bib^c 



BULLETIN 

OF THE 

UNIVERSITY OF TEXAS 

1915: No. 37 



JULY 1 



1915 



A SCORE CARD FOR THE MEASUREMENT OF HANDWRITING 

BY 

C. TRUMAN GRAY 

Instructor Department Education 




Published by the University six times a month and entered as second 
class matter at the postoffice at Austin, Texas 



ifl allograph 



Publications of the Uni v^ersity of Texas 

Pub lictioDs Committee: 

W. J. Battle B. C. Bapiker 

J, C. TowNES A. Caswell Ellis 

W. S. Carter R. A. Law 

KiLLis Campbell J. A. Lomax 

F. W. SiMONDS A. C. JUDSON 



The University publishes bulletins six times a month. These 
comprise the official publications of the University publica- 
tions on humanistic and scientific subjects, bulletins prepared 
by the Department of Extension and by the Bureau of Munci- 
pal Research, and other bulletins of general educational in- 
terest. With the exception of special numbers, any bulletin will 
be sent to a citizen of Texas free on request. All communica- 
tions about University publications should be addressed to the 
Editor of University Publications, University of Texas, Austin. 



A. C Baldwin & Sons, Austin, Texas 



B108-715-lm-8402 



BULLETIN 

OF THE 

UNIVERSITY OF TEXAS 

1915: No. 37 



JULY 1 



1915 



A SCORE CARD FOR THE MEASUREMENT OF HANDWRITING 



BY 



C. TRUMAN GRAY 

Instructor Department Education 




Published by the University six times a month and entered as second' 
class matter at the postoffice at Austin, Texas 



4A° 



V% 



The benefits of education and of 
useful knowledge, generally diffused 
through a community, are essential 
to the preservation of a free gor- 
emment. 

Sam Houston. 



Cultivated mtnd is the guardian 
genius of democracy. ... It is the 
only dictator that freemen acknowl- 
edge and the only security that free- 
men desire. 

Mirabean B. Lamar. 



Do of Do 
DEC : " 1915 



AUTHOR'S PREFACE 

Ab investigation such as is set forth in the following pages 
can result only from the cooperation of a considerable number 
of persons. My appreciation of the work of all those who have 
helped is keenly felt. The greatest assistance has been received 
from Dr. J. Carleton Bell, Dr. Truman L. Kelly, and Dr. L. W. 
Sackett. The criticisms and suggestions of these men have 
been given freely at all times. Special mention should also be 
made of the work of my wife, Bessie Stretcher Gray, who has 
so ably assisted in the many statistical calculations. 

C. T. G. 

Austin, June, 1915. 



TABLE OF CONTENTS. 

Introduction 7 

Educational Measurements 7 

Teachers' Grades 7 

Courtis Tests 7 

Scales 7 

Minimum Essentials 7 

Score Card Method 7 

Score Card in Agriculture 8 

Score Card in Education 8 

Indiana Score Card ' 8 

Elliott's Score Card 9 

Boyce 's Score Card 11 

The Problem I'i, 

The Determination of the Points which Enter into the Score 

Card 13 

Lists from Which Selection is Made 1-4 

Point of View from AVhieh the Score Card is Derived 15 
Other Principles to Be Followed in the Selection of 

Points , 16 

Suggestive Methods for Determining the Points 17 

List of Points 18 

The Evolution of the Points Which Enter into the Score 

Card If) 

The Thorndike Method 19 

Results from Teachers and Supervisors of Writing. 20 

Results from Elementary-School Teachers 27 

Results from Teachers and Students of Education. 31 

Results from all Judges 34 

The Correlation Method 35 

Uses and Forms of the Score Card 44 

Objections to the Score Card 46 

Experimental Work with the Score Card 47 

Bibliography 49 



INTRODUCTION 

One of the most interesting and important movements in 
modern education is that which has for its purpose the meas- 
urement of the immediate results attained by the teaching 
process. Less than a decade ago, it was assumed that the grad- 
ing systems used by the present day schools gave such measure- 
ments, but the work of Starch and Elliott (12), Kelly (10), 
Gray (6), and others shows clearly enough that this method is 
unsatisfactory. 

At present, other methods for securing such measurements 
are being proposed. An extended review of these various 
studies would be in place here, but space and time permits only 
the mentioning of the various plans. 

First: Courtis (5) has used quite extensively a set of tests 
for measuring ability in arithmetic. 

A second method is to be found in a series of scales intro- 
duced by Thorndike (16), and by Ayres (1) for the measure- 
ment of handwriting; by Thorndike (17) for drawing; by 
Buckingham (4) for spelling; by Hillegas (9) and by Ballon 
(2) for composition. In addition to this, there has been a 
tentative scale developed by Gray (8) for the measurement of 
reading. 

The third type of work to be mentioned in this connection is 
that known as Minimum Essentials. The last yearbook of the 
National Society for the Study of Education is devoted entirely 
to this topic. In addition to this volume there are pamphlets 
published by Thompson (15) which contain the essentials for 
memory work in arithmetic and geography. 

Attention may now be called to another method of measure- 
ment which seems to lend itself to the problems mentioned 
above , but which has received very little attention from edu- 
<iators. The method referred to is measurement by means of 
the score card. This plan for measuring has been developed 
and brought to a high degree of perfection by agriculturists, 
and the results obtained by them are recognized as accurate 
and scientific. 



