INTERNATIONAL

LABOUR

OFFICE

STUDIES AND REPORTS
Series J (Education) No. 2

ABILITY IN TYPEWRITING
IN RELATION TO

VOCATIONAL GUIDANCE
By Dora BIENEMAN

PRINTED BY ALBERT KUNDIG

GENEVA

CONTENTS

Page

Prefatory Note
Analysis of the Occupation
Enquiry
Observation and Experiment
Diagnosis of Ability in Typewriting
Analysis of Individual Mental Qualities in Relation
to Ability in Typewriting
The Questionnaire Method
Experiment
Discovery of Tests for Detecting Ability in Typewriting
The Author's Investigations
Tests
Results
Vocational versus Psychological Tests; Moral Qualities
Bibliography

5
10
10
11
15
15
15
18
22
33
34
37
41
47

PREFATORY NOTE

The essential purpose of vocational guidance being to determine
scientifically the abilities of an individual, it obviously presupposes
a knowledge of the abilities required in the various occupations.
These two elements in vocational guidance — occupational analysis and diagnosis of occupational abilities — are treated in two
separate chapters in Professor Claparède's study of vocational
guidance published by the International Labour Office in October
1922, (*); the author there describes the different methods adopted
and discusses the nature and value of tests.
His general survey of methods in general use may well be supplemented by a concrete illustration. The present monograph,
by M Ue Dora Bieneman, certificated pupil of the Jean Jacques
Rousseau Institute, is an example of the practical application of
experimental methods to vocational guidance. After a brief
analysis of the work of typewriting, the author endeavours to
arrive at a method of determining the abilities required for this
work, basing her conclusions on the results of earlier research
by a variety of investigators and on her own experiments with a
group of typists at the International Labour Office. Her work
has a double interest : it emphasises the complexity of the problem
which has to be solved and explains the actual machinery of the
tests.
t1) Problems and Methods of Vocational Guidance, by Dr Edouard CLAPA
REDE. Studies and Reports, Series J (Education), No. 1. Geneva, 1922.
79 pp. 2s.; 40 cents.

Ability in Typewriting in relation
to Vocational Guidance 0

Typewriting as an occupation has increased in importance
almost more rapidly than any other in recent times. The typewriter was invented about two hundred years ago, but it only
began to come into use about forty or forty-five years ago and even
then only in large establishments. Today typewriting is one of
the most important commercial occupations and in fact seems
indispensable. Unfortunately the occupation is overcrowded
with inefficients ; the work seems easy, a matter of habit and practice rather than of special ability. There are far too many youths
and even more girls (for typewriting is becoming more and more
a woman's occupation), who have no very definite inclination or
pronounced capacity and therefore become typists for lack of a
better opening, and are confident of success. In this, however,
they are mistaken. Typewriting is not a purely manual trade,
and in order to succeed and achieve a fairly secure and well paid
position the typist needs to be able to do more than strike the keys
with two or three fingers. She needs special capacities and vocational abilities, not to mention thorough training.
People with long experience of the work, when asked: "Can
anyone be a successful typist ?" replied definitely that application,
perseverance, and long training are not sufficient to ensure success. They mentioned cases of typists who began work at the
same age and had the same training, but while some made good
progress others never rose above the level of copying typists.
At the same time those who made these statements were often
unable to define the special qualities which made for the success
of good typists. The investigator thus found himself faced with
certain questions :
(*) Prepared in the Vocational Guidance Department of the Jean Jacques
Rousseau Institute, Geneva, under the direction of Professor Edouard Claparède.

— 8 —
Does "ability in typewriting" actually exist ?
Is this ability a special function or does it depend on a group
of functions which are found in both good and bad typists, but
in varying degrees ?
Is it possible by mere common sense to analyse this ability and
distinguish its component factors, or must this be done by observation and experiment ? J ï
Given that these questions were answered, it would still be
necessary to find how this ability in typewriting can be detected
and measured in practice. It may be laid down once and for all
that none of these questions can be answered a priori. Of course
certain essential qualities can be noted at first glance: a certain
amount of memory and of attention, some education, great rapidity
of movement, motor dexterity, etc., but mere common sense cannot
discern the relative importance of these factors, which is the most
necessary and which is indispensable. Unless conclusions are
based on discussion of facts and experiments, there is considerable
risk of error.
For example, one might suppose that it is desirable, if not
necessary, for a typist to be equally deft "with both hands.
In the course of the investigations described later, however,
it was found that ambidexterity is not as important as had
been thought. For instance, the typist who was ranked first by
her supervisors in respect of vocational efficiency did not show
very great symmetry between the two hands, while the worker
who ranked bottom as typist was almost at the top (third) for
ambidexterity ('). This shows that there is no necessary relation
between vocational efficiency and ambidexterity.
In considering the medical aspect of the case it is also difficult
to determine a priori the drawbacks of this occupation in relation
to the constitution of the subjects, to foresee the effect of fatigue,
etc. (2) Similarly, when traits of character are considered, it
is impossible to determine which are the most valuable (calm
or vivacity, initiative or "need of direction", etc.), or, indeed, if
these characteristics have any effect on ability. These problems
can only be solved by observation of facts.

i1) See below, p. 38.
(a) The investigations here reported were not carried into this important
but as yet little studied field. Cf. MAUVEZIN: La rose des métiers, p. 291, for
the physical qualities required.

— 9 —
The facts again, may be examined on two different methods:
(a) objectively, by analysing the occupation and deciding what
psycho-physiological functions are involved; (b) subjectively,
by analysing the psycho-physiological processes in the individual
which are involved in typewriting.
In the investigations here reported, the objective analysis was
carried out by enquiry and observation. Under the second heading
research in the psychological characteristics of the occupation was
undertaken in the light of experiments already made in this field.
The results of these investigations are given in the following pages.

Analysis of the Occupation

The occupational analysis of typewriting is a practically new
question. Apparently very few books have been published on the
subject, and these generally only cover part of it. Previous publications were not ignored, but it proved necessary to collect further
information by questioning competent persons in close touch with
the occupation and checking their statements by observation and
experiment.
Enquiry
When questioning employers or workers, the conversation should
be guided by means of questions prepared beforehand in order
to secure definite answers and to avoid the danger of neglecting
any important point. These prepared questions should not be
allowed to hamper the free flow of conversation, but serve mainly
as starting points.
The questionnaire was deliberately made very simple in order
to avoid any danger of suggesting replies and to make the interviews
as conversational as possible. The following questions were asked :
(1)
(2)
(3)
(4)
(5)

What does the work of a typist consist of ?
What are the qualities required for the work ?
What are the most important qualities ?
What are the defects most frequently noted ?
Are there any defects which make it absolutely impossible
to enter the occupation ?
(6) What type of character is the best in this occupation
(vivacious or calm, etc.) ? Intelligence ?
(7) What is the best form of preparation for the occupation ?
(Length of training, etc.) Why ?
(8) Miscellaneous (fatigue, rapidity of work, salaries, etc.).

— 11 —
Needless to say, it is unnecessary to follow the order given above.
The chief thing is to allow the person questioned to express his
views freely. The investigators were fortunate in obtaining valuable
information from the heads of typewriting offices and teachers
of typewriting in the schools of Geneva, but the replies were
frequently incomplete.
Observation and Experiment
These were necessary in order to complete the data and make
them more precise (*).
Before attempting a description of the occupation, a few notes
may be given on the typewriting machine and the methods in use.
The single keyboard (') is now adopted in all the chief makes
of typewriter, and the double keyboard is gradually disappearing.
It is interesting to note that the variety of makes of machine is not
as great a drawback as might be imagined; a good typist can
rapidly adapt herself to a different keyboard. Moreover all the
chief machines are standardised. The English keyboard only
differs from the international keyboard by the inclusion of certain
signs peculiar to English (£ for the pound sterling, etc.), and the
German keyboard by certain German signs (ä, etc.). As one of
those interviewed on the subject picturesquely said: "It is the same
thing with bicycles; whether you buya Swiss or a British machine,
in spite of minor differences you always have to pedal."
As regards method, there is general agreement that it is best
to use all ten fingers. This method has certain difficulties, takes
longer to learn (»), and is awkward to begin with, but once the
habit has been acquired the effect of practice is more permanent (*).
There is no doubt, moreover, that it is possible to attain greater
speed in typewriting with ten fingers than with two or four.
A first-class typist who had not acquired this method stated that
she fully realised the advantage it would be to her and that she
(') See below, III, The Author's Investigations.
(2) On the single keyboard the capitals are on the same keys as the small
letters.
(s) Cf. SCHILLING: Beiträge zur Psychotechnik der Schreibmaschine und ihrer
Bedienung ; Rationalisierung der Maschinenkonstruktion auf psyckotechnischer
Grundlage, in Praktische Psychologie, 1921, Vol. I l l , No. 1, pp. 21-31. Leipzig.
(*) Cf. MENZEL: Beiträge zur Psychotechnik der Schreibmaschine und ihrer
Bedienung;
Vergleichendes Methodik und Didaktik des Maschinenschreibens
auf experimenteller Grundlage, in Praktische Psychologie, 1912, Vol. II, No. 7,
pp. 269-274. Leipzig.

