simultaneous presentation), for grouping would
make it too simple for the subject to secure a memory span above
his "true " one. Brotemarkle (20) and others emphasize the impor-
tance of the control of grouping. Chamberlain (30) has experi-
mentally demonstrated that recall is stronger when the objects are
presented in groups. However, even when grouping is eliminated
in the presentation of the material, subjective grouping often occurs.
The material used should have approximately the same degree
of familiarity for all subjects. Calhoon (23) and Whitley (136) both
stress the fact that apperceptive background should be equalized for
all subjects, as far as possible.
All of the subjects should have the same degree of familiarity with
the items in the series. Do not, for example, test a child by the
digit method if that child has never been taught numbers, for all
available norms have been secured with the use of subjects acquainted
with numbers.
(2) Rhythm of the Presentation of the Material. Closely related
to the problem of presenting the stimuli in groups, is the presentation
of the stimuli in rhythmic fashion. Most investigators point out that
 MEMORY SPAN: A REVIEW OF THE LITERATURE 9
the stimuli used in testing memory span should be presented with
as little rhythm as possible (11, 18, 119, for example).
Probably the only experimental study of the effect of varying rhythm on
memory span has been undertaken by Adams (3). Adams varied the rhythm
in presenting a series of digits by using trochaic, iambic, dactylic, anapestic or
amphibrachic rhythm, or no rhythm at all. He reported that his subjects
(elementary psychology students at the University of Michigan) had higher
memory spans in general when rhythm was used than when no rhythm was
used. The effect of the different types of rhythm depended upon the sex of the
individual, the females doing best with anapestic rhythm, the males with
dactylic.
The fact that the introduction of rhythm into the presentation of the series
of units does increase memory span is further verified by the results of those
investigators interested in the " rhythmic span," in which the units are pre-
sented in rhythmic fashion (see 122, 124, for example). The effect of rhythm
is to group the units in the series, again enabling the individual to secure a
span higher than his " true " one.

(3) Rate of Presentation of the Stimuli. The speed with which

the stimuli are presented has an effect on the memory span score
attained. Terman (126) and other psychologists set the best rate
of presentation of digits at a rate of slightly faster than 1 per second,
while Lightner Witmer, in instructing clinical psychologists at the
University of Pennsylvania, expressed his belief that the " natural
rate of discharge " (the speed best adapted to the individual) should
be used.
Actual experimental investigation also indicates that the speed of presenting
the stimuli affects the score. Peatman and Locke (100) experimentally showed
that the best rate of presentation for digits by either the audito-vocal or visual
method was one digit per two-thirds of a second to one digit per second.
In the auditory digit test, Brotemarkle (20) believes that a rapid increase in
the rate of presentation will result in an increase in score. Lumley and
Calhoon (82) found that a decrease of speed enabled children of the seventh
and eighth grades to raise their scores, but that in the other grades tested
(third to twelfth) there was no consistent effect on performance. Other
experimenters have found that a faster rate of presentation adversely affected
memory span performance (56, 107). One investigator (Bergstrom, 6) reported
that rate of presentation of the stimuli had no effect on the attained memory
span.

Once again, a conclusion about the effect of the variable cannot

be reached. Different research workers make various reports. More
•carefully controlled and standardized work is essential.
(4) The Method of Scoring the Responses. The method of
scoring the responses also has an effect upon the apparent memory
10 ALBERT B. BLANKENSHIP
span of the individual. Variations in scoring are common; scarcely
two investigators have scored alike.
In the audito-vocal memory span for digits, for example, Terman (126)
gave the individual 2 or 3 trials on any particular series, depending upon its
length, and the subject was given credit for that length if one of the series was
reproduced correctly. Starr (122) gave credit if 2 of 4 series were repro-
duced correctly at a given length. Humpstone (65) gave one trial at each level
where the series was arranged in lengths varying from 3 to 10 digits. Credit
was given for the longest series correctly reproduced. M. H. Young (145)
showed that the number of trials given affects memory span attained. When a
child was given 3 chances instead of 2 (with one series necessarily correct for
credit), 55% of the subjects increased their span by one.
In the determination of reverse memory span, Starr (122) gave 4 chances
at each length, and 2 of the 4 had to be correct reproductions for credit to be
given. Terman (126) gave 2 or 3 chances at each level, and only one had to-
be correct. In the visual memory span for digits, Humpstone (65) gave only
one chance at a series of given length, and to get credit, that series had to be
absolutely correct.
Most investigators take the point of view that an incorrect series should
not be scored at all. As Bergstrom (6) points out, if errors in a series longer
than the span attained are scored, the true memory span is not ascertained.
Other investigators feel, however, that all of the reproductions should be
considered (55, 133).
Krueger and Spearman (77) take account of errors in their novel technique
of scoring. They correlate the subject's reproduced series with the original
stimulus series by use of Spearman's " footrule method." Thus the greater the
error of the subject, the lower will be the correlation coefficient. Other methods
of scoring errors are many and complicated, but these schemes and techniques
will not be discussed in this paper.
(5) Fatigue of the Subject. Fatigue may be another extrinsic
factor affecting memory span performance. Though the few investi-
gators mentioning the effect of fatigue on memory span do not dif-
ferentiate between mental and physiological fatigue and boredom,
this does not immediately exclude their observations from consider-
ation, though it does make them much less valuable. Hao (57) and
Whitley (136) both believed fatigue to be a factor in their results,
but Smedley (116) probably delayed experimental work on the prob-
lem when he pointed out that if one attempted to test the effect of
fatigue, the subjects are apt to gain more through practice than they
lose through fatigue.
As throughout the field of memory span investigation, more
careful work is needed before the effects of fatigue can be conclu-
sively shown.
(6) Time of Day. The time of day apparently is another ex-
trinsic variable which produces differences in memory span. From
 MEMORY SPAN: A REVIEW OF THE LITERATURE 11
the available data, there is no way of telling whether the observed
variations in memory span during the day are due to mental or to
physiological fatigue, or boredom, or to some other factors not even
considered as a possible cause. It is for this reason that the variable
is considered under a separate heading.
Marsh (84) found wide individual variations in the time of day at which
greatest efficiency appeared in memory span performance, while Winch (138)
found that efficiency was greatest in the forenoon. Gates (SO, 52) substanti-
ated Winch's results, and Laird (79) extended them to conclude that the per-
formance reaches its low point about 10:00 P.M., when there is an "' end spurt."

