Daniel Starch.

Educational psychology online

. (page 26 of 41)
Online LibraryDaniel StarchEducational psychology → online text (page 26 of 41)
Font size
QR-code for this ebook

when he begins to learn to write. At any rate its importance is
rather secondary in the writing process.

The chief elements in the writing process are those connected
with the steps (5) to (8), particularly those connected with the
steps (5) and (6). Development of skill in any muscular move-
ment which is not instinctive or at least not as mechanically pre-
cise as an instinct, proceeds by trial and error. Attempts are made
at carrying out the desired movements. At first some of them
suceed, most of them do not; and through continued trials the
erroneous attempts decrease and the successful ones increase until
perfect control is established. The adult scarcely realizes the
utter lack of control in the early attempts on the part of the child
in making the writing movements. The nearest approach that
the adult can make toward realizing the actual difficulties of the
child consists in such an experiment as the tracing of an outline
as seen in a mirror. The spatial relations are so completely new
and different that a person has little or no conception of the direc-
tion in which to move. Figure 62 shows a reproduction of the
first tracing of a star outline as seen in a mirror. It does very
little good to reason about it. One may think he is going to move
in a certain direction but finds upon making the movement that
he is going in an entirely different direction. The child learns to
write very much after the same manner. He proceeds by trial and




error. The successful movements somehow become more deeply
fixed in the nervous connections and consequently more and more
numerous. The correct movements finally become associated
with the visual perception of form and direction so that the move-
ment can be carried out at will with precision and grace.

It is a matter of common observation that a child beginning to
write not only makes the movements very slowly but also with
much excess pressure. The latter point was investigated by Meu-
mann. He had children and adults write on a little platform which
was so supported that the amounts and changes of pressure during
the writing were transmitted to a light lever which traced them ac-

FiG. 62. — Record of tracing of a star-outline as seen in a mirror.

curately on a moving drum. It is natural to find that men tend to
exert a greater pressure than women. Gross, working in Kraepe-
lin's laboratory, found that men average nearly twice as great a
maximum pressure as women. These records also serve to illustrate
the fact that each letter has a fairly characteristic pressure rhythm.
Freeman and others have shown by exact methods that the rate
at which the writing point moves differs greatly in various parts of
a given letter. : In general the up and down strokes are the most
rapid while the sharp turns and angles are the slowest. The speed
curves of different persons writing the same letter consequently
show striking similarities.

Sex Diferences. The striking differences between the hand-
writing of different individuals has led many people to believe that


there was a definite relation between the handwriting and the per-
sonality of the writer. The greater pressure exerted by men as a
class suggests that the sex of the writer determines at least certain
characteristics of the writing. This has been investigated inde-
pendently by Binet, Downey, and Starch who find that untrained
subjects can determine the sex of the writer in from 65 to 75% of
the specimens (50% being pure chance). The author made a com-
parison of the writing ability of 2,113 ^oys and girls in the Madison
schools and found the differences exhibited in Figures 63 and 64.
It appears from these graphs that the median of the girls is above
that of the boys in both speed and quality, but particularly in
quality. The difference in speed is very slight.

Miss Downey ('10) attempted to determine sex differences in
handwriting by selecting from envelopes by a chance method 100
samples of men's writing and 100 samples of women's writing and-
by asking thirteen persons to judge whether a given sample had been
written by a man or by a woman. The thirteen persons made re-
spectively the following percentages of correct judgments: 60, 60,
61, 64, 66, 66, 68, 68.5, 70, 70.5, 71.4, 71.5, 77.5. These show
that on the average a judgment of sex as revealed in handwriting
is correct 67% of the times, or two out of three times. Miss Downey
also reports that the writing of the women showed less variability
and more conventionality than that of the men. Those samples of
writing by women wliich were celled masculine were generally from
persons accustomed to doing a great deal of writing.

Correlation of handwriting with other traits. A good deal of mis-
leading character interpretation is based upon various features of
handwriting. Experimental work needs to be done in this field, but
it is probable that there is nothing to the claim of graphologists
that handwriting reveals such traits as energy, clearness and sim-
plicity, vanity, or self-consciousness.

Gesell ('06) reports that there is a close correlation between
quality of writing and intellectual ability, but as a matter of fact
his results show, as pointed out by Thorndike, a correlation of only
about .30.

The author ('15) found for children a correlation of .31 be-
tween writing ability and general scholarship. Thorndike ('10)
reports that for adults the correlation between ^vriting and schol-
arship is zero. The j^robability is that the small correlation exist-
ing in the case of children is due to the fact of receiving instruc-
tion in writing and to the attention given to it. The better pupils



do somewhat better in writing because they probably pay more
attention or make more careful efforts. So far as adults are con-
cerned, poor handwriting is no indication either of high or low
intelligence, since the correlation is approximately zero.