— 12 —
would thus acquire much greater speed. For this, however, a
thorough training is required ('), and few people are ready to
undertake this. Yet training is needed for any other trade or
occupation and very rarely lasts only one or two months, as in the
case of typewriting. In two months the pupil knows his machine
both from the mechanical and the technical point of view (style,
abbreviations, etc.), but obviously he has not yet much speed,
which can only be acquired by practice.
R. E. Hoke (2), in working on the construction of standard
tests in typewriting, found himself held up by a number of preliminary problems. These included :(1) the frequency of occurrence
in the English language of the various letters of the alphabet and
marks; (2) the relative abilities of the eight fingers and the two
hands; (3) the burden of work which the present typewriter keyboard places upon the fingers and hands (when all ten fingers
are used).
Hoke concludes that typewriter keyboards are not adapted
to modern methods of typewriting, as the keys are so arranged
that certain fingers are given far more work than others, while
from his experiments he finds that the ability of the different
fingers is roughly the same. He proposes a new keyboard, which is
better adapted to the frequency of occurrence of letters and marks
in English and to the relative abilities of the fingers and hands.
Although the processes of typewriting are well known, the chief
operations involved (s) may be briefly recalled as follows :
(1) insertion of the sheet of paper in the machine, estimation
of margins, etc.;
(2) comprehension of the text ;
(3) typing, strictly so called (striking the keys, pushing back the
carriage, etc.);
(4) care and maintenance of the machine.
In considering these four groups of operations one may ask
(*) BOLLER: La sténo-dactylographie.

Cf. Troisième cours d'orientation pro-

fessionnelle, organisé par l'Association suisse de conseils d'apprentissage et de
protection des apprentis et la Société suisse d'utilité publique. Geneva, 1920.
(2) R. E. H O K E : The Improvement

of Speed and Accuracy

Typewriting.

Johns Hopkins University Studies in Education, No. 7. Baltimore (U. S. A.),
1922.

(3) The t y p i s t almost always has o t h e r w o r k to do in addition to a c t u a l
t y p e w r i t i n g . This varies w i t h the p o s t she holds. It m a y include sorting
correspondence, filing, etc.
M a n y t y p i s t s are also s h o r t h a n d - w r i t e r s , b u t
this s t u d y deals exclusively with t y p e w r i t i n g .

— 13 —

•

which of them involves more particularly the real ability in typewriting, i.e. the ability which is least affected by training and
practice. Operations (1) and (4) may be passed over, as it was
observed that these depended much more on training than on
any particular ability. At the most one might note under (1)
accuracy of eye and taste in arrangement of text, headings, etc.,
and as yet it has not been discovered to what extent these qualities
are innate or acquired.
The actual striking of the keys involves two distinct factors,
accuracy and speed. Accuracy depends almost entirely on the
nature of the training received. Speed, on the other hand, seems
to be a more individual quality; with the same amount of practice
two persons will develop different speeds.
Comprehension of the text is the most important operation
from the point of view of individual qualifications. This can
readily be understood, as it involves very complex mental processes, and individual differences are most marked in the higher
functions. Comprehension of the text may be acquired by very
different means. The typist has sometimes to copy (*), sometimes
to type from dictation (either verbal dictation or, more rarely, the
dictaphone), or to transcribe notes previously taken from dictation
(either in shorthand or longhand). A good typist should be able
to do any of these three kinds of work, but in practice this is not
always the case. Some typists transcribe their notes excellently
but cannot type accurately from dictation; others, again, type well
from dictation but cannot transcribe their notes correctly. Obviously, therefore, different abilities are involved.
Attempts were made to define the factors which make up these
subdivisions of typewriting work. With the aid of information and
experimental research endeavours were also made to determine
the qualities and abilities required by each of them, with the
following results.
Copying

Typing from dictation

\ Attention
) Visual memory for words
1 Calmness
J Attention
' Auditory memory

f1) The position of copying typist to be found in some offices (e.g. lawyers'
offices) are generally very poorly paid (80-150 francs in Geneva). These
posts are frequently held by second-rate typists, as they require less complete
training.

— 14 —

Transcription

Memory
Aptitude for deciphering
Aptitude for combination
Rapid comprehension
Imagination
Initiative

II
Diagnosis of Ability in Typewriting

As a certain number of investigations have already been carried
out in this field, their results may be noted here, so that recapitulation of what has been done may show more clearly what remains
to be done. When a general view of the ground covered has been
obtained, it will be easier to see what points require emphasis
and where further research is necessary. Thus the experiments of
previous investigators will obviate preliminary hesitation and
endeavour, and a further step will be taken in the co-operation
which is at once an essential and a privilege of all scientific work.
Investigations for the purpose of analysing ability for typewriting will first be described, and subsequently those undertaken
primarily with a view to finding tests for vocational guidance.
The distinction between these two types of research is often very
fine, and many of them serve both purposes simultaneously.
ANALYSIS OF INDIVIDUAL MENTAL QUALITIES IN RELATION
TO ABILITY IN TYPEWRITING

The Questionnaire Method
The questionnaire drawn up by Heinitz (') is one of the most
complete yet produced. It includes 89 questions under four headings ;
I. Questions about the machine.
II. Questions about the "stimulant" (Reizgebung), i.e. dictation,
copying, or transcription; for example:
(9)
(12)

Do you transcribe from shorthand ?
Do you type from dictation ?

(*) Wilhelm HEINITZ: Vorstudien über die psychologischen Arbeitsbedingungen
des Maschinenschreibers, in Zeitschrift für angewandte Psychologie, 1917.
Vol. XIII, Nos. 1-3, pp. 37-90. Leipzig.

— 16 —
III. Questions about the typist.
IV. Questions about training for the occupation.
The following are some of the most important questions in t h e
third section:
(17) Do you know any foreign languages ?
(20) Can you fix your attention on a certain part of the work which
recurs at regular intervals ?
(21) Can you concentrate your attention closely at a given moment ?
(22) Can you rapidly fix your attention on something new ?
(23) Can you give the same amount of attention to a single piece of
work continued for a long time without perceptible fatigue or
diminution of attention through fatigue ?
(25) Can you repeat accurately what you have read shortly beforehand ?
(26) Can you repeat accurately what you have heard shortly beforehand ?
(38) Do you habitually read fast or slowly ?
(44) Are you interested in the construction of machines in general ?
(45) Can you readily distinguish different kinds of paper by touch ?
(55) Can you estimate considerable distanses by eye accurately ?
(56) Do you see immediately when the paper is slightly crooked in your
machine ?
(58) Can you complete incomplete texts easily and accurately ?
(59) Can you easily read bad handwriting ?
(61) Is your attention easily distracted by disturbing impressions ?
(62) Can you react to several impressions at roughly the same time and
grasp them ?
(63) Do you readily become impatient when you are not successful in
your work ?
(66) Are you interested in things which do not affect you personally
very closely ?
Heinitz recognised that his questionnaire was unsatisfactory, as
it required introspection to a degree of which subjects were rarely
capable. This shows the importance of checking the replies by a
series of tests. Written replies to a questionnaire involve many
dangers, and Heinitz therefore preferred an oral questionnaire,
so that it was possible to give supplementary explanations.
He questioned 25 subjects, of whom 23 were professional typists.
Many of his questions produced no significant results ; for example :
(31) Are you deft with your hands (knitting, embroidery, crochet) ?
(32) Do you play a musical instrument ?
Other questions produced doubtful or unsatisfactory replies;
for example:
(28) Have you a good memory for names ?
(29) Have you a good memory for digits ?

Nearly all the subjects naturally stated that they had a very good
memorv.

— 17 —
Heinitz nevertheless collected data precise enough to enable
him to observe certain important points:
(1) the need of a better education than the great majority of
typists at present have, and particularly the importance
of spelling and the study of languages;
(2) the importance of interest in the work, which is often as
important as ability, if not more so ;
(3) the importance of rapid and accurate tactile and motor
estimation of space (in this connection Heinitz considers
t h a t the majority of typists have this ability to estimate
space, and when typewriting co-ordinate a variety of
impulses, forming what he calls "diagrams'', although this
is frequently unconscious even in experienced typists).
Heinitz drew up a list of the abilities to be found in the ideal
typist, though he did not claim t h a t it was exhaustive. The list
is as follows:
(1)

General Abilities
General intelligence
Good education
Knowledge of languages
High degree of concentration
Constant and sustained attention
Good memory for words, names, and digits
Rapid reactions
Motor functions susceptible to practice
Good and rapid work
(2) Special Abilities
Mechanical and technical ability
Good tactile-motor capacities, especially accurate and rapid
memorising of spatial figures (diagrams)
Ability to co-ordinate several elementary processes (Gesamtimpulsen)
Aptitude for visual estimation
Sense of division of space
Ability to decipher difficult handwriting
Ability to prevent one's thoughts straying
Capacity for eliminating past impressions
Ability to retain in the mind a large number of spoken words
(grössere Lautkomplexe)
Ability to correct rapidly and conscientiously
(3) Moral Qualities
Accuracy, neatness, sense of order
Patience
Ability to take an interest in things which are not of direct
use to oneself
Ability to take orders instead of wishing to give them

— 18 —
Some of Heinitz's questions were suggested to him by an article
by Lipmann^). Lipmann's questionnaire is more general than
that of Heinitz. It was drawn up as a basis for questions on any
"middle-grade" (3) occupation (e.g. railway guards, tram drivers,
bookbinders, telegraph and telephone employees, typists), rather
than for any given trade. Lipmann found that the question —
"What abilities are required for your work ?" as a rule only
evoked vague replies, e.g. intelligence, taste, accurate eye, etc.
He therefore tried to make his questions more definite in order
to oblige the subjects to make their replies more definite. His
questionnaire includes 86 questions; the results of those which
elicited affirmative replies with regard to typists are noted here.
The occupation of typewriting requires ability:
to fix at a glance in the memory and to reproduce in full large
sections of text one has read (26) ( s );
to reproduce shortly afterwards what one has read once (27);
to write correctly (spelling and punctuation) (35);
to execute different movements very rapidly one after the
other (41);
to group series of different movements which recur frequently (43);
to react rapidly to different stimuli by different movements,
but in each case by the prescribed movement (44).
Lipmann came to the same conclusion as Heinitz, namely, that
it is necessary to have a series of tests in oider to verify by experiment information obtained by questionnaire. This leads to the
second method of analysing individual mental processes and their
relation to ability in typewriting.
Experiment
In this method the attempt is made to discover the psychophysiological factors involved in the exercise of a given trade
(!) Otto LIPMANN: Zum psychologischen Charakteristik der ^mittleren"

Be-

rufe, in Zeitschrift für angewandte Psychologie, 1916, Vol. XII, pp. 91-107.
Leipzig.