8 



Bulletin of the TJniversity of Texas 



Such score cards consist of two essential parts : first, a list 
of the elements of the thing to be measured; and second, an 
evaluation of these points in such a manner that the sum of the 
values equals one hundred. A typical score card, that for 
wheat, is shown be'low. It will be noticed that it contains five 
main points with three minor points, and an evaluation of each 
of the points. 

SCORE CARD FOR WHEAT 

Texas A. and M. College. 



SCALi: OF POINTS 



o 



1. Weight per bushel 

2. Soundness 

3. Purity 

4. Size and plumpness of kernels 
6. Uniformity in — 

(a) Hardness 

(b) Color 

(c) Size of kernels 

Total 



lOO 



Many different phases of educational activities, such as 
drawing, compositions, writing, school buildings, etc., seem to 
lend themselves to just such measurement as is provided in 
the score card ; and yet as far as the author is aware, no such 
attempts have been made. It is true, howevc'r, that attempts 
have been made to use this method in the measurement of the 
efficiency of teachers. In the year 1908 the State Department 
of Education in Indiana published a score card for determining 
the "success grade" of any teacher who had had experience in 
teaching. The different points seemed to have been selected 
in an entirely arbitrary manner, and the values used were 
doubtless based upon one man's judgment as to their relative 
importance. The plan was considered too complicated by su- 
perintendents, and as a result, it was never put into general 
use. 

In the year 1910 Professor E. C. Elliot presented to the 
state convention of city superintendents of Wisconsin "A 



Score Card for Measurement of Handwriting 9 

Tentative Scheme for the Measui'ement of Teaching Efficiency.". 
Again it can only be inferred that the points entering into the 
scheme and their values have both been determined in an 
arbitrary way. The following is the schedule of points and 
their relative weights : 

I. Physical Efficiency — 12 points (12)" 

1. Impression — general 2 

2. Health — general 2 

3. Voice 2 

4. Habits — personal 2 

5. Energy 2 

6. Endurance 2 

II. Moral — Native Effijciency — 14 points (14)" 

1. Self-control 2 

2. Optimism— enthusiasm 2 

3. Sympathy — tact 2 

4. Industry — earnestness 2 

5. Adaptability 2 

6. Sense of humor 2 

7. Judicial mindedness 2 

III. Administrative Efficiency — 10 points (10)" 

1. Initiative 2 

2. Promptness and accuracy 2 

3. Executive capacity '. . . . 2 

4. Economy (time, property)- 2 

5. Cooperation (associates and supervisors) 2 

IV. Dynamic Efficiency — 24 points (24)" 

1. Preparation 4 

Including: 

(a) Intellectual capacity 

(b) Academic education 

(c) Professional training 

2. Professional attitudes and interest 2 

3. Human nature, attitudes, and interest 

(Appreciation of values — intellectual, so- 
cial, and moral in child life) 2 

4. Instructional skill 12 

Including: 

(a) Attention and interest of pupils 

(b) Formality V. vitality of instruction 

(c) Motor V. verbal methods 

(d) Application of the technique of teach- 

ing; organization and presentation 
of subject matter; the recitation as 
artistic product. 

(e) Application of the technique of living; 

participation and contribution of 
pupils; the recitation as a demo- 
cratic activity. 

(f) The tools and machinery of instruc- 

tion; effective adaptation. 

(g) Assignment of work 

5. Government and directive skill (discipline) ... 4 



10 Bulletin of the University of Texas 

V. Projected Efficiency — 6 points ( 6 ) " 

1. Continuing preparation 2 

(a) daily 

(b) weelily 

(c) annual 

2. Tlie school program 2 

3. Increase of professional equipment 2 

(professional reading and study; travel) 

VI. Achieved Efficiency — 2 4 points (24)" 

1. Achievement 

(a) Illustrative results 8 

(b) Examinations; success and attainments 

of pupils 12 

2. Stimulation of individuals and community. ... 4 

VII. Social Efficiency — ^10 points (10)'' 

1. Intramural interests 2 

2. Extramural interests 

(a) Cultural and ethical 2 

(b) Civics 2 

(c) School patrons 4 

Another plan which has been worked out with much care 
has been published by Boyce (3). The points which enter into 
the record are chosen arbitrarily but with much care and 
thought, and no definite values are assigned. The author seems 
to think it better to use such terms as "Very Poor," "Poor," 
-etc. Some work on the relative importance of the various 
points has been done by using correlation coefficients. A copy 
'Of tlie record with the estimates of three different judges upon 
Ihe same tea(3hei' follows : 



Score Card for Measurement of Ilandivriting 



11 





Efficiency 


Record 
























GENERAL RATING 


1 


1 1 1 


1 1 1 1 1 M 


_JD1X|0 


1 


Oi'ALiTiES or Merit 


Vt.v 
Pooa 


Poor 


Medium 


Goop 


Ex. 