Professional graphologists have claimed that a great number of
specific traits of writing are determined by corresponding traits
of character on the part of the writer. Thus the lines of writing
of ambitious persons are supposed to slope upward from left to
right. Hull ('19) investigated some of the more persistent claims by
computing correlations between the traits of character of 1 7 univer-
sity fraternity men, as judged by their fellows, and exact measures
of samples of their writing. The correlations in each case were
approximately zero, showing these claims to be entirely unfounded.

The Measurement of Efficiency in Writing

(a) Essential dements to be measured. The two important aspects
of writing that must be measured are speed and quality, including
under the latter legibility and form or beauty.

(b) Methods of measurement. Speed of writing is now generally
measured in terms of the number of letters written per minute.
Quality may be measured by cither one of several scales, the Thorn-
dike Scale ('10), the Ayres Scale ('12), the Starch Scale ('19), and
others. The Thorndike Scale consists of a series of 18 steps or
qualities of handwriting, each step consisting of one or more spec-
imens of writing of the appropriate quality. Step zero represents
an attempt at writing but as such is entirely illegible and devoid
of beauty. Step iS is a perfect copper plate specimen. The steps
from o to iS represent equal units of increase in quality. The
Ayres Scale consists of 8 steps designated as 20, 30, up to 90.
Each step contains three specimens of equal quality, a vertical,
a medium, and a slant sample. The recent revision of the Ayres
Scale, the Gettysburg edition, contains only a medium slant
specimen for each step. The successive steps represent uniform
increments of legibility in writing. The Starch Scale is composed
of a series of 20 steps arranged in the order of merit or excellence
of writing as judged by 400 persons. (See the original monographs
for detailed description of the preparation of these scales.)

A sample of handwriting is measured by any one of the scales by
putting it alongside the scale and determining which step it is most
like in general quality. Speed and quaUty should ordinarily be



measured simultaneously in the same sample because these two
aspects of writing have a functional relationship. A test of writing
should, therefore, be made by having the pupils write a short, simple
sentence repeatedly as many times as they can in, say, two minutes,
doing it as well as they can. Speed is then measured by the number
of letters written per minute and quality is rated by one of the scales.
Freeman ('14) has prepared a set of five analytical scales for the
purpose of rating handwriting from the standpoint of uniformity of
slant, uniformity of alinement, quality of line, letter formation, and
spacing. Each scale contains samples of three successive degrees






3 69



S 62






S 48


<u 41



ra 34

/ Z^J*"*^


y ^, - -


/ /-

/ /


/ /


_j 1 1


12 3 4 5


Fig. 63. — Se.x difference in speed of writing. After Starch ('13).

of merit. These methods of rating ought to be useful in calling
attention to defects in particular features of writing.

The A. N. Palmer Company has published a series of five or six
samples of successive degrees of value for each grade. A percent-
age value for posture, movement, speed, and formation is given for
each sample. The value of the sample as a whole is expressed by the
average of these -four estimates.

The Zaner and Bloscr Company has also issued a set of specimens
for evaluating handwriting consisting of a series for grades one and
two, another series for grades three and four, and a third series for
grammar and high school classes. Each series has a number of
samples whose rating is expressed in terms of percentage values.
Comments concerning the defects or excellencies are appended to
the various specimens.



The manner in which the vahies of the samples were determined
is not indicated for either the Palmer or the Zaner scales. This im-
pairs their scientific value. From the practical standpoint they are
commendable in that they suggest specific attention to, and evalua-
tion of, important elements in handwriting.

(c) Results and uses of measurements. In general the results and
tises of measurements in handwriting are the same as those pointed
out for reading. It is possible by means of these measurements
to determine more precisely the actual writing abihty of a pupil,
class, or school and to compare it with standard averages for cor-

FiG. 64. — Sex differences in quality of writing. After Starch ('13, p. 461).

responding grades in schools generally. These standards of at-
tainment for the ends of the respective school years are as follows:

TABLE 96. After Starch ('16, p. 83)
Standards of attainment in writing

Speed (letters per minute) .
Quality (Thorndike scale) .

Quality (Ayres scale)

Quality (Starch scale)



3 4






38 47






8.2 8.7






33 37






9-7 IO-3




12. 5

By reference to these standards of attainment it is possible to
define quite accurately the speed and quality of writing of a pupil
or class by saying, for example, that a given pupil in the fifth grade
is able to write 65 letters per minute at quality 9, Thorndike scale.
The value of exact measurements of handwriting, as of any educa-
tional products, consists in the diagnosis of ability as it actually
exists in different pupils and schools, in the measurement of the in-


fluence of different factors and conditions upon learning to write,
and in the determination of the mutual relationships of various
aspects of writing. A survey of our present knowledge concerning
these matters will be given in a later section.