(2) Middle-grade occupations are those requiring a certain a m o u n t of general
intelligence a n d some combination of physical abilities, b u t these m e n t a l
>rocesses are limited and d e t e r m i n e d mechanically b y conditions of w o r k

Íaid

down beforehand.

PIORKOWSKI: Beiträge zur psychologischen Methodo-

logie der wirtschaftlichen Berufseignung; Beihefte zur angewandten Psychologie,
No. 11, 1915, p. 16.
(3) T h e n u m b e r of t h e question in L i p m a n n ' s questionnaire is given in
brackets.

— 19 —
by experiments on the workers. Lahy's researches (') provide
an example of this form of experimental control and explain
its nature.
In order to determine the mental functions related to superior
efficiency in any occupation it is first necessary to discover what
functions are highly developed in the most efficient workers in the
occupation. First-rate workers must be found and then graded
according to their vocational efficiency in order to observe the
correlations between this and test results.
For this preliminary grading, which should be as objective as
possible, Lahy obtained three opinions. His 28 subjects (17 men
and 11 women) did a vocational test. They copied a continuous
text of 1,702 letters or signs. The work was then judged by
directors of typewriting offices and their judgment was confirmed
by several typists not taking part in the investigations. Speed,
number of errors, and arrangement were taken into account, but
as several factors of equal importance (e.g. speed and accuracy)
were involved in this grading, Lahy selected one — speed — and
treated it as the chief factor.
He then approached the problem of the particular qualities
possessed by a good typist by what he called "overall" investigations; he studied a group of functions in his subjects which seemed
to him to be involved in typewriting. Having thus drawn up
a list of functions he selected tests for measuring them and applied
these tests to the typists who had been graded on the typewriting
test. In the course of his work he found it necessary to eliminate
some tests which gave no result, especially those intended to
measure higher functions. "Neither abstraction, judgment, nor
imagination have any bearing on superior efficiency in typists".
(Lahy nevertheless makes reservations regarding the value of
these results, as the higher functions are extremely difficult to
investigate.) He then compared the grading for vocational
efficiency with the results of the tests. As will be seen later (2),
Lipmann calculated from Lahy's figures correlations showing
more exactly the relations between vocational efficiency and the
results of the experiments.
I1) LAHY: Les conditions psycho-physiologiques de l'aptitude au travail
dactylographique, in the Journal de psychologie et de pathologie générale, Vol. XV,
No. 4, July 1913. La méthode à suivre pour la sélection des travailleurs; les
dactylographes, in the Anals de l'Institut d'orientaciô professional, Vol. II,
No. 3, May 1921, pp. 65-78. Barcelona. The second article is a summary
and continuation of the report of experiments given in the first.
(2) See below, pp. 21-22.

— 20 —
The qualities examined by Lahy in his first investigations were
as follows:
Muscular sensibility (determined by a Toulouse myoesthesiometer
— weight lifting).
Tactile sensibility.
Auditory reaction-time (with a d'Arsonval chronometer).
Muscular symmetry of the two hands (measured by a Chéron dynamometer).
Memory for concrete phrases.
Memory for digits.
Attention.
Later on he also investigated memory for letters.
In general, the following qualities appear to be constant in a
good typist:
Good memory for digits and phrases.
Sustained attention.
Fine tactile and muscular sensibility.
Tendency to ambidexterity.
Rapidity of auditory reaction, however, does not appear to
indicate superior vocational efficiency.
Following on his first experiments, Lahy pursued his study of
motor rapidity and fatigue (tapping tests) and of mental spelling.
To carry out this last test rapidly, "the subject must be thoroughly
familiar with spelling. If he hesitates or stops there is a loss of
time which detracts from his speed in typewriting". The real
reason why a good education is always considered necessary for
a typist is, in Lahy's opinion, because a knowledge of spelling and
grammar is a necessary condition of speed in mental spelling.
The reason commonly advanced was " a kind of moral need, since
a text must be copied without spelling mistakes in order to satisfy
the professional conscience of the worker". Lahy went into the
matter more deeply and the reason he advances appears more
satisfactory at first sight. From the spelling test he went on to
investigate ability to decipher manuscripts (l). His experiments
on ten adult subjects led him to believe t h a t there is an ability for
deciphering.
The final list of abilities to be tested which he compiled is as
follows :
Motor rapidity (tapping).
Susceptibility to fatigue (deduced from the preceding test).
(*) Cf. BAER and BOVET: Lectures des textes manuscrits, in the Intermédiaire des éducateurs, Mar. 1920, pp. 28-33. Geneva.

— 21 —
Alternation of hands and fingers in the tapping test.
Simple or compound spelling (a) in understanding directions ; (b) speed
in oral spelling.
Deciphering manuscript.
Memory.
Tactile and muscular sensibility (possibly).
Reaction time.
Attention.
On the basis of these investigations he concluded that "superiority
in an occupation is not indicated by one particular sign or by the
predominance of a particular function, but by a group of signs
which vary in each individual". The good typist is not necessarily
the one who possesses this group of signs indicating the superiority
of her psycho-motor abilities, but "the one who is most successful
in adapting this superior ability to the combined effort required
by the complex operations of the work of the occupation".
Lipmann's account (') of Lahy's first investigations is a valuable
addition. He calculated the correlations between the various
functions as measured by the tests and vocational efficiency as
measured by the typewriting test. The highest correlation was
shown by memory for phrases, while the test for auditory reactiontime showed a negative correlation. The correlation between
vocational efficiency and the combined results of the tests of memory,
muscular and tactile sensibility, and ambidexterity is higher than
any of the correlations with a single test, thus confirming Lahy's
statement that "superiority in an occupation is indicated by a
group of signs".
Lipmann's figures are given below. The correlations are calculated by a special method, giving the median and quartiles
of the differences between the two series expressed as percentages.
This method gives the following figures:

Perfect positive correlation
Absence of correlation
Perfect negative correlation

Lower
quartile

Median

¿JP^

0
12.5
25

0
25
50

0
37.5
75

t1) Otto LIPMANN: Die psychophysischen Merkmale der Eignung zum Maschinenschreiben, in Zeitschrift für angewandte Psychologie, 1920, Vol. XVII,
Nos. 1-3, pp. 159-165. Leipzig.

— 22 —
The results of Lipmann's experiments were as follows:
Lower
Upper
Median
quartlle
quart ile
Memory for phrases
0
29
6
25
15
Muscular symmetry
8
27
15
Memory for digits
8
31
19
37
Tactile sensibility
8
21
40
25
Muscular sensibility
8
42
33
Attention: errors
17
54
35
duration
8
Reaction-time
21
W. W. Tuttle's experiments may also be mentioned ('). His
subjects were a class of twenty students beginning the study of
typewriting. The tests used were the following:
(1) motor control (tapping any key on the keyboard of the typewriter
as rapidly as possible for five seconds) ;
(2) sense of rhythm;
(3) attention and accuracy;
(4) memory span;
(5) ability to follow directions;
(6) substitution.
Tuttle then calculated the correlation between the results of
these tests and those of a typewriting test administered at the
end of the first term. He obtained negative correlations with the
memory-span test and very low correlations with tests 2 and 5.
He concludes that the correlation between the scores in the typewriting test and those in the tests of motor control, attention and
accuracy, and substitution is sufficiently high to indicate a strong
relation between capacity to learn typewriting and the tendencies
tested. (The coefficient of correlation between the scores for the
typewriting test and those of the three tests mentioned was 0.621.)

DISCOVERY OF TESTS FOR DETECTING ABILITY IN T Y P E W R I T I N G

It has already been shown that Heinitz, Lipmann, and Lahy
endeavoured b y different methods to achieve the same object: the
determination of ability in typewriting. Rogers, Cody, Link, and
Burt approached the problem from another point of view. Their
object was to obtain a series of objective tests controlled by experi(') W. W. TUTTLE: The Determination of Ability for Learning Typewriting,
in the Journal of Educational Psychology, Vol. XIV, No. 3, Mar. 1923, pp.
177-181. York, Pennsylvania.