Though the causes of such variation are not clear, the implication
for clinical psychologists is clear. In order for the subject to per-
form in his best possible manner, the test should be made in the
forenoon.
(7) The Attitude of the Subject. Since the attitude of the sub-
ject is another important factor ordinarily within the control of the
experienced examiner, it too is listed here as an extrinsic factor.
Too many excellent chapters have been written on the technique of
establishing rapport with the subject for the present writer to go
into detail. For such a discussion, the reader is referred to almost
any current text on intelligence testing.
It will be sufficient to mention work in which the attitude of the
subject has been found to have a definite effect on the memory span
attained. Bronner, Healy, Lowe, and Shimberg (19) and Hao (57)
report that the personal attitude of the subject definitely affects
results, and Squire (121) found the use of pictures effective in testing
the memory span of children, for it increased their interest.
(8) Distraction. Naturally enough, one would expect that the
greater the distraction present in the situation, the poorer would be
the performance of the individual, and this is actually the case (92,
117, 134). The reason for this effect is apparent. Inasmuch as
attention is one of the processes involved in the successful functioning
of memory span, if the processes of attention are directed towards
some other stimulus, they cannot operate effectively in the memory
span function. Distractions must be kept at a minimum for reliable
results, as Lumley and Calhoon (82) indicate.
(9) Practice. Practice on the part of the individual is another
extrinsic factor affecting the apparent length of the memory span.
Although it is now commonly assumed that the memory span is
a congenital ability (65, 116), investigations reveal that a temporary
increase in memory span score will result from practice.
,12 ALBERT D. DLANKENSHIP
Gundlach, Rothschild and Young (56) and Ide (67) found that some indi-
viduals' memory span scores were increased, those of others not visibly affected
by practice. Winch (137) and T. L. Bolton (16) reported marked improve-
ment with practice on the part of their subjects. Foster (43), experimenting
with 6 different materials, stated that there was a definite practice effect in his
subjects, but that the gain was specific, and limited to the particular type of
material used. This is probably further evidence that memory span is not a
general ability, but is specific for different types of material.
The greatest practice effects on memory span thus far have been demon-
strated by Martin and Fernberger (85), who discovered that the memory span
of one individual increased 47%, that of another 36%, after periods of practice
spread over several months. Foster's (43) subjects gained from 6% to 44%.
Dallenbach (35) and Gates (54) were interested in determining the per-
manence of the practice effect reported. Dallenbach, after training subjects
for a period of 17 weeks, observed a practice effect 41 weeks after the drill
had been discontinued. Gates trained a group of subjects over a period of
78 days (spread over 5 months) and at the end of training, this group had
raised its average memory span by 2 digits. After 4^2 months of no practice,
the group had fallen back to its original average.
Reed (108), however, claims that practice effects are negligible, and
Whipple (134) experimentally found that if adaptation and assimilative devices
are held constant, there is no practice effect. We must conclude, nevertheless,
that practice does have an effect on memory span score as it is now commonly
obtained by experimental or clinical methods. The reasons again are fairly
obvious, and are so well discussed by Foster (43) that a detailed discussion is
unnecessary. Foster believes gains to be due to (1) confidence and effort,
(2) familiarity with the material, (3) learning to distribute the attention effec-
tively, and (4) efficient methods of work and organized procedure.
(10) Subjective Grouping of the Units in the Series. It has
already been noted that presenting the units in the series of stimuli
by any method of grouping or rhythm will enable the subject to
secure a higher memory span than he would otherwise have. Often
the subject himself is entirely responsible for grouping the units, and
may thus increase his apparent memory span. In Martin and Fern-
berger's study (85) it was noted that any memory span over 5 was
secured through subjective grouping of the units. Oberly (98)
found that the memory span limen, as indicated by grouping on the
part of the subjects, was from 6 to 13.8 units.
It is certain that subjective grouping will increase the memory
span of the individual and thus contribute to the unreliability of the
method. The many cases of unusual immediate memory are prob-
ably explained by such grouping, though in the case of some indi-
viduals, this grouping is merely a matter of associating some of the
units in the series with others.
F, D. Mitchell's report of Inaudi(91; also reported by Binet, 8), who
correctly repeated 42 digits on one occasion, and his report of the blind Swiss
 MEMORY SPAN: A REVIEW OF THE LITERATURE 13
who repeated ISO digits, are of this order. Even the famous Dr. Finkelstein,
who appeared on Ripley's "Believe It or Not" program (111) could not have
repeated the IS digits he did on that occasion except through some method of
grouping or meaningful association.

(11) Temporary Pathological Condition of the Individual. Some

temporary pathological conditions of the individual will detrimen-
tally affect memory span score. If the pathological condition is a
permanent one, it may then be classed as an intrinsic factor, beyond
the control of the examiner.
Kohnsky (76), controlling practice effects, found that pupils, several
months after having dental treatment, increased their memory span scores.
Paulsen (99) found, after equalizing for practice effects, that subjects who had
been suffering from intestinal toxemia increased their memory span scores after
treatment for the condition. These results, though apparently definite, need
confirmation before final conclusions can be drawn. If the results are con-
firmed, we are probably justified in assuming that such temporary states have
some adverse effect on the processes involved in memory span, making them less
efficient Another temporary pathological condition of the individual is that of
hypnosis. P. C. Young (146) found no differences in "digit span" or
" memory span" under hypnosis from that in the waking state, but his terms
are not well denned and his conclusions thus have little significance.