Economic Procedure in Learning to Write

What influence do the various factors, conditions, and methods
in learning to write have in promoting or hindering the develop-
ment of skill in handwriting? This question could be answered
finally and fully only by the careful isolation of each factor under
experimental conditions and by determining its effect upon the
progress of learning to write. Substantial beginnings have been
made in the direction of answering some of these questions, but little
or nothing is known about most of them.

(i) Perception of the Forms Written or to Be Written. This
topic requires consideration of two general questions: (a) What
are the most advantageous conditions for the visual perception of
the forms to be written? (b) What sort of form or model should be
presented? The former question may be answered by the observa-
tion of certain obvious rules, namely, that the writing surface should
be placed before the eyes at the proper distance, especially not too
near, so as to avoid eye strain, and in a position directly in front of
the eyes so that the points on the paper to be successively fixated
may be at equal distances from both eyes, thus avoiding unequal
accommodation in the two eyes. The paper should not be glazed,
so that it will not produce a glare, and for young children the sur-
face should be rather rough so that it will easily take pencil marks.

The second question is more complicated. The sort of models
to be presented is obviously highly important since imitation,
both voluntary and involuntary, probably plays a large part
in the acquisition of writing skill. The author ('11) made an
experiment in which he attempted to measure the unconscious
effect of different models of writing upon the normal writing of
adults. Four samples of writing were obtained from each of 106
university students. In order to avoid any suggestion of imita-
tion, written rather than oral directions were given stating that
they were to produce samples of their writing and that they should
proceed at once to write the passages presented without further
thought or questions. The four passages put before each person
consisted of (i) a typewritten selection, (2) an extreme vertical


model, (3) an extreme slanting model, and (4) a large model with
many flourishes. The purpose of the typewritten passage was to
obtain at the outset a sample of the normal writing of each person.
The other three models were taken from school copy books.

After the experiment was finished, each person was asked whether
he had tried purposely to imitate the various models. Three
persons stated that they had intentionally modified their styles
of writing. Their records were thrown out. The samples pro-
duced by the remaining 103 persons were carefully measured to
ascertain their slant and size. Slant was measured by means of
a specially prepared, transparent device with ruled lines for de-
termiring the angle of inclination of certain tall letters, such as 1,
f, and p, with the base line on which the words were written. Size
was measured by determining the horizontal width of letters by
measuring the length of words and dividing by the number of
letters in the word.

These measurements showed that the average tendency for this
group of persons was to make the letters distinctly more vertical
when the vertical model was before them and more slanting when
the slanting model was before them as compared with their nor-
mal styles of writing. They also tended to write slightly larger
when the large model was before them. The amounts of these
changes were as follows:


Average inclination of 1 in the normal writing 65 . i degrees

Average inclination of 1 written from vertical copy 68.

Average inclination of 1 written from slanting copy 61 .

Change from normal to vertical 3 .

Change from normal to slant 3 .

Total range of change 7 •

Average width of letters in normal writing 4.

Average width of letters written from large model 4.

When we realize that the handwriting of adults is a pretty
firmly fixed habit, the amount of unconscious imitation is consid-
erable, being a total of 7.3 degrees in slant and of .52 millemeters
in width. We may infer that with children whose writing habit
is in process of formation, the element of unconscious imitation
plays a much larger part. Furthermore, it seems quite probable,
although no experimental proof is at hand, that the style and
quality of writing of the teacher distinctly influences the writing
of the pupils, especially so because the writing done by the teacher












in the presence of the pupils for the purpose of showing them how
to write, is hkely to be more efficacious in securing imitation than
a static model in a copy-book would be. It would seem, therefore,
highly imperative that every elementary school teacher should be
a reasonably good writer.

In connection with the survey of penmanship in the Grand
Rapids, Michigan, schools, Freeman (Judd, 'i6) reports that

"Grand Rapids adopted about five years ago a new system of pen-
manship. Up to that time the writing was not regarded as satisfactory.
A part of the difticulty was thought to be due to the inability of the
teachers themselves to write well enough to furnish a good example to
the pupils. Accordingly, by action of the Board of Education, all teachers
in the elementary schools were required, as a condition of promotion, to se-
cure a Palmer certificate. This rule has been recently enforced with strict-
ness and the writing in the schools is reported to be greatly improved."