— 23 —
ment (i. e. with a high correlation with vocational efficiency), which
can be used in selecting typists. They did not endeavour to
analyse the occupation or to base their tests on abilities which
might be supposed to be necessary.
Rogers (*) investigated the method of empirical vocational
tests (2), which is much in favour in the United States. He hoped
if possible to discover tests of practical value in the selection and
even the guidance of applicants for employment in any commercial
occupation. He drew up a list of ten tests measuring rapidity of
mental functions.
(1) Five tests of logical relations, the first four requiring the subject to
give the opposite of a word, the object of a verb, the subject of a
verb ("action-agent"), and the verb of a subject ("agent-action").
The fifth is a mixed relations test.
(2) A colour-naming test (100 squares of different colours — red, blue,
black, yellow, and green — each one square centimetre in area,
placed at distances of one square centimetre from each other, are
shown to the subject, who has to name the colours as rapidly as
possible).
(3) A form substitution test (the subject has to name a number of geometrical forms — square, circle, triangle, cross, star, etc. — as
rapidly as possible). (Cf. test No. 2.)
(4) Number-group checking (84 numbers of 6 digits each are shown to
the subject, who has to mark the numbers containing both an 8
and a 9).
(5) A "hard directions" or sentence completion test. The blanks have
to be filled in in accordance with the context.
Nine of these tests were taken from a series standardised by
Woodworth and Wells for the American Psychological Association,
while the tenth is one of Trabue's language tests. For the calculation of correlations with his tests, Rogers then subdivided the work
of a typist into (1) stenography, (2) typewriting, (3) general knowledge (grammar, spelling, punctuation, etc.). His subjects —
45 students of typewriting — were then graded for stenography
by their instructor, and for typewriting by six typewriting tests
repeated once a month for six months. The subjects wrote from
dictation for ten minutes and the average number of words correctly typed in one minute was noted. For general knowledge
(x) ROGERS : Psychological Tests for Stenographers and Typists, in the Journal
of Applied
Psychology, Vol. I, Sept. 1917, pp. 268-274. Worcester, Mass.
(2) HOLLINGWORTH, the American psychologist, has described the basis
of this method as follows: "If the test records and ability of any particular
type of work show a high positive correlation, so that an individual who is
good, medium, or poor in the one is, as a matter of fact, also found to be
good, medium, or poor in the other, then, without further analysis, the one
may be used as the sign of the other." (Quoted in ROGERS: op. cit.)

— 24 —
there was a test of grammar, spelling, etc. The following table
shows the correlations which Rogers obtained:

CORRELATIONS CALCULATED MONTHLY BETWEEN TYPEWRITING WORK
AND TESTS

Test

Oppo- Verb- Agent- Action- Colour Mixed
relasites object action agent naming tions

NumDirecber
tions checking

Form Numsubsti- of ber
subtution jects

Stenography

0.45

0.36

0.19

0.23

0.34

0.31

0.46

0.07

0.40

Grammar

0.40

0.37

0.35

0.38

0.43

0.54

0.22

0.16

Typewriting:
October
November
December
February
March
April

0.17
0.11
0.07
0.15
0.19
0.54

0.41 0.29
0.43 0.19
0.46
0.40
0.55
0.31
0.21 -0.02
0.57
0.28

0.42
0.43
0.29
0.41
0.00
0.40

0.30 - 0 . 0 9
0.43
0.21
0.45
0.17
0.41
0.25
0.29
0.04
0.61
0.00

0.11
0.14
0.19
0.13
0.34
0.32

0.45
0.47
0.37
0.53
-0.01
0.30

0.21
0.27
0.11
0.37
0.30
0.42

42
40
40
40
29
27

Average

20.5

43.8

24.8

32.0

41.0

13.0

21.0

35.0

28.0

One or two interesting points may be noted with regard to the
correlations with the typewriting tests. (1) The correlation with
any one test remained fairly constant during the first four months.
(The correlations for March cannot be compared with those for
the preceding months, as the typewriting text was exceptionally
difficult; moreover only 29 of the subjects took part in the test,
while only 27 took part in that for April.) (2) Practice evidently
tends to raise the correlation.
Rogers also calculated the correlation between the results of
the typewriting tests and groups of two, three, or more tests.
The correlations so obtained were noticeably higher than those
given in the table.
(1) Verb-object, action-agent, number
0.56
(2) Verb-object, agent-action, action-agent, colours,
number
0.56
(3) Verb-object, action-agent, colours, number
0.58
(4) Verb-object, agent-action, action-agent, colours,
number, form
0.62
(5) Verb-object, colours, number
0.63
He concludes that the method of empirical tests is the best
criterion yet discovered for vocational guidance and selection in

— 25 —
typewriting. He does not claim that the limited number of tests
which he investigated (some of which give a low correlation)
are adequate, but he considers that if a larger number of tests and
of subjects were utilised there is no doubt that other tests showing
much higher correlations than his would be discovered.
Cody (*) describes the National Business Ability Tests, of which
he has published two parallel series. In 1913 a national committee
was organised in order to consider means of promoting efficiency
both in schools (particularly commercial schools) and business
offices. The committee, which consisted of educators, manufacturers, business men, etc., concentrated its attention on developing
and trying out a series of elementary tests of ability to perform
common operations in a business office. The purpose was to determine the education required by any office employment and, at the
same time, to find a means ol selecting candidates for vacancies.
The tests selected were frequently tried -out on a large number
of subjects, until a standardised series was obtained in 1915.
In 1916 parallel equivalent series were issued. These series are
essentially selection tests, i.e. their purpose is to distinguish a good
worker from a bad one; they cannot detect aptitude for an occupation in a person who has not yet entered it, which is the function
of vocational guidance tests.
Cody's tests for typists are: (1) to copy a typewritten or printed
business letter on the machine in five minutes; the number of
strokes and the number of errors are counted; (2) to copy a printed
or typewritten text on the machine very exactly for mimeographing; the number of errors is counted.
Link's book (s), the first part of which deals with psychological
tests, includes a chapter on typewriting. Typewriting is a specialised form of office work, and in drawing up tests three factors
must be taken into account (Link is only concerned with vocational selection):
(1) the general ability required for any office work;
(2) manual dexterity gained by practice;
(3) the innate ability which produces the rapid and accurate
typist.
(*) Sherwin CODY: Commercial Tests and How to Use Them. Yonkers-onHudson, N. Y., World Book Company. 1920. See especially pp. 37 and 160-163.
(2) H. C. LINK: Employment Psychology., New York, Macmillan. 1919. See
especially pp. 88-92 and 422.

— 26 —
In order to find tests to meet these conditions Link conducted
an extensive series of experiments. Tests were given to more than
300 girls and boys in a commercial high school, to 76 pupils of
two business schools, to a group of 22 office typists, to a group of
19 stenographers, to more than 400 candidates for positions as
typists and stenographers, and others ; in all more than a thousand
people were tested. On the basis of these experiments Link drew
up a list of tests which, in his opinion, may serve as practical
guides in the selection of applicants for employment, and in distinguishing good from bad typists.
Link's system of rating is so ingenious as to merit description
in detail. His method can be applied equally well to any test.
Tests with a Time Limit
The point of reference in rating (i. e. maximum possible score)
is the number of units completed in the given time (figures cancelled, blanks completed, etc.). Thus, in the sentence completion
test (given below) there are 31 blanks to be filled in. Dividing the
number of blanks filled in by any subject by 31 gives that subject's
score in the test. For example, if a subject filled in 25 blanks, his
25
result would be ^ = 0.80. In order to allow for errors, the number of errors is subtracted from the total obtained. Thus in the
case quoted, if the subject filled in one of the blanks incorrectly
(one such mistake counting as two errors), the result would be
25 — 4
Tests without Time Limit
The point of reference in this case is the perfect completion of a
certain number of units in a given time; e. g. the point of reference
in the spelling tests is 84 words corrected in 126 seconds. The
score of any one subject is obtained by dividing 126 seconds by
the time actually taken. If, for example, he took 130 seconds,
126
his rating would be 7™ = 0.97. In order to allow for errors and
omissions, these are subtracted from the total possible number of
units. Link's formula is as follows :
Ru x (RU — E)—_r
T
RU = the number of units called for by the point of reference

— 27 —
(e. g. in the spelling test 84 words corrected) ; Ru = the reference
126

time per unit (e. g. ^-¿- = 1.5); E = the number of errors; T = the
time actually taken. Suppose, for example, that a subject took
127 seconds for the spelling test and omitted 14 words, the point
of reference being 84 words corrected in 126 seconds, we obtain
the following result by the application of Link's formula:
1.5 x (84-14)
127

°' 8 3

This being Link's method of rating his tests, some description
may be given of the actual tests he uses for typists. He always
begins with a "shock absorber", the purpose of which is to put the
subject at his ease; it is not counted in estimating the final results.
As a shock absorber in the typists' test Link uses a typewritten
letter with several glaring mistakes in spelling, punctuation, and
capitalisation. The subject is asked to read the letter and check
the mistakes. This test should be very simple.
In the substitution test the subject has to relate different letters
according to a key given at the top of the sheet, to which he can
refer. It is, in effect, a test of attention. The key is as follows:
t g u p m k
e v e ]
z d
Below the key is a table of 144 letters (each of the six letters in
the first line of the key repeated 24 times). The first lines of the
table may be given by way of example:
g t u p m k t g u m p k
m p t k m p g t m k u g
Under each of these 144 letters the subject has to place the
corresponding letter in the second line of the key, thus below g
he places v, below t, c, etc. Link's point of reference is 90 letters
substituted in 120 seconds; errors count double. For example, if
a subject wrote 76 letters and made two errors, the formula would
be:

Sentence Completion (Context) Test
The subject has to fill in blanks in the sentences given him,
e. g. "The kind lady . . . . the poor man a dollar." The point of

— 28 —
reference is perfect completion of 31 blanks in 240 seconds. A
wrong word in a blank counts double, and each word omitted
counts one.
Spelling Test
A list of words in ordinary use, some of which are misspelt, is
given to the subject. He has to mark the words which are spelt
incorrectly. Subsequently the examiner dictates to the subject
those words which have been wrongly marked or overlooked.
Point of reference : perfect completion of 84 words in 126 seconds.
Errors count one, if made in the first part of the test, two if they
are repeated in dictation.
Typewriting Test
A business letter has to be copied. The sample given by Link
contains 150 5-space units. Point of reference: 150 units in 150
seconds; errors or omission of words count five; mistakes of punctuation or capitalisation two; smudges two, etc. For example, a
subject who completed the test in 160 seconds with an error-score
of 20 would have the following final score :