(12) Effect of Drugs. Drugs may also produce a temporary

condition which will affect memory span results. Since drugs pro-
duce a toxic state, the condition could well be included under the
previous heading. Froeberg (45) and Hull (62) found a loss in
memory span performance in non-smokers after smoking. Hull
found that the habitual smokers showed a very small loss in efficiency
due to smoking. From these facts Hull decided that habituation ap-
peared to have produced a partial tolerance for tobacco with regard
to its effect on the memory span. The explanation of the effect of
drugs on memory span is obvious. The toxic states produced
adversely affect mental processes, and through so doing, decrease the
memory span score attained.

B. Intrinsic Factors
In addition to the factors here called " extrinsic" (but only
extrinsic in that they are largely within the control of the examiner,
and if not properly controlled, tend to produce an erroneous memory
span), there are also certain "intrinsic" factors affecting memory
span. It is these in which the psychologist is primarily interested.
These intrinsic factors are those within the individual which work
to produce his " true " or permanent memory span.
14 ALBERT B. BLANKENSHIP
(1) Age oj the Individual. The age of the individual is a factor
which definitely affects memory span.
Memory span has been found to increase with age by a number of investi-
gators (8, 19, 27, 38, 40, 42, 56, 63, 69, 80, 121, 137). Norms for various age
levels have been secured by McCaulley (86), Lumley and Calhoon (82), M.
Murphy (95), Starr (123), Smedley (116), and Terman (126, 127).

It should be pointed out that if the mental age of the individual

does not increase, the memory span will not. So far as is known,
memory span increases along with intelligence up to a similar age.
At what age does memory span cease to develop?
Carpenter (27), in using subjects from 6 to 14 years of age, reported a
consistent improvement from year to year. Fischler and Albert (42) found
an increase of memory span to adulthood. Kuhlmann (78) claimed that
memory span increased up to maturity, but neither Fischler and Albert nor
Kuhlmann interpret their terms. Wessley (135, p. 176) found that the maximal
memory span occurred at the age of 12 to 14, while Smedley (116) and
Chambers (31), although finding a general increase with age, found no par-
ticular age at which memory span was maximal. Bourdon (18) reported that
maximal efficiency occurred at the age of 14. Hao (57) placed the age at 13
or 14 for girls, 15 or 16 for boys.

A few investigators claim that memory span increases to a point

somewhere between the sixteen- and twenty-six-year level, though a
large number of workers believe that memory span remains constant
after the individual reaches a point somewhere between 12 and 16
years. Once again, the investigators have used such diverse methods
of administration and scoring, and such different material, that the
results are scarcely comparable. It is not surprising that no definite
conclusion can be reached regarding this and other points in question.
(2) Sex of the Individual. Sex may be another intrinsic factor
affecting the memory span; there is some disagreement on this point.
Burt (22), T. L. Bolton (16), Gates (51, 52), Kirkpatrick (75), and
Woolley (140) all reported consistent superiority of the females in memory
span tests. Gundlach and his coworkers (56), testing memory span with
flashing lights, observed only a very slight superiority of the females over the
males. Lodge and Jackson (81) and Travis (128), however, using prose pas-
sages to test memory span, report the superiority of females over males.
No significant sex differences in memory span have been observed in chil-
dren of kindergarten age (67), at the six-year level (38), in primary school
children (30, 137), and in children in general below the age of 15 (31).
Fischler and Albert (42), testing for audito span with digits, consonants, and
phrases as material, and for visual span with forms and pictures, found no
significant sex differences, either in children or in adults.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 15
Adams (3) reported a slight superiority of men over women in forward
.memory span for digits, presented by the audito-vocal method. Chambers (31)
.noted a superiority of males over females above the age of IS, and
Watkins (133) reported a superiority of boys over girls in memory span
ability. Dallenbach (35, 36) found that when mental age was held constant,
.males consistently surpassed females in " visual apprehension."

Thus we can reach no conclusion as to the role of sex in memory

.span. All we can do at this time is to note that sex may be a factor.
Again, the methods used are so different that results vary.
(3) Race of the Individual. Recent investigations indicate that
the race of the individual is another factor which may affect memory
span. Apparently the Chinese are superior to the whites, who may
be in turn superior to the Negroes. Data concerning the memory
.span of other racial groups have not yet been reported. Hao (57),
-and Pyle (106) observed that Chinese children were superior to
white children in immediate memory. Pyle (105) also found that
negro children were definitely inferior to white children in rote
memory. Clark (33), however, observed his negro subjects to be
•superior to the whites. The results here are purely exploratory, and
need further confirmation, but at least there is some evidence that
there are race differences in memory span.
(4) Permanent Pathological Condition of the Individual. When
the physical condition of the individual becomes permanently modi-
fied, the memory span has been found to be lower than that for a
normal individual.
Epilepsy is such a condition; W. G. Smith (117), as early as 1905, reported
the inferiority of a group of epileptic subjects to normal subjects in memory
span, while Niude (97) substantiated this conclusion with a study of 2,000
epileptics.
Smith (119) also reported, in another article, that normal subjects were
definitely superior in memory span to those in a pathological (insane) group.
Pintner and Paterson (102) found that deaf children, as a group, had abnor-
mally poor memory spans. They concluded that this was due to the lack of
auditory experience. Bond and Dearborn (17), testing auditory "memory"
for different types of material, reported that normal subjects were distinctly
superior to the blind subjects they tested from the Perkins Institution. But
Hayes (58) failed to substantiate this report when it was found that the blind
subjects were superior to normal individuals in memory for auditory digits,
"but that for other types of material and methods of presentation, superiority
of the blind or the normal group varied with the age group tested.