I. Persona] Equipment— 






























X 


n 




|o 




— 


2. Health 






























X 




□ 


k 


3. Voice 








<x 




















n 















4. Intellectual capacity 


































1 


m 1 




























X 




nic 


^ 






6. Adaptability and resourcefulness _ 






























V' 







r 






7. Accuraty 


































nixio 




8. Industry 






















, 
















n 





9. Enthusiasm and optimism 
























X 








n 












10. Integrilv and sincerity , 






































p^ 


1 1 . Self-control 






























X 






p\ 






12. Promptness. 




















i 














» 






13. Tact 






























(^ 













14. Sense of justice 






























X 





a 








II, Social and Professional Equipment— 






































m 




2. Prnfcssional preparation 






























ra 





D 


1 


3. (Irasp of subject-matter 


































X 


ol 


























□ 







X 












S- School and commimitv Interept 
































IS 











6. Ability to meet and interest parents. . 






















X 




D 

















7. Interest in lives of pupils 






























X 




a 









8. Co-operation and lovaity 











































Q. Professional interest and growth. .... 




































a 










































n 




L_ 




































X 






III. School Management— - 




























a 


X 





K 





- 






























































X 
































^ 






X 















IV. Technique oj Teaching — 






























1 


tsi 





































□ 


X 






































XiD 










































































- 


Ql 


□ 4? 


— 


— 




























































□1 






& 
































Di 






is 


































X 


ol 







































X 




Oj 













V. Resulls— 




























□ 






a 










































n 







— 




























X 




n 
































S] 



















































X 




D 













1 






__ 




























1 



Graph II.— Efficiency of a teacher recorded by three different judges 
Superintendent: □ ; Principal: X ; Supervisor: Q 



12 Bulletin of the University of Texas 

Other plans of the same nature have been published by 
Witham (18), Stowe (14), Ruediger and Strayer (12). 

This brief survey of the work done in this field seems to 
indicate that while the general plan of measuring the efficiency 
of a teacher by the score card is considered good, still the per- 
fecting of such a plan will require much patience and effort. 
If these different score cards for grading the efficiency of teach- 
ers be examined, very few if any suggestions upon method will 
be found. In other words, no plan other than that of arbitrary 
selection is given for the determination of the points which 
should enter into the score card. Moreover, there is no pro- 
vision made for the evaluation of the diff'ereut rubrics except 
b}^ means of the correlation coefficient ; and there is no plan 
suggested for transferring these coefficients into such values 
as are used in the Indiana plan. 

However, the fact that the score card is being brought to 
bear upon one of the most important problems in education, — 
the measurement of the efficiency of teachers, and the fact that 
it has been used so successfully in other fields, lead the author 
to believe that this method has a much larger place in education 
than has been accorded it. The purpose of the following pages 
is to present a score card for measuring one product of the 
teaching process juid to pres; nt the methods by which the score 
card has been derived. 

THE PROBLEM 

More specifically our problem is the formation of a score 
card for the measurement of handwriting. As has been sug- 
gested this requires that the essential elements of writing be 
selected and that these various elements be weighted. 

One of the chief advantages of this method of measurement 
when applied to writing lies in the fact that it requires an 
analysis of writing. To say to a child that his writing is very 
poor and deserves only fifty out of a possible one hundred 
points means little to either the child or the teacher and cer- 
tainly it gives the child little if any basis upon which his efforts 
for improvement may be founded. The fact is that usually 
writing is not entirely bad, and if the teacher is competent 
she can analyze it so as to point out the good and the poor 



Score Card for Measurement of Handwriting 13 

points. If her analysis reveals that slant, spacing or size are 
to be criticized she at once gives both herself and the pupil a 
basis for future work. 

The strongest argument which has been urged against both 
of the present writing scales is that they are each made upon 
a basis which in no sense attempts any analysis of writing. In 
one case the basis is general merit and in the other the basis 
of legibility is used. Grades in terms of such points may have 
many uses, but certainly from the standpoint of the teacher, 
of the pupil, and of the superintendent it would be valuable to 
have at least a part of the grades given in terms of the elements 
of writing. 

If two systems of schools, or two systems of teaching writ- 
ing, or two teachers of writing are to be compared as to the 
results shown in writing, the results would mean much more 
if certain elements of writing were agreed upon and the com- 
parison made in terms of these. 

The weighting of the dilferent elements of writing is also an 
important problem. There are many teachers who emphasize 
movement above everything else, others emphasize slant, while 
not infrequently there are those who think that neatness is 
the prime thing in all writing teaching. 

In view of such divergence of opinion it is certainly of 
value to know how a representative body of writing teachers 
or elementary school teachers rank and evaluate these various 
elements. Such knowledge could be used by teachers, super- 
visors, and superintendents in dealing with the problems of 
writing. 

To sur/imarize : The problem is a two-fold one, involving the 
determination of the elements of writing and the relative values 
of the elements chosen. 

THE DETERMINATION OF THE POINTS WHICH ENTER 
INTO THE SCORE CARD 

A study of the different score cards in agriculture reveals 
no definite plan of procedure for this phase of the problem at 
hand. Sometimes such cards are made entirely by a depart- 
ment in an agricultural college, or they may be authorized by 



14 



Bulletin of the University of Texas 



some association, or in still other instances they have been pro- 
duced by a single individual. As has already been shown, the 
score cards which have been used in the measurement of the 
efficiency of teachers give little if any suggestion upon the 
determination of the points which enter into them. 