Link endeavours to express the results of a series of tests in
terms of a single rating, determining the value of each test by its
index of correlation. To take his own example, three of his tests
gave correlations of 0.56, 0.63, and 0.72 respectively. Obviously
the third of these should be given more weight in the final result
than the second. In order to do this Link gave each test a proportional value by dividing each correlation by the sum of all the
correlations of the series. In the above instance, for example,
this gave Q.30, 0.32, and 0.38, that is to say, the first test has a
weight of 30 per cent, in the total, etc. In the typists' test the
substitution test is prorated 10 per cent., the spelling and completion tests 25 per cent, each, and the typewriting test 40 per cent.
In the case of a new typist just out of school, who has had little
practice as yet, it is advisable to count the substitution test as
15 per cent., the spelling and completion tests as 30 per cent, each,
and the typing test as 25 per cent. The result of each test is then
multiplied by its proportional value, e.g. if the score in the sub-

— 29 —
stitution test is 0.80, in the spelling test 0.83, in the completion
test 0.84, and in the typewriting test 0.81, their prorated values
would be respectively: 0.80 X 0.10 = 0.08; 0.83 x 0.25 = 0.20;
0.84 x 0.25 = 0.21 ; 0.81 x 0.40 = 0.32. The sum of these results
is 0.81, which is the rating for the whole series of tests. A good
typist should obtain from 0.60 to 1.00.
Burt (') also undertook experiments to find tests for selecting
shorthand-typists who are already trained in their work and are
applying for employment. With slight modifications the same
tests could be used for vocational guidance of young people on
beginning their training in order to discover whether they are
likely to succeed.
Burt's method of procedure was as follows:
(1) A preliminary series of tests was carried out on a small
group of persons (shorthand-typists engaged on scientific work and
students of psychology with a knowledge of typewriting) in order
to eliminate or modify unsatisfactory tests and to determine and
improve the general procedure for the tests retained.
(2) Of the tests thus selected the more important (chiefly
those dealing with educational attainments and general intelligence)
were then applied to about a hundred school-children between 13
and 15 years of age, most of them in attendance at secondary
schools in London. It proved impossible to find a sufficient number
of adults for this purpose. The object of these experiments was to
grade the various tests in order of increasing difficulty.
(3) The tests were then arranged in four groups:
(a)
(b)
(c)
(d)

graded tests of intelligence ;
graded tests of educational attainments;
tests of linguistic ability and general information;
tests of typewriting and shorthand.

(4) Thirty typists in a large office than took these tests. The
supervisor of the office arranged the subjects tested in order of
merit according to their general ability as clerical workers, according to their ability in shorthand writing and in typewriting. The

(!) Cyril BURT: Tests for Clerical Occupations, in the Journal of the National
Institute of Industrial Psychology, Vol. I, No. 1, Jan. 1922,'pp. 23-27; No. 2,
Apr. 1922, pp. 79-81. London.

— 30 —
correlation between the tests and the three rankings wer£ then
calculated. Among others the following results were obtained:
Test

General intelligence
Typewriting
Shorthand

Mixed
sentences

Opposites

Arith- Com- Defini- Synometic pletion tion
nyms

Analogy

Spelling

0.44
0.37
0.45

0.52
0.35
0.32

0.41
0.28
0.33

0.32
0.31
0.35

0.35
0.44
0.43

0.48
0.46
0.49

0.36
0.27
0.36

0.31
0.33
0.39

The highest correlation (0.79), as might be expected, was that
between general ability for clerical work as estimated by the
supervisor and general intelligence as indicated by the tests. The
lowest (0.60), which was nevertheless significant, was that between
ability in typewriting and the results of the typewriting tests.
Burt states that this is consistent with what he felt in the course
of the investigation, namely, "that ability in typing was far the
hardest quality to test satisfactorily" ('). The tests which gave
the highest correlation with ability in typewriting were sentence
completion (0.46), mixed sentence test (0.37), and the spelling
test (0.44); these results, Burt adds, "seem to throw a little light
on the nature of the mental capacities involved" (2).
The important investigations of Muscio and Sowton (a), although
their object was to arrive at vocational tests for typewriting, deal
chiefly with questions of method, notably the problem of constituting homogeneous groups of subjects for purposes of vocational
tests. They draw attention to the fact that efficiency in an occupation is not a measure ot capacity or natural aptitude for it.
There are a great many factors, such as practice and application,
which affect efficiency and may conceal actual capacity. This
fact presents a great obstacle to the construction of vocational
tests. If an employer is asked to grade the workers who have
taken the tests, he does so according to their efficiency or commercial value, while what interests the psychologist is ability or
psychological capacity.
f1) Journal of the National Institute of Industrial Psychology, Apr. 1922,
p. 80.
(2) Ibid. p. 81. Cf. the results of the present author's experiments, p. 33
below.
(8) B. Muscio and S. C. M. SOWTON: Vocational Tests and Typewriting, in
the British Journal of Psychology, Vol. XIII, Apr. 1923, pp. 344-369. Cambridge.

— 31 —
The experiments were carried out on approximately 220 subjects,
belonging to different groups — typists in various offices, pupils
of commercial schools, etc. The tests used were as follows:
(1) attention;
(2) memorising;
(3) immediate memory span;
(4) speed of reproduction of associations;
(5) intelligence;
(6) tactile sensibility;
(7) motor capacity;
(8) acquired associations (knowledge).
The correlation between the ranking of the subjects in the tests
and t h a t in typewriting efficiency varied from one group to another
for a single test. For example, the spelling test gave the following
correlations :Group
Correlation coefficient
A 2 a.
A 2 b.
A3.
G.
C2.
C3.
Dl.
D2.

20
14
18
60
22
30
12

young women
young women
young women
girls 15 to 21 years of age
women 22 to 36 years of age
girls under 21
pupils who had learned typewriting
for about 1 months
14 pupils who had learned typewriting
for about 11 months

0.45
0.81
0.58
0.15
0.28
0.16
0.68
0.13

The extent of the variation in these figures shows t h a t it is rash
to draw conclusions from the coefficients of correlation obtained
from a single group of subjects, unless the group be very much
larger than can generally be obtained in practice, i t is suggested
that the chief cause of the variation was that the efficiency gradings
for typewriting were not determined by typewriting capacity alone,
but were influenced by other factors which differed in different
groups of subjects. A further cause of variation was the varied
nature of the groups.
Muscio and Sowton conclude that coefficients of correlation,
even when obtained from experiments on several groups of subjects, do not indicate with any precision to what extent the capacity
tested by the test is or is not essential to the occupation. They
suggest that "useful results might be obtained for vocational
purposes by giving more attention to the theoretical analysis of
occupations (such analysis being carried out with the purpose of
arriving at the elementary functions exercised in the occupations)".
They recommend experimental research of the type which they
undertook, as likely to yield useful suggestions. The results would

— 32 —

prove nothing finally, and not much importance should be attributed to correlation coefficients, but they might in some cases
confirm theoretical analysis. The results of theoretical analysis
of the operation of typewriting were confirmed by the results of
the tests of immediate memory span (sentences) ; a test of attention;
and two intelligence tests (directions and completion). Muscio
and Sowton suggest that these tests might be useful for vocational
guidance in typewriting.

Ill

The Author's Investigations

The foregoing sections have shown that, although a number
of interesting investigations have been undertaken, experiments
to determine the nature of typewriting work are relatively few
in number. The investigations described in the following pages
were undertaken in order to obtain additional information and
thus to contribute to the study of the subject. It was also thought
advisable to control one or two results already obtained which
appeared particularly suggestive. The investigation was conducted
by means of experiment. Through the courtesy of the authorities
of the International Labour Office it was possible to conduct
experiments on 28 typists belonging to the French Typing Branch
of that Office. A threefold advantage was thus secured :
(1) the subjects formed a homogeneous group working under
identical conditions and employed more exclusively on typewriting
than is the case in the majority of offices;
(2) the experiments were carried out on all the subjects under
the same conditions;
(3) the subjects were ranked by their supervisors, which provided data essential for controlling the tests. Two independent
rankings were obtained, differing in several respects; these were
of great assistance.
The experiments were begun in the spring of 1922 and completed
in the autumn of the same year. Some of them involved the use
of apparatus (e. g. tapping and the d'Arsonval chronometer). In
one such case the apparatus was taken to the International Labour
Office, but in the majority of cases the typists came to the Psychological Laboratory for these special tests. The other tests

— 34 —
were conducted at the International Labour Office, where a room
was placed at the disposal of the examiners (').
As the building in which the tests were held was of wood, it was
impossible to obtain as much quiet and seclusion as in the laboratory. Further, as the typists came to the tests direct from their
work many of them were either hurried, having a great deal of
work to do, or more or less tired. In many respects the conditions
under which the experiment was conducted were not ideal and no
doubt were a source of more or less inevitable errors.

TESTS

The choice of tests was based on information previously obtained
by investigation of the typist's occupation, information which, as
the examiners were well aware, was unsatisfactory, but the object
of the tests was precisely to complete this information. In addition,
certain experiments already made by Lahy were repeated in view
of the interest attaching to them. (These tests are marked in the
following list with an asterisk.)
After certain preliminary attempts in order to perfect the
method of procedure, the following series of tests was finally
adopted :
(1) Tactile Sensibility. Nine of Binet's cards were used as
Weber dividers, the distances between the points varying by 5
millimetres (5, 10, 15, etc.). The experiment consists of pricking
the right forearm of the subject 45 times, 5 times with each card (').
(2) Muscular Sensibility. The subject has to arrange six
(') I am glad to take this opportunity of thanking the heads of the French
Typing Branch, who did everything in their power to assist the work of the
examiners, for the great courtesy shown us.
(*) The limen of sensation was determined by a method used by Professor
Claparède. The frequency with which the subject feels the two points of the
dividers, and not merely one, is noted for each distance between the points.
The limen is taken as the distance between the points corresponding to a
frequency of 50 per cent. This is easily found by constructing a graph with the
distances between the points of the dividers as abscissae and frequency as
ordinates. If, as in the above experiment, the dividers are applied five
times for each distance between their points, the limen is the distance
between the points corresponding to a frequency of 2'/ 2 ; it is found by
drawing a horizontal line from the point on the axis of the ordinates representing 2'/,. The limen is given by the point of intersection of this line with the
frequency curve. If the lines intersect at more than one point, the limen
is given by the point of intersection with the largest abscissa (i.e. the
greatest distance between the points of the dividers).