Apparently a pathological condition of the individual may operate

to improve memory span; at least certain pathological cases demon-
strate unusual memory spans, whether or not this is due to the
16 ALBERT B. BLANKENSHIP
pathological condition. Barr (4) discusses Kitri, an " idiot savant "
with echolalia, who repeated, after the first hearing, words and accents
correctly in English, French, German, Spanish, Italian, Japanese,
Latin, Greek, and Norwegian. Tredgold (129) tells of an imbecile
who could repeat verbatim a newspaper he had just read.
These cases are not demonstrations of " true " memory span, for
the individuals apparently reproduce the materials through some
form of a memorial image.

V. CORRELATION OF MEMORY SPAN AND INTELLIGENCE AND WHAT

IT MEANS FOE THE MEMORY SPAN AS A
CLINICAL TEST
(1) Relation Between Forward M.emory Span and Intelligence.
Earliest observations of the relation between memory span and intel-
ligence were made by Jacobs (69), who noted that pupils who stood
high in class tended to have high memory spans, while Smedley (116)
corroborated this report. Early experiments with feebleminded indi-
viduals pointed to the same fact—that memory span was directly
related to intelligence (48, 72).
Early estimates of intelligence placed those with high memory
spans near the head of the list (18, 22, 78, 137). Later investigators
on the subject made use of the correlation coefficient and more objec-
tive measures of intelligence. Table II summarizes the coefficients
secured between memory span and intelligence by various investi-
gators using different types of forward memory span.

TABLE II
CORRELATIONS REPORTED BETWEEN MEMORY SPAN AND INTELLIGENCE
Auditory Presentation
Investigator Digits Sentences Commissions Nouns
Abelson(2)7 .45,-18 .53, .65 .18, .19
Clark (33) 03
Garrett(49) 21
Wissler (139) 8 16
Visual Presentation
Digits Letter Square Nonsense Syllables
Garrett (49) 18
Wyatt(141) .18 .59
7
Abelson's first figure represents results with girls, the second figure his
results with boys.
8
Wissler correlated auditory memory for digits with class standing rather
than with intelligence.
 MEMORY SPAN: A REVIEW OF THE LITERATURE Y7
This table shows that there is some relation between memory span
and intelligence, and other results further indicate this relationship.
A contrast of results, probably reflecting differences between the 2 sets of
subjects in intelligence, are those of Terman (126) and McCaulley (86).
Terman sets the audito-vocal memory span for the normal six-year-old at
5 digits, whereas the backward children tested by McCaulley secured a modal
span of only 4 digits. Starr (123) reported that the retarded, sub-normal, and
low defective children all tested below normal in memory span. Squire (121)
also found retarded children to be inferior in memory span. Bingham (13) and
Humpstone (63) found that college students in general had higher memory
.spans than average adults, as tested by other investigators.
All of these findings indicate a definite relation between memory
span and intelligence. But at the present time, results are so varying
in nature that the true degree of correlation between the. two is
impossible to predict. Terman, both in his original revision of the
Binet Test (126) and in his recent revision (127) feels certain
enough of the high degree of relationship to include memory span
tests throughout the scale.
(2) Relation Between Reverse Memory Span and Intelligence.
Bobertag (14), in 1911, was the first to suggest the reverse memory
span test. Little work has been done up to the present time in
making use of the reverse span, except for placement in the 2 Stanford
revisions of the Binet Test (126, 127). Fry (46) has been the only
worker to run correlations between the reverse memory span and
intelligence. He secured a coefficient of .75 for reverse audito-vocal
digit span and intelligence (as measured by Army Alpha). This is
higher than any correlation secured between forward span and
intelligence.
(3) Value of the Memory Span Test as a Diagnostic Measure.
The results of a memory span test, then, are ordinarily indicative of
the level of intelligence of an individual. Binet and Henri (10),
A. M. Jones (73), Ninde (97), Learning (80), and others place
memory span ability at the base of all intellection. Starr (123)
states that memory span " expresses the index of proficiency of all
the mental competencies involved." Ninde says, " It goes without
saying that a certain degree of associability is essential to all intelli-
gent behavior and it is of special value in the development of the
intellect" (97).
Most research and clinical workers agree that the value of the
memory span test lies in its clear differentiation of the upper and
lower groups of the distribution (Brotemarkle, 20; Starr, 123;
18 ALBERT B. BLANKENSHIP
Chambers, 32). There is too much overlapping at the middle and
at the extreme upper end of the distribution of age and diagnosis.
Most clinical workers place more value in low spans than in high
spans. Opinions of clinicians in regard to " critical spans " are of
interest. In the forward memory span, a normal child of 5 or 6
should have a span of 2 or more (Easby-Grave, 38). A forward
span of 5 is taken as a prerequisite to do high school work, while an
even higher span is probably a prerequisite to do more advanced
work (Learning, 80). Other "critical spans" are listed by
Sherman (114), Ninde (97), and McCaulley (86).
The memory span test as an indication of the individual's intelli-
gence has several clinical advantages. Ninde (97) points out that it
is simple and easy to administer. It does not place an emphasis on
language ability, nor is it a long, extended test which is apt to tire
the individual. Witmer believes that it is one of the most significant
clinical tests, and Starr (122) states that " it is without doubt one
of the most valuable tests employed for diagnostic purposes."
But its very simplicity is one of the dangers of the memory span
test. The inexperienced examiner is apt not to follow specifically
the particular directions which he is using. In addition, the scoring
must be done precisely according to the method used in securing the
norms which the worker is using. There are so many additional
extrinsic factors affecting memory span that if careful clinical con-
ditions are not observed, the results may be meaningless.
Another danger is that the investigator may place too much sig-
nificance on the memory span test. Bronner, Healy, Lowe, and
Shimberg (19) think that the importance of the memory span test
has been greatly overemphasized. Of course a memory span test
alone should never be used for diagnosis; the results on the memory
span test are merely suggestive, and should always be supplemented
by other test results and by qualitative observation.