The plan followed in the present investigation was first to 
get as large a list as possible of the elements of writing and 
then to select a limited number from this list. After considera- 
ble time had been spent in going through books on writing and 
corresponding with teachers and supervisors of writing and 
others who have much to do with writing either directly or 
indirectly the following list of the elements of writing was 
compiled : 



Beauty 

Shading 

Legibility 

Speed 

Formation of letters 

Execution 

Position 

Slant 

Neatness 

Endurance 

Uniformity in turns 

Uniformity in retraces 

Uniformity in slant 

Uniformity in spacing 

Uniformity in endings 

Ease 

Individuality 

Shape of letters 

Lightness 

Parts omitted 

Conformity to an ideal 



Spacing of letters 

Spacing of words 

Spacing of lines 

Alignment 

Movement 

Form 

Size 

Uniformity 

Accuracy 

Smoothness 

Uniformity in angles 

Uniformity in loops 

Uniformity in size 

Uniformity in beginnings 

Uniformity in height 

Touch 

Effort 

Proportion 

Strength of line 

Parts added 



An examination of this list reveals the many diverse ideas 
and opinions whicli are held in regard to writing. Evidently 
all of these points can not be used and our problem becomes 
one of selection. A closer examination of the list seems to 
show two rather distinct points of view. There are certain 
rubrics which are evidently intended to emphasize the manner 
in which the writing has been produced, while others are in- 
tended to emphasize writing as a product. To illustrate, the 
teacher who says that position and movement are the most im- 
portant things in writing is thinking of writing as a process, 
while the teacher who says that "formation of letters" or 



Score Card for Measurement of Handwriting 15 

spacing is the most important is evidently thinlving in terms of 
results. 

To get a basis for a selection of the elements of writing it 
seems necessary to make a choice between these two points of 
view. Such a choice was determined by a consideration of 
certain problems which may be attacked by means of the score 
card in the future. 

In other words there are certain important investigations in 
writing which are of such a nature that the amount of work 
required to get results by means of a score card constructed 
upon the basis that writing is a process would be almost pro- 
liibitive, while such work would proceed much more rapidly 
if the score card is constructed upon the basis that writing is 
a product. Many illustrations for this point might be given. 

First, it seems that the teaching of writing is suffering from 
an over abundance of methods and systems. The elimination 
of poor methods can only be brought about by scientifie and 
accurate investigations which compare these different methods. 
There is a basis for such a comparison if each system uses its 
best technic[ue and then has its results compared with the 
results of other systems. Such comparisons are made many 
times in terms of technique and so do not give satisfactory 
results because the principles underlying the teaching of writ- 
ing are not well-defined. 

Second, the writing of a large number of children should be 
scored so that norms can be established and a basis given for 
the comparison of different school systems and of the different 
elementary and high school grades as to their writing. If such 
items as movement are to be included in the scoring, the work 
becomes almost impossible, because it means that each pupil 
must be scored separately while he is writing and to make this 
scoring valid it should be done by one person. The time and 
labor which this process would require make it almost im- 
possible to score any considerable number of records, but under 
the conditions here proposed a superintendent could have each 
child in his school system write a sample with the technique 
used in that system, and the samples could be graded later by 
experts. 

In addition to these two reasons for considering writing as 



16 Bulletin of the University of Texas 

a product a third may be added. If the teacher of writing 
grades her pupils upon the basis that writing is a product, she 
establishes a relation between the process and the result. One 
reason that the child does not make progress in the matter of 
position is that no relation between position and results is 
seen. If the child were graded low on slant, say, and he were 
told that better slant could be had with proper position, cer- 
tainly his interest in position would increase. 

It is not the intention of the author to minimize the value of 
writing as a process nor does he wish to give the impression 
that such should not be kept in mind by the teacher, but there 
are many good reasons both from the standpoint of scientific 
investigations and of good teaching why writing may be con- 
sidered as a result. 

Certain principles which should be followed in the selection 
of the rubrics may now be discussed. The most obvious of 
these is that if the card is to be usable, the number of elements 
selected must be small. A serious objection to the score cards 
for the grading of teachers is that they contain such a large 
number of rubrics that the time required almost prohibits the 
general use of such a plan of grading. However, the number 
must be sufficiently large to cover the most essential if not all 
the different phases of writing. Later discussion will set forth 
the nine main points with five points subsidiary to one of the 
nine which number seems to meet both of the foregoing require- 
ments. 

Another principle which must be considered requires that the 
terms be mutually exclusive or as nearly so as possible. If we 
consider "beauty" for a moment it can be seen that it is a 
function of many other elements, and would not be selected 
because it includes too much. Other elements of which the 
same thing is true are legibility and neatness. On tlie other 
hand if such a thing as "spacing of letters" is considered, it is 
seen immediately that this refers to just one thing and all 
others are excluded. 

In addition to these principles certain methods for selecting 
the elements suggest themselves. It would be possible to 
determine the correlation between general merit in writing and 
each of the points in the list given above, and also between 



Score Card for Measurement of IhnidwrHing 17 

each two of the points in the list. Two objections may be urged 
against this method. First, if such results were obtained it 
would require a choice among those points for which the cor- 
relation is high, and then among those which did not correlate 
so highly, etc. After all, this resolves itself into an arbitrary 
choice which is just the thing to be avoided. The second ob- 
jection would be the very large amount of work which would 
be required in collecting and compiling sufficient data for the 
solution of such a problem. 

Another general plan which might be used would be to sub- 
mit such a list to a representative body of teachers and have 
them mark what they considered to be the ten or twelve which 
are of most importance. The objection to be raised here is 
that it is not known in advance how many elements must be 
chosen and it is very doubtful if mutually exclusive points 
could be procured in this way. 

A third plan which suggests itself is to have the points agreed 
upon by some one or two representative bodies of teachers. 
This method is approved in certain phases of agricultural work. 
However, until the score card establishes itself in education as 
an efficient method of measurement, very little can be expected 
from this source. 