— 35 —

graduated weights, differing obviously between themselves, in
order of weight as quickly as possible. The time taken by each
subject is noted.
(3) Memory. Form Recognition ('). The subject is shown
for one minute a board on which are shown 25 simple figures
mainly composed of straight lines. The subject then has to
identify these figures on another board where they are mixed
with 25 similar figures. Two series are used. The number of
figures correctly identified is noted and the number of errors
subtracted.
(4) Memory. 15 Words (*). The examiner reads a series of
15 words, one every two seconds. The subject then writes down
those he remembers, regardless of order.
(5) Memory. Repeated Sentences. As auditory memory is of
great importance in typewriting this test was compiled to supplement the 15-word test. The examiner reads sentences varying
from 10 to 40 syllables in length. The subject has to repeat them
exactly one by one. The examiner notes the longest series of
which the subject succeeds in repeating at least two sentences
(each series includes from 4 to 5 sentences).
(6) Attention. Cancellation of Signs (Barcelona form of the
test of Toulouse and Piéron) (3). The subject is given a sheet of
paper bearing printed squares each of which has on its outline a
short line pointing towards the centre and in a different position in each square (the Barcelona model has rather smaller squares
than the original form of the test). The subject has to strike
through those squares where the line is in one of four positions
given in a sample put before him. The examiner notes both the
quantity of squares cancelled, which indicates rapidity of work,
and the quality of cancellation (the proportion of squares correctly
cancelled to the total number to be cancelled).
(7) Permutations ('). This test is really one of intelligence, and
in Professor Claparede's words a test of "inventive intelligence".

(') Mary T. WHITNEY. The Empirical Study of Certain Tests for Individual
Differences, in Archives of Psychology, 1911. New York.
(') CLAPARÈDE: Tests d'aptitude. Geneva, 1920. Reprinted from Archives
de Psychologie, Vol. XVII, No. 68, Dec. 1919.
(3) TOULOUSE and PIÉRON : Technique de psychologie expérimentale, Vol. I,
p. 266. Paris, 1911.
(*) CIAPAREDE: op.

cit.

— 36 —

The subject has to write as many permutations as possible of the
four letters, a, b, c, d, in one minute.
(8) Mixed Sentences (Jean Jacques Rousseau Institute).
The subject is given five sentences in which the words have been
jumbled and has to re-arrange them as quickly as possible (e.g.
"les grandes dans triste il est des malheureux bien villes de penser
abri que sans sont" — Il est triste de penser que dans les grandes
villes bien des malheureux sont sans abri). This is an oral test.
The subject has the sentence before him and the examiner notes
the number of sentences re-arranged in one minute.
(9) Sentence Completion (Jean Jacques Rousseau Institute).
The subject has to read sentences in which the vowels are omitted
and represented by points. This test is comparable to Link's
sentence completion test (e.g. L.s r.v..r.s v..nn.nt d.s m.nt.gn.s
.t s. d.r.g.nt v.rs 1. m.r: Les rivières viennent des montagnes et
se dirigent vers la mer). The examiner notes the time taken and
the number of sentences read correctly.
(10) Reaction-time (with d'Arsonval chronometer). Fifty auditory and 50 visual reactions were measured and the mode (most
frequent reaction-time) and the average variation were noted.
(11) Tapping ('). The subject taps as rapidly as possible
for one minute, first with the right and then with the left index
finger, on a key attached to a counter. The number of taps is
noted.
(12) Muscular Strength (Collin dynamometer). Two grips
with the right and two with the left hand were measured. These
were taken with the subject standing, his arm extended, at intervals of about 10 seconds.
(13) Cutting (Jean Jacques Rousseau Institute). The subject
is given a sheet of paper on which are printed 10 lines, 5 straight
and 5 wavy, 17 centimetres long, narrowing to a point from their
base, which measures one centimetre. The subject has to cut
along these lines as rapidly as possible without going outside the
printed area. The time taken in cutting is noted.
(14) Speed in Writing^). The subject writes in ink as many
times as possible in one minute the sentence "La terre est ronde".
(') First used by GILBERT, and especially studied by F. L. WELLS; see
American Journal of Psychology, 1908 and 1909.
(2) CLAPABEDE: op.

cit.

— 37 —

(15) Spatial Estimation. The subject is given a manuscript
letter and the heading and first words of the same letter typewritten. He has to estimate by eye the space which the whole
letter would occupy when typewritten. The error in centimetres
is measured.
(16) Spelling. A continuous printed text, without punctuation
and with 18 errors (grammatical and spelling mistakes), is given
to the subject. He is allowed three minutes to correct the misspelt words and insert punctuation. The examiner counts the
number of corrections properly made.

RESULTS

The 28 typists (') who had undergone these tests were then
ranked on the results of each test as shown in table I (s). This
table includes in the three right-hand columns the ranking of the
subjects by their two supervisors and the ranking according to
intelligence given by one of these supervisors.
The object of these rankings, which were entirely independent
of the tests, was to check the results of the latter. The only
tests which were regarded as of value in diagnosis were those
the results of which ranked the subjects more or less similarly
to their ranking by vocational efficiency. The degree of agreement
between the two types of ranking is shown by correlation coefficients which are given in table II.
In the first place it may be pointed out that series II on the
whole shows a higher correlation with the test results than series I.
In the three series there are from five to seven coefficients which
may be regarded as significant, in view of the conditions under
which the tests were held. Among the insignificant results may be
noted the very low, if not negative, correlation shown by the test
of reaction-time (3). Tests of motor ability also generally showed
low correlation.

(') I t was not possible to apply all the tests to all the subjects.

(2) W h e r e more t h a n
b y t h e middle figure of
first are graded as 1.5,
m o s t of t h e columns in

o n e subject h a d t h e s a m e ranking this is expressed
t h e group which t h e y form, e.g. two subjects r a n k i n g
t h r e e firsts as 2, e t c . This explains t h e fact t h a t i n
table I all t h e 28 possible rankings do n o t occur. Cf.

F O N T E G N E a n d S O L A R I : Le travail
(3) Cf. L A H Y : o p . cit., p . 69.

de la téléphoniste,

p . 113.

Geneva, 1919.

gSiisiKj><!^^CiHW5dO^O^gt H W'- , M ÄC>^HÖOW>
<-*.*»pOfO"*
h*.
.-»-tO —
•-*•-»•.
000)U^(»>«C^rO>JOOrOOOJO;H.O
O V
CJ*CJ»
CJ»
CJ*
ÜIW

•*•
o - 4 « 4
b*

. t o •-*>-»•
H».—.H»,
H».
fo
©coco-3-v3CJ*coroco'-*,—'-«]|
b*bi

CJ»CJ»CJ*eJ*

CJ»

CJ*

•-».,

CJ»

CJ*

CÍ*

oo t o - a ^ i c o ^ J c o * * - J c o c o c o 1 * - r o t o
CJ»CJ*CJ*CJ»

CJ*CJ»

CJ»

er*

»*

>-*•

-JOSCD

CJ»

Memory for phrases

u » en e n

H*.
-J

* * . _ > . , - * . K > t o •-*-**-cO©*»tO*«C0Gö

Cancellation:
quality

1 o s - J O Í » O -»> o s er* o s - i -*»

cnb»

M.^K>

CJ»

**-

CJ»

Permutations

ej»cncj*

. •-io

0 0 0 ( C - * - « * - J ^ ï O * » |

CJ»

15 words

CJ* CJ»

Cancellation :
quantity

CJ*

M-

*-,ír-*-O0

CJ»

Form recognition

«••-*.
to>-».(o
M»
e o o t o o o o CJ» c j * en * ^

.— M .

rf> I C O t O C O * > i t > I

tO t O • - .

CJ»

to
I

. .-»-»*
CSrOOO

po t o t o c o t o ~ J

Ci*

CT*

CJ»

Cj* CJ» CJ* CJ* CJ*

Weights

1 * - CJ» - 4 * • t o >-*. oo * - c o co - J

toto»»
to
to
>-.
¿a
•*»*.»*.
co t o o o C J * o < • * • * * - a c o * - t o <=
o s ^ ^ j
b*
I
I

t o t o CJ» •-». * o - j

to-».
t o -a 1
en

CJ» CJ»

b»CJ»CJ»

u*u*

•—
»o
>-».
fo
f o ^
to
•*tO^*.K)Ä-<y*CO-40COC»CJ'

Tactile sensibility

•-••-*•
«-••*•
to
•-»•
c o o s ic- o s CJ» •-»• oo c o » o co o o

CJ*

f O f O M f c ^ ^ M f c f O ^ . , 0
Ü1ÜHHCH

* * ,
>*rS,
1 CO 0 0 CO 1 O O r O l
b*b*CJ»
CT*

N*fc*.
to
>P- CJ» o 1 U * J U
uicn

**. Ï O _ * ^ PO
K>
>-*to , w
K> ,
w o o 3 - * o o K K D - » - í o a i O )
>e-coro |
Or

, .