VI. SUMMARY
Though 146 references are listed in the bibliography, it is appal-
ling to note how little real knowledge there is in the field of memory
span. Practically all of the questions raised in the present paper
have to remain unanswered; many researches have been undertaken,
but few facts have been proved.
It has been pointed out throughout the paper that the primary
causes for this state of affairs are the widely diverse methods of
administering the test, the many kinds of materials used, the different
 MEMORY SPAN: A REVIEW OF THE UTERATURE 19
groups of subjects used, the methods of scoring, etc. The question
of whether or not memory span is a specific ability is essential; the
effect of other factors cannot be answered until this is determined.
For if memory span is a specific ability, it seems obvious that investi-
gators using different materials and methods can expect only to get
different answers to the same questions.
Probably the one thing most experimenters do agree on is a
functional definition of memory span. But for other questions there
are all sorts of answers. We do not know whether memory span is
a specific or a general trait. We are sure that memory span is
affected by certain extrinsic and by certain intrinsic factors, but we
are not sure just what to include under each list, since all sorts of
results have been claimed for any one variable.
Oddly enough, however, the test has been shown to have a fairly
high reliability, and clinical investigators think enough of it seldom
to omit it in an examination. It is favored by clinical investigators
because of its close relation to intelligence (which has been fairly
definitely shown), its simplicity, its brevity, and its lack of emphasis
on language ability.
But, nevertheless, the whole field is wide open for a real experi-
mental attack, for there is not a single aspect of the subject which is
a closed chapter.
BIBLIOGRAPHY
1. ABBOTT, E. E., Memory Consciousness in Orthography. Psychol. Monog.,
1909, 11, No. 1, 127-158.
2. ABELSON, A. R., The Measurement of Mental Ability of "Backward"
Children. Brit. J. Psychol, 1911, 4, 268-314.
3. ADAMS, H. F., A Note on the Effect of Rhythm on Memory. Psychol.
Rev., 1915, 22, 289-298.
4. BARR, M. W., Mental Defectives. Philadelphia: P. Blakiston's Son &
Co., 1913. Pp. x+368.
5. BENNETT, F., The Correlation Between Different Memories. / . Exper.
Psychol, 1916, 1, 404-418.
6. BERGSTROM, J. A., Effect of Changes in the Time Variables in Memoriz-
ing, Together with Some Discussion of the Technique of Memory
Experimentation. Amer. J. Psychol, 1907, 18, 206-238.
7. BIGHAM, J., Memory. Psychol. Rev., 1894, 1, 453-461.
8. BINET, A., Introduction a la psychologie expeiimentale. Paris: Bailliere,.
1894. Pp. 146.
9. BINET, A., Attention et adaptation. Ann. Psychol, 1899, 6, 248-404.
10. BINET, A., and HENRI, V., La memoire des mots. Ann. Psychol, 1894,
1, 1-23.
11. BINET, A., and SIMON, T., Methodes nouvelles pur le diagnostic du niveau
intellectuel des anormaux. Ann. Psychol, 1905, 12, 191-244.
20 ALBERT B. BLANKENSH1P

12. BINET, A., and SIMON, T., L'intelligence des imbeciles. Ann. Psychol.,
1909, 15, 1-147.
13. BINGHAM, W. V., Some Norms of Dartmouth Men. / . Educ. Psychol.,
1916, 7, 129-142.
14. BOBERTAG, O., Ueber Intelligenzpriifiing (nach du Methode von Binet
und Simon). Zsch. f. angew. Psychol, 1911, 5, 105-203; 1912, 6, 495-
538.
15. BOLTON, E. B., The Relation of Memory to Intelligence. / , Exper.
Psychol., 1931,14, 37-67.
16. BOLION, T. L., The Growth of Memory in School Children. Amer. J.
Psychol, 1892, 4, 362-380.
17. BONP, N. J., and DEARBORN, W. F., The Auditory Memory and Tactual
Sensibility of the Blind. / . Educ. Psychol, 1917, 8, 21-26.
18. BOURDON, B., Influence de l'age sur la memoire immediate. Rev. Philos.,
1894, 38, 148-167.
19. BRONNER, A. F., HEALY, W., LOWE, G. M., and SHIMBERG, M. E.,
A Manual of Individual Mental Tests and Testing. Boston: Little,
Brown & Co., 1927. Pp. x+287.
20. BROTEMARKLE, R. A., Some Memory Span Test Problems. Psychol. Clin.,
1924, IS, 229-258.
21. BURNHAM, W. H., Memory, Historically and Experimentally Considered.
Amer. J. Psychol, 1888-1889, 2, 39-90, 225-270, 431-464, 568-622.
22. BURT, C, Experimental Tests of General Intelligence. Brit. J. Psychol,
1909, 3, 94-177.
23. CALHOON, S. W., Influence of Length of Lists upon Ability Immediately
to Reproduce Disconnected Word Series Auditorially Presented. / .
Exper. Psychol, 1934, 17, 723-738.
24. CALHOON, S. W., Influence of Syllabic Length and Rate of Auditory
Presentation on Ability to Reproduce Disconnected Word Lists. / .
Exper. Psychol, 1935, 18, 612-620.
25. CALHOON, S. W., A Comparison of Ability to Reproduce One-Syllable
Words and Digits Auditorially Presented. / . Exper. Psychol, 1935,
18, 621-632.
26. CALKINS, M. W., A Study of Immediate and of Delayed Recall of the
Concrete and of the Verbal. Psychol. Rev., 1898, 5, 451-456.
27. CARPENTER, D. F., Mental Age Tests. J. Educ. Psychol., 1913, 4, 538-544.
28. CAT-TELL, J. McK., Mental Tests and Measurements. Mind, 1890, IS,
373-380.
29. CATTELL, J. MCK., Tests of the Senses and Faculties. Educ. Rev., 1893,
5, 257-265.
30. CHAMBERLAIN, A. H., A Memory Test with School Children. Psychol.
Rev., 1915, 22, 71-76.
31. CHAMBERS, W. G., Memory Types of Colorado Pupils. / . Philos., Psychol.,
& Sci. Method, 1906, 3, 231-234.
32. CHAMBERS, W. G., Individual Differences in Grammar Grade School
Children. / . Educ. Psychol, 1910, 1, 61-75.
33. CLARK, A. S., Correlation of the Auditory Digit Memory Span with
General Intelligence. Psychol. Clin., 1923, 15, 259-260.
34. COLLINS, M., Some Observations on Immediate Color Memory. Brit. / .
Psychol, 1932, 22, 344-352.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 21