The last plan to be suggested is a selection which grows out 
of experience with the card itself. This plan coupled with 
arbitrary choice is the one used by the author. A number of 
students began the use of the card when it was in a very crude 
state with only tentative values, with the idea of determining 
what points should be used in order to give a complete account 
of writing. The general plan followed was to have these 
students grade separately a number of samples of writing each 
week- and to follow the grading by a conference with the 
author. They were cautioned to watch for points in tlie writ- 
ing which were not covered by the card that they were using. 
In light of the experience gained in the above experiment, and 
by keeping in mind the foregoing principles the following list, 
of points was selected: 

1. Spacing of letters. 

This includes the uniformity and the length of the space. 



18 Bulletin of the University of Texas 

i. e., the space may be too short, which leads to crowding, or 
it may be too long, or it may lack uniformity. 

2. Spacing of lines. 

The standard here is that the lines should be a uniform 
di!::tance apart and that this distance should be neithei- too 
great or too small. If the paper is ruled this is made much 
easier for the writer, but even then, many persons do not suc- 
ceed in properly spacing their lines. 

3. Spacing of words. 

This includes points similar to those under "1" above. 

4. Slant. 

Here is included such points as uniformity and degree of 
slant, i. e., the writing should not be extreme in either direc- 
tion. This element of writing is closely related to position. 

5. Size. 

Writing should be uniform in size and neither too large 
nor too small. This is well expressed by the term proportion. 

6. Alignment. 

This suggests that the writing should follow a line and 
that this line should be perpendicular to the edge of the paper. 
If the paper is ruled this is made much easier for the writer, 
but even then many children have difficulty in meeting this 
requirement. 

7. Neatness. 

This was one of the last points to be added because it was 
iirst thought that it was a function of many other elements, 
and so need not be included. It was finally included so as to 
take account of such points as blotches, carelessness, and re- 
tracing. In this sense it is closely related to effort. 



8. Heaviness. 



(ka) 



As the illustration indicates, this term is intended to in- 
clude its opposite lightness. It suggests the width or quality 
of the line. It depends largely upon the movement used and 
in a less degree upon speed. 
9. Formation of letters. 

This is intended to emphasize the degree in which the 
letters conform to the correct form. It is rather a comprehen- 
sive term and iive subsidiary points are included under it. 



Score Cord for Measurement of Handwriting 19 

(a) Parts omitted. 

Often seen in the letters a, g, and d as ^^ '^ ^^ 

(b) Letters not closed. ' 
Often seen in the letters mentioned above as 



cc 



c^czl 



(e) Parts added. 

Flourshes and dashes. 

(d) Smoothness. 

Indicates the execution. The letters should 
neither be cramped nor jerky. 

(e) General form. 

The following illustration would not be good form 



tC,/^ 



THE EVALUATION OF THE POINTS WHICH ENTER INTO 
THE SCORE CARD 

Two methods for obtaining such values suggested themselves, 
and in this initial attempt at such work, it was thought best to 
determine such values by each of the methods. 

The first inethod to be mentioned is a modification of the 
Thorndike method used in deriving scales for drawing and 
writing. It was used also by Hillegas in his scale for composi- 
tion and a modification of the same method was used by Buck- 
ingham in his scale for spelling. 

The second method is known as the regression equation. 
The theory of this method is discussed in detail by Yule (19) 
and is used extensively in connection with an educational 
problem by Kelley (11). 

In order to get data for the Thorndike method it was deter- 
mined to get the judgments of a considerable number of per- 
sons upon the relative importance of the points listed on page 
18. These judgments were rendered by three distinct groups 
of judges: first, teachers and supervisors of writing; second, 
elementary school teachers; and third, teachers and students 
of education. 



20 Bulletin of the University of Texas 

RESULTS FROM TEACHERS AND SUPERVISORS OV 

WRITING 

A list^ of the members of The National Association of Pen- 
manship Supervisors was obtained and to one hundred twenty- 
five of its members the following letter was sent: 
My dear Sir or Madam : 

Some of the teachers of writing in Texas have asked 
me to cooperate with tiiem in arranging some plan to 
enable them and those under their supervision to 
grade writing in a more exact manner. Will you not 
cooperate in this matter by giving us the benefit of 
your expert opinion upon the points mentioned in the 
inclosed sheet. Any criticisms or suggestions will be 
appreciated. 
Accompanying each letter was a sheet like the following, 
in which it will be noted that the points upon which the judg- 
ments are to be rendered are listed: 

If a sample of handwriting is considered as a product or a 
result, the points listed below may be taken into consideration 
when such a sample is graded. Will you please rank the points 
in the list as to their importance in grading by placing the 
digit "1" before the one which you consider the most im- 
portant of the nine points, the digit "2" before the one which 
you consider to be second in importance, and so on until you 
have ranked the entire list. If you think two or more should 
have the same rank, put the same digit before each. 
Spacing of letters. 
Spacing of words. 
Spacing of lines. 
Slant. 
Size. 

Alignment. 

Neatness (blotches or carelessness). 
Formation of letters. 



Heaviness 1 0O (\J ) 



'This list was very kindly furnished to me by Mr. G. G. Gud- 
mundson of Bonne, Iowa, who is secretary of the above-mentioned 
association. 



Score Card for Measurement of Handwriting 21 

Under formation of letters the five points listed below may 
be considered as subsidiary points. Will you please rank these 
five points as you did the one above by placing the digit "1" 
before the one which you consider the most important, etc. 
Parts omitted. 
Letters not closed. 
General formi 
Parts added. 
Smoothness. 