Should the model presented to the pupils be vertical or slanting?
Should it be plain or contain flourishes, decorative curves and
shading? Should it be angular or rounding? The answers to
these questions are at present largely matters of opinion and con-
venience rather than matters of scientific determination. Some
years ago vertical writing came into general use because it was
thought to be more legible and less productive of spinal curvature.
But it has largely disappeared for the obvious reason that almost
everyone naturally falls into the habit of writing a medium slant,
no matter what style of writing was taught to him previously.
The average slant for adults, as shown in Table 97 is about 65
degrees with the base line or about 25 degrees with the vertical
line. Whether slant writing actually tends somewhat more to
produce spinal curvature is doubtful. The difference in legibility
between vertical writing and a medium slant writing is also prob-
ably very small. The letters should probably be of a medium
slant and should be relatively plain and free from flourishes since
these take time and add nothing to the general value of the writ-
ing, and finally, they should probably be moderately rounding
because extreme roundedness is likely to reduce speed and ex-
treme angularity is likely to reduce legibility.

Graves ('17) classified 604 samples of handwriting according to
slant and then studied the speed and quality tendencies of the
vertical, medium slant, and extreme slant group. The final aver-
ages of speed and quality are shown in the following table:



TABLE 98. After Graves

Words Written
IN 5 Minutes

Vertical gi.6

Median slant 96 . i

Extreme slant loi . 7

(Ayres scale)


There is revealed clearly a positive connection between slant and
speed on the one hand and poor quality on the other. That is the
"extreme slant" writers write more rapidly and more poorly than
the "vertical" writers.

(2) Length of Period of Practice. What is the most productive
practice period in learning to write? Even such a question as
this, which is capable of definite experimental solution, has been
answered only in part. The answer given by school programs in
the time allotted therein for writing, is based largely on opinions
instead of facts.

Quality of handwriting at roughly the same rate in seven school systems.

Thorndike ('10)
Median results for eighth-grade pupils

At 20-29 words in
At 30-39 words. .
At 40-49 words . .
At 50-59 words. .
At 60-69 words. .
At 70-79 words . .






4 min.

- -145-





- -12.3-



. .12.0

. .12.0.





. .11.1.

. .11 . 1 .




..II. 5.




. .10.8













Median results, for seventh-grade pupils

At 10-19 words. .
At 20-29 words. .
At 30-39 words. .
At 40-49 words . .
At 50-59 words. .
At 60-69 words . .
At 70-79 words . .

. 9-8

- -13-3-
. .11.4.
.. 9.8.


-10. s
■ 9-9

II. 7



II .0
II. 8
II. 4







13 5
II 8

Systems A and B devote no time to writing as such in grades 7 and 8.
System C devotes 50-60 minutes weekly

" D " 73-100 " "

Systems E and G devote 60-90 "
System F devotes 75 " " " grade

F " 30


Thorndike ('lo) compared the writing in seven school systems
as given in Table 99, and concluded that time was practically
negligible. He says:

"What these facts do prove is: First, that at least three systems (C,
D, and E) get little or no bolter results at a time cost of about 75 minutes
a week than two systems (A and B) do at zero time-cost; second, that
one system (F) at no greater time-cost than C, D, and E gets results
about 25% better than they do; and third, that practice for quality may
secure it only at the cost of speed. The teachers in A and B are better
paid than those in the other cities, so that the success of these schools at
no time-cost might not be generally attainable.

"Leaving F out of account, the differences of these school systems in
the method of teaching handwriting, in the time devoted to it, and in
the ideals of the system in respect to it are of inconsiderable influence
upon efhciency. One makes its pupils write very well at very slow rates,
the others vary a little in quality with small inverse variations in speed.
On the whole, in spite of the achievement of system F, efficiency in hand-
writing seems, like spelling, and unlike arithmetic to be under present
conditions not very much influenced by the management of the schools."
(Thorndike, "Handwriting," p. S3-)

Freeman ('15) had writing tests made in 47 cities and then
compared the attainment in these schools with the amount of time
devoted to the writing-period in each school. His results are set
forth in Figure 65. Each school is represented in the chart by a
short vertical line. This line is placed at a position above the base
line so that it represents the relative rank of that school in attain-
ment in penmanship among the 47 schools.

These results are interesting and valuable, but it is questionable
whether they prove that time makes no difference. The difificulty
with a wholesale set of figures such as these is the impossibility
of separating the various elements and determining their ef-
fects individually upon the ultimate attainment in writing. The
schools which devoted 90 to 100 minutes per week to writing and
obtained no better results than the schools which devoted 40 to 50
minutes per week, may contain other factors which kept their
proficiency down, such as, poorer teaching, different classes of
pupils, the quality of writing done in other subjects outside of the
writing period, which probably has as much if not more influence
upon proficiency in writing than the writing period itself, and so
on. In fact, we might even imagine that if these same schools had
devoted only 40 to 50 minutes per week, they might have been



much worse in writing than they actually were. The real expla-
nation may perhaps lie in the possibility that the schools having
longer writing periods may not use the time to as good advantage
as those having shorter periods. The latter, by virtue of having
only a short time to devote to the subject, may work more in-
tensely and profitably.

Online LibraryDaniel StarchEducational psychology → online text (page 26 of 41)