Subject

•-•

Mixed sentences

CD ( O 1 * • * « •

CJ»

CJ*

i * f O i i»».i-*toto
•—•
**.
**.
toto
to
to
tO"-».(o
o * o • * • e o oo CJ» » o •-»• o s t o c o ~ j 1 c o * 4 CJ» * » H», o s «•* c o * » * * CJ* - * i —a t o c o
CJ*

CJ*

CJ»

CJ*

Sentence completion

CJ»

W M - *
<-»• »O »O (O • * • <-fc ( O l O , <-*.»^
<-». t-fcJsO
f».pO
H"
N*
H*. CO CO CO O i-». î O CO CO CO CJ» <-*
CO CO OO CO CO CJ* CO CO » J CO O OS CO O l CO
CHCJ»

CJ»CJ*CJ»CJ»

CJ»CJ»CJ»

CJ»

M.M.
M-M.
M - M W »O . * - t O > [ OO*-COCDCJ»CD-4CO>*CTSOO©
»*toco

-aosto

CJ*CJ*

COCJ*

Reaction-time
% av. var.

1 -Jto 1

Reaction-time
mode

* - «•*•
^50>*•

CJ*

toto«-»- I t o f O * - * - * c f t t o o s N > o s * * - c o t o |
CJ*

CJ»

co t o t o

1 to*oco

CJ»

CJ*

M-i-ktO
.-.•*.
(O—.
( O , tOtO
«-*
**r5Hi.hktO'*M.
co c o CJ» CJ* os - J - J * » c o CJ» **• t o I * o CJ» * * t o * • o o o •*• t o c o - J CJ* O CO
CJ*

çj*

CJ»

»o
»o•*• >oi-*to
i-».ro
osco*»cocj»toto*-cDrffCOiP*

b»
V

Ù<

, to
•*
OlP•^^

•-•«-*
f o i ^
•*
**.
*»-joo>ts»—i-^**cjiic-cocj'

b» '

. •-»
H*
en os o

CJ»

COOStO-^rf>'-*HfctOCJi0OCO-J

'
o s * « co

- » . i-*. • - . * * • * . i-», to t o
toto
CO © Ï O * - - J CO * * O CO CJ* CO

Hfc**(0**

. - " - ^
, ^ ^ I O
*-coit*
*>i-*.co
' b»
b»
CJ»

OOSCJ»

- J O " ^

"*•.
^^•*•j
'

CO*»

Hk(0

^tOl-^tO

] it> * - e o o s c o K> c o •-»• c o c o t o - J - * c o t-^ co * * * - c o o o e o o s |

FfcMh
tO-^CJ»

HfcM»^*.
M.
toto
•*
to
to
^^to>-'^^^s
<— - J CO OS OS CO OS CO * * O t O • * • » CO • * • O OS O * • • * • - J CD

b»
b»

b*b*
b*

b»
b»cjib»

j
i

f!npfflripnt
w u c m w c u i

of asymmetry
Cutting :
straight lines

1
|

fíiittinff
\JUtblllg

wavy

b»

• * • * - .
I

b»

CJ»CJ»

»*h*.

W r i t i n g speed

i-* * * •*.•*. ^ t o t o
^^to
oococc»*c»'-fc'*-CJ'05coo

<-»•
to.«-»-—
M.^^
( o —
« i < 4 o
t-fccooo*-H«-oscjstoco>-*oocD
1
b»

eOtOCO

CJ*

to-*-—
to>—ro
, H*
CJ* ^11—CJ» O H*. OS OS t 3 CJ» CJ* t O
OOCOOO

**.
O

CJ»

M.
^ » k
tonato
tococo«-i*'->-»-cj»coroco*'-to

Spatial
estimation

*
.

lines

Tapping, right
Tapping, left
Spelling

b*
CJ»

Punctuation

>-*
•*.•*.
t o — »-tototo
^ t o
toi-*
t o ^ ' - ' t o ,
B^i^cOCJ»COCDO«<'CO"*COCO<h}>P*CJ»00*-lrJtOtOOSOOSCr<^CO>-*' 1

^.M-H*.
•*.—••*>*.
^ t O t O » ^
i--tO
to
t0)-*tO»O ,
CDi*tOi^<I«J05^00SCOCO<IC7'CJ»CJieOOOtO>-*OsOOhD*»COCS*-- 1

I
II
HI
(Intelligence)

Supervisor's
ranking

h*, i-fc (O I-*-iO
H*.
(Ok*.(Oi-k*k**(-».tO
t0<^tO
tOtO
C0t0C0*J-*C0'-*.C0-Ji-*CJ»CJ»~a-JOC0'4O-^J0SCJ«C000hfc-4CJi"*'Q0
CJ»
CJ* CJ* CJ* CJ» CJ* CJ« CJ»
CJ» CJ* CJ» CJ* CJ* CJ* CJ»
CJ» CJ* CJ»
CJ»CJ»

— 39 —

TABLE II.

CORRELATION BETWEEN TEST RESULTS AND SUPERVISORS'

Ranking I

Ranking II

(based on vocational efficiency)

RANKING

Ranking III
(intelligence)

1
Spelling
0.63
Punctuation
0.47
Permutations
0.42
Cancellation : quality 0.42
Form recognition
0.39
Sentence completion 0.35
Cancellat. : quantity 0.32
Memory for phrases 0.30
Reaction-time :
%
average variation 0.26
Weights
0.24
Tapping: right hand 0.20
Tactile sensibility
0.18
Cutting: wavy lines 0.18
Asymmetry of hands 0.17
15 words
0.11
Writing speed
0.10
Cutting: straight lines 0.08
Tapping: left hand 0.06
Reaction-time : mode 0.05
Mixed sentences
0.04
Spatial estimation
-0.05

0.62
Spelling
0.62
Memory for phrases
0.59
Permutations
0.48
Cutting: wavy lines
Cancellation : quality 0.47
0.44
Sentence completion
0.42
Punctuation
0.39
Cancellat.: quantity
0.37
Writing speed
Cutting: straight lines 0.37
0.32
Mixed sentences
0.30
Form recognition
0.26
Tactile sensibility
0.25
15 words
0.22
Tapping: right hand
0.21
Weights
0.15
Asymmetry of hands
Reaction-time : %
average variation
0.09
Spatial estimation
0.01
Tapping: left hand -0.03
Reaction-time : mode
-0.04

Memory for phrases 0.77
Spelling
0.66
Permutations
0.56
Sentence completion
0.46
Writing speed
0.44
Punctuation
0.44
Asymmetry of hands 0.40
' Cancellation : quality 0.38
Tapping: right hand
0.33
Cancellat.: quantity 0.30
15 words
0.26
Mixed sentences
0.26
Form recognition
0.19
Spatial estimation
0.15
Weights
0.09
Cutting: wavy lines 0.08
0.05
Tactile sensibility
Cutting : straight lines 0.04
Reaction-time : mode 0.01
Reaction-time : %
average variation -0.01
Tapping: left hand -0.07

The six best tests in series II, i.e. those with the highest correlation, were selected. These were spelling, memory for phrases,
permutations, cancellation (quality), sentence completion, punctuation. (The test in cutting wavy lines was ignored, as the correlation
in series II, though relatively high, does not accord with the low
correlation of the same test in series I.) The correlation of these
six tests with the ranking of the supervisor was then calculated,
the coefficient being 0.84 (see table III). The correlation of the
same six tests in series I is 0.67. The correlation of the six best
tests in series I — spelling, punctuation, permutation, cancellation
(quality), form recognition, and sentence completion — is 0.62.
These co-efficients are higher than those for most of the tests
taken singly. This result appears to confirm Lahy's conclusion that
"the capacity of a subject depends upon a group of indications
much more than on very marked superiority in one of the abilities
investigated'^ 1 ).
t1) Laky: op. cit. p. 69.

— 40 —

TABLE
TOTAL

TEST RESULTS COMPARED W I T H S U P E R V I S O R S ' RANKING ( ' )

Series 1
Subject

I
S
R
M
F
W
G
T
Zb
K
N
U
Z
D
Za
B
H
Q
V
0
C
E

IH.

Series 2

Series 3

SuperGrading SuperGrading
visor's Subject by test
by test
visors'
Subject
results ranking I
results ranking II

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15.5
15.5
17
18
19
20
21
22

Correlation

2
1
3
7
6
8
9
12
10
11
5
13
16
14
4
18
22
19
17
20
15
21
0.84

I
1
2
S
3
R
4
M
5
F
6
W
7
G
8
T
9
Zb
10
11
K
12
N
13
U
14
Z
15.5
D
15.5
B
17
18
Za
19
H
20
Q
21
V
22
0
C
ECorrelation

6
1
5
12.5
8
12.5
12.5
15
3.5
9.5
9.5
3.5
12.5
7
17
2
20
20
17
22
20
17

0.67

I
S
K
R
W
T
G
M
F
D
Zb
U
N
H
B
Z
Za
V
Q
0
E
C

Grading Superby test
visor's
results ranking I
1
2
3
4
5.5
5.5
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

Correlation

6
1
9.5
5
12.5
12.5
15
12.5
8
7
3.5
3.5
9.5
20
17
12.5
2
17
20
22
17
20

0.62

To sum up, therefore, it was found that in a certain group of
tests the good typists were the most successful and the bad ones
the least successful. These tests were the following: spelling and
punctuation, permutations, sentence completion, memory for
phrases, form recognition, and cancellation.

(*) Series 1 gives the ranking in the six tests which had the highest correlation with the ranking of supervisor II. Series 2 shows ranking in the same
tests with the ranking of supervisor I. Series 3 shows the ranking in the six
tests which had the highest correlation with the ranking of supervisor I.