From these letters seventy-five replies were received. The 
letters which accompanied many of the return sheets were very 
interesting and contained many valuable suggestions. 

One of the most striking things revealed by this part of the' 
investigations is the many diverse opinions held with reference 
to writing. A few expressed ideas very similar to those held 
by the graphologist, while others in direct contrast expressed 
views in hannony with the best psychological theories concern- 
ing the subject. 

Still others expressed the opinion in no uncertain terms that 
the investigation was not scientific and refused to pass the 
judgments called for in the letter. 

Some few expressed the belief that the one thing which is 
needed by writing teachers above everything else is an ade- 
quate system of measurement. They pointed out that this is 
demanded by both superintendents and boards of education, 
and that as yet writing teachers have made no contribution to 
this most important problem. 

On the whole, the interest expressed in not only this line of 
w^ork but any other which has for its purpose the betterment 
of handwriting was wholesome and augurs well for the future 
of writing, 

Table 1 gives the distribution of the judgments upon the im- 
portance of the fourteen points. It will be recalled that the 
directions were such that two or more of the points could be 
given the same rank, and most of the judges saw fit to rank 
some of them in this way. This makes it necessary to desig- 
nate certain ranks by such digits as 1.5, 2.5, etc. This comes 
from the artificial device of statistics which gives each of two 
points having the rank of 2, for example, a rank of 2.5. Tn 



22 



Bulletin of the University of Texas 



this table the different points are listed at the left and the 
digits at the top indicate the different ranks. 

TABLE 1 

Showing the distribution of judgments upon the relative importance of the various 
points listed below. The lesults for "spacing of letters" can be read as follows: No 
judges rank it as 1, three rank it as 1.5, etc. 



Spacing of letters. .. 

Spacing of words 

Spacing of lines 

Slant 

Size 

Alignment 

Neatness 

Formation of letters 
Heaviness 

Parts omitted 

Letters not closed. ._ 

General form 

Parts added 

Smoothness 



1 


1.5 


2 


2.5 


3 


3.5 


4 


4.5 


5 


5.5 


6 


6.5 7 


7.5 


8 


8.5 





.3 


3 


2 


10 


6 


10 


2 


7 


1 


13 


1 


11 


2 


3 








] 





1 


4 


6 


9 


3 


9 


2 


9 


1 


11 


5 


12 


1 








1 





.■5 


4 


4 


2 


6 





4 


3 


11 


3 


15 


3 


9 


1 


8 


.■^ 


12 


8 


9 


2 


7 





4 


4 


2 





4 





1 


3 


9 


2 


10 


6 


11 


1 


7 


1 


7 


3 


4 


2 


4 





1 


1 


1 





4 


1 


4 


3 


3 





8 


3 


11 


2 


7 


6 


8 


n 


10 


3 


6 


4 


1 





7 


1 


4 


3 


3 





10 





?9 


5 


7 


2 


9 


5 





2 


4 





2 





4 





3 


1 


4 


7 


6 


5 


3 


4 


5 


1 


5 


1 


7 


2 


8 


2 


6 


1 


2 


. 


5 


5 


16 


5 


27 


12 


3 
















4 


2 


7 


7 


31 


1 


15 


2 


6 
















ar) 


12 


29 


3 


4 





3 





4 






















2 





5 


4 


13 


11 


40 
















36 


13 


12 


1 


5 





4 


1 


3 

















Some things about this table should be noted. The distri- 
bution of "formation of letters," "slant" and "size" are 
skewed to the left, which indicates that the final ranking of 
these will be high. In contrast with these the curves for 
"alignment" and "spacing of lines" are very much skewed 
to the right, indicating that the final ranking of these points 
will be low. The most symmetrical curves are for "spacing 
of letters" and "spacing of words," while that of "neatness" 
is peculiar in that it has three rather distinct modes, one at 
1.5, another at 5, and a third at 8. Probably the' most inter- 
esting distribution is that of "heaviness." It is almost rec- 
tangular with no central tendency. The close relation of this 
rubric to movement has been pointed out by many teachers and 
it was supposed in the beginning that this would show a very 
decided skew toward the left. 

The next step in the procedure is to determine the sequence 
of the various rubrics, beginning with the one which is most 
important and so on down to the one which is least important. 
Such a sequence can be determined in a tentative way by an 
inspection of the distributions, shown in Table 1. This is to 
be followed by a count of the number of judges who place the 
first point above the second, the second above the third, etc. 
These results are shown in Tables 2 and 3. 



Score Card for Measurement of Handivriting 



23 



TABLE 2 

"E" stands for "equal." "M. I." stands for "more important." The table should 
be read 9 persons judge "formation of letters" equal in importance to "neatness," and 
44 persons judge "formation" of letters more important than "neatness," etc. The 
digit "2" at the bottom of this column means that 2 persons judged "formation ol 
letters" least important. 



Neatness 

Slant 

Size 

Heaviness 

Spacing of letters 
Spacing of words 
Spacing of lines-- 
Alignment 

Least important- 




M.I. 



39 



E. M.I. 



E. M.I. 



E. M.I. 



35 



TABLE 3 

E stands for "equal." M. I. stands for "more important." The table should be 
read: Fifteen persons judge "smoothness" as equal in importance to "general form" 
and thirty-seven persons judge "smoothness" as more important than "general form," 
etc. The digit 3 at the bottom of this column means that three persons judge 
smoothness as least important, etc. 