35. DAIXENBACH, K. M., The Effect of Practice upon Visual Apprehension

in School Children. / . Educ. Psychol., 1914, 5, 321-334, 389-404.
36. DALLENBACH, K. M., The Effect of Practice upon Visual Apprehension
in the Feebleminded. / . Educ. Psychol, 1919, 10, 61-82.
37. DAVIS, E. A., Knox Cube Test and Digit Span. / . Genet. Psychol., 1932,
40, 234-237.
38. EASBY-GRAVE, C , Tests and Norms at the Six-Year-Old Performance
Level. Psychol. Clin., 1922, 15, 261-300.
39. EBBINGHAUS, H., Ueber eine Neue Methode zur Priifung Geistigen
Fahigkeiten und ihre Anwendung bei Schulkindern. Zsch. f. Psychol.
u. Physiol. d. Sinnes., 1897, 13, 401-457.
40. FAKSON, M. R., A Report on the Examination of 100 6B Children in
Philadelphia Schools. Psychol. Clin., 1928, 40, 128-152.
41. FERNBERGER, S. W., Elementary General Psychology. Baltimore: Williams
& Wilkins, 1936. Pp. xi+445.
42. FISCHLER, D., and AT.BF.RT, I., Contribution a l'etude des tests de memoire
immediate. Arch, de Psychol., 1929, 21, 293-306.
43. FOSTER, W. S., The Effect of Practice upon Visualizing and upon the
Reproduction of Visual Impressions. / . Educ. Psychol., 1911, 2, 11-22.
44. FRANZ, S. I., and GORDON, K., Psychology. New York: McGraw-Hill
Book Co., 1933. Pp. xvii+494.
45. FROEBERG, S., Effects of Smoking upon Mental and Motor Efficiency. / .
Exper. Psychol., 1920, 3, 334-346.
46. FRY, F. D., The Correlation of Reverse Audito-Vocal Memory Span with
General Intelligence and Other Mental Abilities of 308 Prisoners in the
Eastern State Penitentiary of Pennsylvania. Psychol. Clin., 1930, 19,
156-164.
47. FULLER, H. H., The Art of Memory. St. Paul: National Publishing Co.,
1898. Pp. ix4.-481.
48. GALTON, F., Supplementary Notes on " Prehension" in Idiots. Mind,
1887, 12, 79-82.
49. GARRETT, H. E., The Relation of Tests of Memory and Learning to Each
Other and to General Intelligence in a Highly Selected Adult Group.
/ . Educ. Psychol, 1928, 19, 601-613.
50. GATES, A. I., Diurnal Variations in Memory and Association. Univ.
Calif. Publ in- Psychol, 1916, 1, 323-344.
51. GATES, A. I., Correlations and Sex Differences in Memory and Substi-
tution. Univ. Calif. Publ. in Psychol, 1916, 1, 345-350.
52. GATES, A. I., Variations in Efficiency During the Day, Together with
Practice Effects, Sex Differences, and Correlations. Univ. Calif. Publ.
in Psychol, 1916, 2, 1-156.
53. GATES, A. I., The Mnemonic Span for Visual and Auditory Digits. / .
Exper. Psychol, 1916, 1, 393-403.
54. GATES, A. I., The Nature and Limit of Improvement Due to Training.
27th Yrbk. Nat. Soc. Stud. Educ, 1928, Part I, 441-460.
55. GUILFORD, J. P., and DALLENBACH, K. M., The Determination of Memory
Span by the Method of Constant Stimuli. Amer. J. Psychol, 1925, 36,
621-628.
56. GTJNDLACH, R., ROTHSCHILD, D. A., and YOUNG, P. T., A Test and
Analysis of " Set." / . Exper. Psychol, 1927, 10, 247-280.
22 ALBERT B. BLANKENSHIP