Vocational versus Psychological Tests ; Moral Qualities

Certain objections will naturally arise in the minds of those who
have read the preceding sections, and one of these may be dealt
with briefly here. Would it not be better to use vocational tests
(copying letters on the typewritter, etc.) which appear to measure
ability in typing more directly than the analytical (*) tests used in
the above experiment ? This question leads to another. The chief
object of the investigation was to find tests which give information
on the future capacities of a subject, for example, of a child beginning its training. Which are the most useful tests for this diagnosis,
vocational or analytical tests ?
Chapman (*) investigated the process of learning in typewriting
and came to negative conclusions with regard to vocational tests.
He conducted experiments on classes of boys and girls who were
learning typewriting. They took tests once a week during the
school year from the stage at which they had had 20 hours' practice
until they had completed 180 hours' practice. The test involved
copying texts of the same length in five minutes. The number
of words correctly typed was then noted. The average curve
obtained by Chapman is given on page 43.
Chapman endeavoured to discover up to what point it was
possible to predict the final ability of an individual on the basis
of his initial ability. He calculated the average score of each
subject at three different stages of the experiment and ranked his
subjects according to these averages.
(') For the explanation of these terms see INTERNATIONAL LABOUR OFFICE:

Problems and Methods of Vocational Guidance, by Professor E. CLAPARÈDE,
pp. 43 and 44. Geneva, 1922.
(2) J. Crosby CHAPMAN: The Learning Curve in Typewriting, in the Journal
of Applied Psychology, Vol. I l l , No. 3, Sept. 1919, pp. 252-268. Worcester,
Mass.

— 42 —

TABLE IV. EFFECTS OF PRACTICE IN TYPEWRITING ON TEST SCORES
AND RANKING OF 1 9 SUBJECTS

Ranking

Average score
Subject

A
B
G
D
E
F
G
H
I
J
K
L
M
N
O
Q
R
S
T

Period of practice in hours

20-35

75-91

136-151

20-35

64
65
80
52
45
61
60
76
82
48
51
46
52
113
42
83
53
54
77

230
206
219
176
140
146
184
200
205
182
203
198
127
211
162
196
184
157
227

256
227
256
218
228
180
271
255
251
223
228
213
148
244
226
272
214
231
274

8
7
4
13
18
9
10
6
3
16
15
17
13
1
19
2
12
11
5

76-91
1
5
3
14
18
17
11.5
8
6
13
7
9
19
4
15
10
11.5
16
2

136-151
4.5
12
4.5
15
10.5
18
3
6
7
14
10.5
17
19
8
13
2
16
9
1

It may be pointed out that each average represents from three to
five measurements over a period of 16 hours' practice. No test was
given before the subject had had 20 hours' practice, and the test
was strictly vocational. The correlations between initial ability
and final ability (0.65), between initial ability and intermediate
ability (0.66), and between intermediate ability and final ability
(0.66) are statistically satisfactory; that is to say, if a school had
to select 50 pupils from 100 applicants who had each had 20 hours'
practice in typewriting, it might place some reliance on the results
obtained by the pupils at the end of this period. Vocational
guidance tests, however, are intended for the guidance of the
individual before he begins his training in order to avoid hesitation
and false starts and not to select groups of workers who have
completed their training. One may well agree with Chapman
"that the use of the initial ability as an indication of future
skill would be very precarious. It is likely that a mental test
of 15 minutes would give a better selection for the purpose of
even the narrow function of typewriting".

— 43 —

AVERAGE CURVE

280240200
I60
120

8O
4.O

O
2O

10 O

12 O

14rO 16O lô<D

The author of this study repeated Chapman's experiment on a
group of school-children from 13 to 14 years of age who had never
used a typewriter. They were required to type various short
phrases and the alphabet five times in succession, the examiners
noting the time taken on each occasion. Table V shows the
ranking of the subjects on the first and last repetition of the
alphabet (').

(') Cf. INTERNATIONAL LABOUR OFFICE: Problems and Methods of Vocational

Guidance, by Professor E. CLAPAREDE, p. 76. Geneva, 1922.

— 44 —
TABLE V. EFFECTS OF PRACTICE IN TYPEWRITING ON TEST
SCORES AND GRADING OF 1 8 SUBJECTS

Subject

Time taken for
first alphabet
(seconds)

Ranking

Time taken for
fifth alphabet
(seconds)

Bur.
Mar.
Val.
Fin.
Tac.
Bis.
Mon.
Cor.
Des.
Bar.
Flu.
Bad.
Rad.
Egg.
Pau.
Bro.
Dei.
Vit

50
51
56
72
73
75
81
85
86
95
101
102
106
108
113
130
171
227

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

57
33
31
35
27
51
34
51
45
51
31
40
37
39
48
64
92
130

AVERAGE RESULTS AT DIFFERENT STAGES OF THE

Hours'
practice

25 35

Ranking

15
4
2.5
6
1
13
5
13
10
13
2.5
9
7
8
11
•
16
17
18

EXPERIMENT

95 105 115 125 135 145 155 165 175

Words

typed
in 5
54 81 105 128 151 174 191 216 215 222 216 228 230 235 246 245
minutes

These data corroborate Chapman's conclusions, and it is clear
that it would be rash to advise a child to become a typist on the
basis merely of his initial capacity measured by a typewriting test.
The vocational test seems inadequate. In the particular case
of typewriting the work depends to a very large extent on practice,
but motor rapidity is not the only ability involved. Other functions come into play which depend less on practice and more on
the individual, and it is these functions which the analytical test
attempts to measure.
There is another problem which arises (l). In practice it is
impossible to measure moral qualities by tests, and' it is obvious
(') Cf. Problems and Methods of Vocational Guidance, p. 76.

— 45 —

that it is difficult to obtain precise information from a subject
regarding his moral qualities during an interview or a psychological
examination. Information given by other persons is hardly more
conclusive Nevertheless ambition, orderliness, patience, initiative, etc., of vital importance in any trade, are sometimes as important as vocational abilities strictly so called. In effect two
factors make a good worker: abilities and moral qualities. The
value of tests is that they detect and measure special abilities, but
they are practically incapable of detecting and measuring moral
qualities. Consequently, psychological tests cannot claim to be
an infallible method of vocational guidance and selection. Nevertheless they are not as valueless as might be thought in the sphere
of moral qualities.
It must be recognised that a moral quality is not absolute but
rather relative, depending on many external circumstances, physical, economic, or social. Examples could be multiplied. An
employee may show a great deal of initiative if he occupies a post
for which he is well fitted, but he may fail to show any initiative
at all if he has to do work for which he is not fitted. If a man
is employed on work which does not interest him and for which
he has not the necessary abilities, he can hardly be expected to be
accurate and conscientious. People who are interested in all kinds
of work, or dislike all kinds of work, are rare. In the great majority
of cases moral qualities are influenced by two conditions: (a) a
liking for a particular kind of work (there are people engaged in a
trade or occupation who like their work for its own sake); (b) a
liking for work for the sake of the money to be earned by it (e. .g
unskilled workers). In both cases it is important that the work
should be adapted to the capacities of the worker; it is very probable that he will then show application and energy. On the other
hand, if he feels himself incapable of doing his work he may become
discouraged, lazy, and indifferent, and will be ready to throw up
his work at the first opportunity.
Thus, in attempting to determine the work for which each person
is best fitted, psychological tests also assist in solving the complex
problem of the influence of moral factors in employment.

— 46 —

The experiments made up to the present regarding ability in
typewriting show that no special test has been found for detecting
this ability in any certain fashion or for foretelling success in the
occupation. It has nevertheless been shown that good typists
obtain better results in certain tests than in others. Those tests
which correlate with ability in typewriting measure higher or
intellectual factors rather than manual or motor functions. This
is therefore one of the types of work involving manual operations
but depending to a large extent on intellectual functions. In
practice an applicant would be advised to take up typing as an
occupation much more on the ground of his intellectual faculties
(memory, comprehension of printed matter, spelling, and grammar)
than according to the results of tests measuring motor functions.
It is obvious that a series of tests is necessary to measure these
different faculties and that guidance cannot be given on the results
of a single test. In conclusion, a quotation may be given from
remarks of Professor Imbert of Montpelier in the Année psychologique of 1912: "We have a right to affirm that in general the
superiority of one worker over another, even in wholly mechanical
work, is due far more to intellectual than to physical qualities."

BIBLIOGRAPHY
This list does not include books or articles already cited in the text.
BOOK, W. F. The Psychology of Skill, with special reference to its Acquisition
in Typewriting. University of Montana Publications in Psychology, 1908.
BURGHAGEN, Otto. Der perfekte Maschinenschreiber. Hamburg, Verlag der
Schreib-Maschinen Zeitung. 1907.
FRANKFURTER, W. Arbeitsversuche an der Schreibmaschine. Psychologische
Arbeiten, VI, p. 419. Leipzig, 1914.
J. M. LAHY : Recherches expérimentales sur la frappe du dactylographe. A paper
read at the International Congress on Psychology held at Oxford from
28 July to 2 August. It summarises the practical results of the author's
experiments, which will be published later in detail.
HEINITZ, W. Untersuchungen über die Fehlleistungen beim Maschinenschreiben,
in Zeitschrift fur angewandte Psychologie, Vol. XVIII, 1921, p. 33. Leipzig.
HERBERTZ, R. Zur Psychologie dés Maschinenschreibens, in Zeitschrift für
angewandte Psychologie, Vol. II, 1909, p. 551.
Monographie sur la profession de dactylographie. Marseilles, Bureau
d'orientation professionnelle. 1921.
HILL, Rejall and THORNDIKE: Practice in the Case of Typewriting, in Pedagogical Seminary, Vol. XX, Dec. 1913.
SWIFT, B. J. The Acquisition of Skill in Typewriting, in Psychological Bulletin,
Vol. I, 1904, p. 295.
WELLS, F. L. On the Psychomotor Mechanisms of Typewriting, in American
Journal of Psychology, Vol. XXVII, 1916, pp. 47-70.