General form 

Letters not closed 

Parts omitted 

Parts added 

Least important- 



OS 

E.lM.I. 

15 37 
3 6 
2l 9 
O' 6 



M.I. 



M.I. 



19' 49 



E. M.I. 



By taking one-half of the "equal" judgments and addin?' 
them to the "more important" judgments in each case it is 
possible to reduce these results to a percentage basis. By re- 
ferring to Table 2 it is seen that 44 judges rank "formation of 
letters" as more important than "neatness," and 9 rank the same 
points as equal. Adding 4.5 to 44 it is found that 64.6% of 



24 Bulletin of the University of Texas 

the 75 judges have placed formation of letters above neatness 
in importance. In this same manner Tables 4 and 5 are de- 
rived. 

TABLE 4 

Spacing of lines is judged more important tliaa alignment by 54% of the judges 

Spacing of words is judged more important than spacing of lines by--77% of the judge-: 
Spacing of letters is judged more important than spacing of words by 73% of the judges 

Heaviness is judged more important than spacing of letters by 54% of the judges 

Size is judged more important than heaviness by 54% of the judges 

Slant is judged more important than size by 60% of the judges 

Neatness is judged more important than slant by 54% of the judges 

Formation of letters is judged more important than neatness by 65% of the judges 

TABLE 5 

Parts omitted is judged more important than parts added by 77% of the judges 

Letters not closed is judged more important than parts omitted by 64% of the ju<lges 

General form is judged more important than letters not closed by 78% of the judges 

Smoothness is judged more important than general form by 59% of the judges 

Referring now to Thorndike's table- it is possible to get a 
statement of the difference between each pair of items in 
Tables 4 and 5 in terms of the probable error (D) of the dis- 
tribution of the judgments. By the use of this table, Tables 
6 and 7 are derived. 

TABLE 6 

Difference between values for spacing of lines (b) and align- 
ment (a) = .149D. 

Difference between values for spacing of words (c) and 

•spacing of lines (b) =1.094D. 

Difference between values for spacing of letters (d) and 

spacing of words (c) = .909D. 

Difference between values for heaviness (e) and spacing of 

letters (d) = .149D. 

Difference between values for size ff) and heaviness (e)=: .149D. 

Difference between values for slant (g) and size (f) = .376D. 

Difference between values for neatness (h) and slant (g) . .:= .149D. 

Difference of between values for formation of letters (i) and 

neatness (h) = .571D. 

TABLE 7 

Difference between values for parts omitted (y) and parts 

added (z) =1.094D. 

Difference between values for letters not closed (x) and 

parts omitted (y) = .532D. 

Difference between values for general form (w) and letters 

not closed (x) =1.143D. 

Difference between values for smoothness (v) and general 

form (w) = .337D. 

Stating this more simply by using the letters in parenthesis. 
Tables 8 and 9 are derived. 



2. Teachers College Record, Vol. 14, No. 5, page 25. 



Score Card for Measurement of Hatidwriting 25 

TABLE 8 



b- 


■a= .149 D 


c - 


b= 1.094 D 


d- 


•c= .909 D 


e - 


d= .149 D 


f- 


e= .149 D 


g- 


■f= .376D 


h- 


-g= .149 D 


d- 


•h= .576 D 


y- 


- z= 1.094 D 


X - 


- y= .532 D 


w - 


- x=1.143 D 


V - 


-w= .337 D 



TABLE 9 



From Table 8 the following set of equations is derived : 

(1) b-a= .149 D (1) 

(2) c-a=1.243 D (2) 

(3) d-a== 2.152 D (3) 

(4) e-a=2.301D (4) 

(5) f-a = 2.450D (5) 

(6) g-a=2.826 D (6) 

(7) h-a= 2.975 D (7) 

(8) i-a = 3.551 D (8) 

Then, b+c+d+e+f +g+h+i-8a = 17.642 D 

Adding to this a — a=0 

Then a+b+c+d + e+f+g+h+i-9a = 17.642 D 

In tlie earlier part of the discussion is was suggested that 
the sum of the points must equal 100. Substituting this value, 
the equation becomes 

9a = 100- 17.642 D (9) 
Now D is defined as that difference in importance upon which 
75% of the judges agree and may have one of many numerical 
values, so it becomes necessary to determine a definite value 
for it. Since it is defined as a difference it must have a value 
greater than zero, for in judgments as here rendered a negative, 
difference or a zero difference could not be considered, and if 
tlie values of the different elements are to be positive, the 
value of D cannot be greater than approximately 5.6. 

It seemed impossible to determine the value any more ac- 
curately than this unless "a" in equation (9) was known, and 
so a value for "a" was next determined. In order to get data 
for this, the following sheet with the ranks which they had 
given at an earlier ^time were given to a number of persons. 
Returns were received from thirt^^ judges. 



26 Bulletin of the University of Texas 

Dear Sir or Madam : 

The digits placed before each of the points listed below 
indicate the ranking which you gave the same points some 
time ago. AVill you at this time please give the points 
numerical values in accordance with these ranking, so that the 
sum of the nine values will equal 100. 