57. HAO, Y. T., The Memory Span of 600 Chinese School Children in San
Francisco. Sch. & Soc, 1924, 20, S07-S10.
58. HAYES, S. P., The Memory of Blind Children. Teach. Forum, 1936, 8,
55-59, 71-77.
59. HAWKINS, C. J., Experiments on Memory Types. Psychol. Rev., 1897, 4,
289-294.
60. HENMON, V. A. C, The Relation Between Mode of Presentation and
Retention. Psychol. Rev., 1912, 19, 79-96.
61. HOLMES, O. W., Mechanism in Thought mid Morals. Boston: Osgood,
1871. Pp. 101.
62. HULL, C. L., The Influence of Tobacco Smoking on Mental and Motor
Efficiency. Psychol. Monog., 1924, 33, No. 150. Pp. 161.
63. HUMPSTONE, H. J., Some Aspects of the Memory Span Test: A Study in
Associability. Philadelphia: Psychol. Clin. Press, 1917. Pp. i-f-30.
64. HUMPSTONE, H. J., The Analytical Diagnosis. Psychol. Clin., 1919, 12,
171-173.
65. HUMPSTONE, H. J., Memory Span Tests. Psychol. Clin., 1919, 12, 196-
200.
66. HUNTER, W. S., Learning: II. Experimental Studies of Learning. In
Murchison, C. (Ed.), Foundations of Experimental Psychology.
Worcester: Clark Univ. Press, 1929. Pp. x+907.
67. IOE, G. G., Educability Level of 5-Year-Old Children. Psychol. Clin.,
1920, 13, 146-172.
68. JACOBS, J., The Need for a Society for Experimental Psychology. Mind,
1886, 11, 49-54.
69. JACOBS, J., Experiments on " Prehension." Mind, 1887, 12, 75-79.
70. JAMES, W., The Perception of Time. / . Spec. Philos., 1886, 20, 374-407.
71. JAMES, W., The Knowing of Things Together. Psychol. Rev., 1895, 2,
105-124.
72. JOHNSON, G. E., Contribution to the Psychology and Pedagogy of Feeble-
minded Children. Fed. Sem., 1895, 3, 246-301.
73. JONES, A. M., An Analytical Study of 120 Superior Children. Psychol.
Clin., 1925, 16, 19-76.
74. JONES, W. F., An Experimental-Critical Study of the Problem of Grad-
ing and Promotion. Psychol. Clin., 1911, 5, 63-96, 99-120.
75. KIRKPATRICK, E. A., An Experimental Study of Memory. Psychol. Rev.,
1894, 1, 602-609.
76. KOHNSKY, E., Preliminary Study of the Effect of Dental Treatment upon
the Physical and Mental Efficiency of School Children. /. Educ.
Psychol, 1913, 4, 571-578.
77. KRUEGEE, F., and SPEARMAN, C, Die Korrelation zwischen verschiedenen
geistigen Leistungsfahigkeiten. Zsch. f. Psychol, 1907, 44, 50-114.
78. KUHLMANN, F., The Present Status of Memory Investigation. Psychol.
Bull, 1908, 5, 285-293.
79. LAIRD, D. A., Relative Performance of College Students as Conditioned
by Time of Day and Time of Week. / . Exper. Psychol, 1925, 8, 50-63.
80. LEAMING, R. E., Tests and Norms for Vocational Guidance at the 15-
Year-Old Performance Level. Psychol. Clin., 1922, 14, 193-217.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 23

81. LODGE, R. C , and JACKSON, J. L., Reproduction of Prose Passages.

Psychol. Clin., 1916, 10, 128-145.
82. LUMLEY, F. H., and CALHOON, S. W., Memory Span for Words Pre-
sented Auditorially. / . Appl. Psychol., 1934, 18, 773-784.
83. MACDOUGALL, R., Recognition and Recall. / . Philos., Psychol., & Sci.,
Method, 1904, 1, 229-233.
84. MARSH, H. D., The Diurnal Course of Efficiency. Arch. Philos., Psychol.,
& Sci. Method in Colorado Univ. Contr. to Philos. & Psychol., 1906,
14, 1-99.
85. MARTIN, P. R., and FERNBERGER, S. W., Improvement in Memory Span.
Amer. J. Psychol, 1929, 41, 94-97.
86. MCCAULLEY, S., A Study of the Relative Values of the Audito-Vocal
Forward Memory Span and the Reverse Span as Diagnostic Tests.
Psychol. Clin., 1923, IS, 278-291.
87. MCGEOCH, J. A., Memory. Psychol. Bull, 1928, 25, 513-549.
88. MEUMANN, E., Intelligenzpriifung an Kindern der Volkschule. Exper.
Pad., 1905, 1, 35-101.
89. MILLER, K. G., Competency of Fifty College Students. Psychol. Clin.,
1922, 14, 1-25.
90. MITCHELL, D., Variability in Memory Span. / . Educ. Psychol, 1919, 10,
445-457.
91. MITCHELL, F. D., Mathematic Prodigies. Amer. J. Psychol, 1907, 18.
61-143.
92. MUNSTERBERG, H., Die Association Successiver Vorstellungen. Zsch. f.
Psychol, 1890, 1, 99-107.
93. MUNSTERBERG, H., and BIGHAM, J., Memory. Psychol. Rev., 1894, 1,
34-38.
94. MURPHY, G., General Psychology. New York: Harper & Bros., 1933.
Pp. x+657.
95. MURPHY, M., The 10 Year Level of Competency. Psychol. Clin., 1928,
17, 33-60.
96. NICHOLS, H., Our Notions of Number and Space. Boston: Ginn & Co.,
1894. Pp. vi+201.
97. NINDE, F. W., The Application of the Auditory Memory Span Test to
Two Thousand Institutional Epileptics: A Study in Relative Associa-
bility. West Chester, Pa.: Temple Press, 1924. Pp. 36.
98. OBERLY, H. S., A Comparison of the Spans of " Attention " and Memory.
Amer. J. Psychol, 1928, 40, 295-302.
99. PAULSEN, A. E., The Influence of Treatment for Intestinal Toxemia on
Mental and Motor Efficiency. Arch, of Psychol, 1924, 11, No. 69.
Pp. 45.
100. PEATMAN, J. G., and LOCKE, H. M., Studies in the Methodology of the
Digit-Span Test. Arch, of Psychol, 1934, 25, No. 167. Pp. 35.
101. PINTNER, R., The Standardization of the Knox Cube Test. Psychol.
Rev., 1915, 22. 377-401.
102. PINTNER, R., and PATERSON, D. G., A Comparison of Deaf and Hearing
Children in Visual Memory Span for Digits. J. Exper. Psychol, 1917,
2, 76-88.
24 ALBERT B. BLANKENSHIP

103. POHLMANN, A., Experimentelle Beitr'age sur Lehre vom Gedachtnis.

Berlin: Gerdes und Hodel, 1906. Pp. 191.
104. PRELIMINARY REPORT of the Committee on Physical and Mental Tests.
In Proceedings of the Fifth Annual Meeting of the American Psycho-
logical Association, Boston, December, 1896. Psychol. Rev., 1897, 4,
132-138.
105. PYLE, W. H., The Mind of the Negro Child. Sch. & Soc, 1915, 1, 357-
360.
106. PYU2, W. H., A Study of the Mental and Physical Characteristics of the
Chinese. Sch. & Soc, 1918, 8, 264-269.
107. RACHOFSKY, L. M., Speed of Presentation and Ease of Recall in the Knox
Cube Test. Psychol. Bull., 1918, IS, 61-64.
108. REED, H. B., A Repetition of Ebert and Meumann's Practice Experiment
on Memory. / . Exper. Psychol, 1917, 2, 315-346.
109. RICHARDS, T. W., Psychological Tests in First Grade. Psychol. Clin.,
1933, 21, 235-242.
110. RICHET, C, Les origines et les modalites de la memoire. Rev. Philos.,
1886, 21, 361-390.
111. RIPLEY, R. I., "Believe It or N o t " Program, broadcast over the National
Broadcasting Company Network from Radio Station WJZ, New York
City, January 10, 1937.
112. SCRIPTURE, E. W., Tests on School Children. Edtic. Rev., 1893, 5, 52-61.
113. SHARP, S. E., Individual Psychology: A Study in Psychological Method.
Anier. J. Psychol., 1899, 10, 329-391.
114. SHERMAN, I. C, A Note on the Digit Test. Psychol. Clin., 1923, IS, 124.
115. SKERRETT, H. S., The Educability of a Two-Year-Old. Psychol. Clin.,
1922, 14, 221-224.
116. SMEDLEY, F. W., Report of the Department of Child-Study and Pedagogic
Investigation (Chicago Public Schools). Report of U. S. Comm. of
Educ, 1902, 1, 1095-1138.
117. SMITH, W. G-, The Relation of Attention to Memory. Mind, 1895, 4,
47-73.
118. SMITH, W. G., The Place of Repetition in Memory. Psychol. Rev., 1896,
3, 21-31.
119. SMITH, W. G., The Range of Immediate Association and Memory in
Normal and Pathological Individuals. Arch. Nenrol, 1903, 2, 767-805.
120. SMITH, W. G., A Comparison of Some Mental and Physical Tests in
Their Application to Epileptic and to Normal Subjects. Brit. J.
Psychol., 1905, 1, 240-260.
121. SQUIRE, C. R., Graded Mental Tests. / . Educ. Psychol., 1912, 3, 363-380.
122. STARR, A. S., An Analytical Study of the Intelligence of a Group of
Adolescent Delinquent Girls. Psychol. Clin., 1923, 14, 143-158.
123. STARR, A. S., The Diagnostic Value of the Audito-Vocal Digit Memory
Span. Psychol. Clin., 1924, 15, 61-84.
124. STARR, A. S., The Significance of the Ratio Maintained Between the
Forward, Reverse, and Rhythmic Memory Span as Obtained in Three
Thousand Individual Examinations. Psychol. Bull., 1929, 26, 172-173.
125. STRONG, C. A., Consciousness and Time. Psychol. Rev., 1896, 3, 149-157.
 MEMORY SPAN: A REVIEW OF THE LITERATURE 25

126. TEEMAN, L. M., The Measurement of Intelligence. New York:

Houghton Mifflin Co., 1916. Pp. xviii+362.
127. TERMAN, L. M., and MERRILL, M. A., Measuring Intelligence. Boston:
Houghton Mifflin Co., 1937. Pp. xiv+461.
128. TRAVIS, A., Reproduction of Two Short Prose Passages: A Study of
Two Binet Tests. Psychol. Clin., 1916, 9, 189-209.
129. TREDGOLD, A. F., Mental Deficiency. New York: Wood, 1922. Pp. x x +
569.
130. TREVELYAN, G. O., The Life and Letters of Lord Macaulay. New York:
Lovell, 1876. Pp. 307+311 (two vol. in one).
131. VASCHIDE, N., Sur la localisation des souvenirs. La localisation dans les
experiences sur la memoire immediate des mots. Ann. Psychol., 1897,
3, 199-224.
132. WARDEN, C. J., The Factor of Movement in the Presentation of Rote
Memory Material. Amer. J. Psychol, 1926, 37, 257-260.
133. WATKINS, S. H., Immediate Memory and Its Evaluation. Brit. J.
Psychol., 1915, 7, 319-348.
134. WHIPPLE, G. M., The Effect of Practice upon the Range of Visual Atten-
tion and of Visual Apprehension. / . Educ. Psychol., 1910, 1, 249-262.
135. WHIPPLE, G. M., Manual of Mental and Physical Tests. Vol. II. Balti-
more: Warwick, 1924. Pp. 336.
136. WHIIXEY, M. T., An Empirical Study of Certain Tests for Individual
Differences. Arch, of Psychol., 1911-1912, 3, No. 19. Pp. 146.
137. WINCH, W. H., Immediate Memory in School Children. Brit. J. Psychol,
1904, 1, 127-134.
138. WINCH, W. H., Mental Fatigue in Day-School Children as Measured by
Immediate Memory. / . Educ. Psychol, 1912, 3, 18-29, 75-82.
139. WISSLER, C, The Correlation of Mental and Physical Tests. Psychol.
Monog., 1901, 3, No. 6. Pp. 63.
140. WOOLLEY, H. T., An Experimental Study of Children. New York: The
Macmillan Co., 1926. Pp. xv+762.
141. WYATT, S., The Quantitative Investigation of Higher Mental Processes.
Brit. J. Psychol, 1913, 6, 109-133.
142. WYLIE, A. T., A Brief History of Mental Tests. T. C. Rec, 1923, 23,
19-33.
143. XIIXIEZ, P., La continuite dans la memoire immediate des chiffres et des
nombres en serie auditive. Ann. Psychol, 1895, 2, 193-200.
144. YOUNG, K., The History of Mental Testing. Ped. Sem., 1924, 31, 1-48.
145. YOUNG, M. H., A Comparative Study of Audito-Vocal Digit Spans.
Psychol. Clin., 1928, 17, 170-183.
146. YOUNG, P. C , An Experimental Study of Mental and Physical Functions
in the Normal and Hypnotic States. Amer. I. Psychol, 1925, 36, 214-
232.