Spacing of letters 

Spacing of words 

Spacing of lines 

Slant 

Size 

Alignment 

Neatness (blotches or carelessness) 

Formation of letters 

Heaviness 

Will you also distribute the number of points which you 
give to "formation of letters" among the following subsidiary 
points : , 

Parts omitted 

Letters not closed 

General form 

Parts added 

Smoothness 

It will be remembered that "a" stands for the element 
which has the ninth rank, and so only values for the nintli 
rank are here concerned. The thirty values ranged from 1 to 7 
with an average of 3.3 and mean variation of 1.4. The mode 
of the distribution is at three. 

If "a*" is given the value 3.3 in equation (9). the value of D 
becomes 3.9. Using the above values for "a" and "D" in 
equations (1) to (8), the following values are obtained: 

Expressed in 

whole numbers 

Alignment (a) = 3.33 3 

Spacing of lines (b) = 3.92 4 

Spacing of words (c) = 8.26 8 

Spacing of letters (d) = 11.87 12 

Heaviness (e) = 12.46 12 

Size (f) = 13.05 13 

Slant (g) = 14.54 15 

Neatness (h) = 15.13 15 

Formation of letters (i) = 17.42 18 



Sum =99.98 100 



Score Card for Measurement of Handtvriting 27 

Referring now to Table 9, the following set of equations is 
procured : 

y-z=: 1.094 D (a) 

x~z — 1.626 D (b) 

w-z = 2.769 D (c) 

v-z — 3.106 D (d) 
Adding these equations we get v+w+x+y — 4z = 8.595 D 
Adding z — z=0, 

Tlie equation becomes v+w+x+y+z - 5z =: 8.595 D 

Since tlie sum of v+w+x+y+z = 18, tlien, 

5z= 18 - 8.595 D (e) 

The distributions concerned here are entirely different from 
those just considered and so both "z" and "D" must be de- 
termined. To determine "z" the data obtained for the rubrics 
at the bottom of the sheet on page 26 is used. The average 
value for the lowest rank in this case is 1.44, with an average 
variation of .54. The range was from 1 to 3. Substituting 
this value for "z" in equation (e), "D" becomes 1.27. Usin'.r 
this value of "D" in equations (a) to (b), the following values 
are derived : 



Expressed in 

whole numbers 

Parts added (z) 1.4 4 1 

Parts omitted (y) 2.81 3 

Letters not closed (x) 3.48 3 

General form (w) 4.9 2 5 

Smoothness (v) 5.35 6 



18.00 18 

RESULTS FROi\I ELEMENTARY SCHOOL TEACHERS 

Blanks like those sent to the writing teachers (see page 20) 
were given to all the elementary-school principals in Austin, 
Texas. Time was taken to explain carefully just what was 
wanted, and then in turn the matter was presented by the 
principal to the teachers. Seventy-five blanks were returned. 
Table 10 gives the distribution of the judgment upon the im- 
portance of the various points. This table has the same plan 
as Table 1, and so needs no explanation. 



28 Bulletin of the University of Texas 

TABLE 10 

Showing distribution of judgments by elementary school teachers upon the relativt 
Importance of the various points listed below. For key see Table 1. 



Spacing of letters. -- 

Spacing of words 

Spacing of lines 

Slant 

Size 

Alignment 

Neatness 

Formation of letters 
Heaviness 

Parts omitted 

Letters not closed 

General form 

Parts added 

Smoothness 



1 


1.5 


2 


2.5 


1 
3 3.5 


4 


4.5 


5 


5.5 


6 


6.5 


7 


7.5 


8 


8.5 





2 


12 


8 


11 


6 


13 


5 


4 


2 


7 





1 


1 


1 


1 


4 





3 


6 


12 


6 


11 


4 


11 


2 


9 


1 


1 


1 


3 


1 








1 





9 


2 


6 


1 


10 


5 


13 


4 


15 





5 











2 


1 


7 


2 


4 


1 


5 


1 


11 


3 


16 


3 


8 


1 


1 





3 





6 


4 


11 


2 


12 


1 


7 


2 


9 


3 


8 


4 


1 





4 





3 


1 


8 


2 


4 


4 


4 


5 


10 





13 


2 


14 


6 


12 


4 


9 


2 


7 


2 


5 


1 


2 





3 


1 


4 





39 


8 


11 


1 


7 


1 


3 


1 


1 








1 




















2 


2 


3 


2 


1 





1 


1 


3 





7 


1 


18 


5 


9 





7 


2 


20 


11 


18 


8 


7 



















2 


22 


1 


25 


1 


14 


2 


6 
















50 


10 


7 





5 





2 





1 

























3 


5 


10 


20 


9 


28 
















7 


8 


28 


2 


7 





7 


1 


15 

















It should be noted that none of these distributions presents 
any such peculiarities as tbat of "neatness" or "heaviness" iti 
Table 1. Those indicating high rank are skewed to the left, 
while those indicating the low rank are skewed to the right, 
and those for the middle rank approach the normal curve of 
distribution. The next step in the procedure is to determine 
the sequence of the various rubrics, begimiing with the one 
which is most important and so on down to the one which is 
the least important. Such a sequence can be determined in a 
tentative way by an inspection of the distribution shown in 
Table 10. This is to be followed by a count of the number of 
judges who place the first point above the second, the second 
above the third, etc. These results are shown in Tabh^s 11 
and 12. 



Score Card for Measurement of Handwriting 



29 



TABLE 11 

Showing distribution ol "equal to" and "more important" judgments by elementary 
school teachers upon the various points listed below. For key see Table 2. 



Neatness 

Spacing of letters- 
Spacing of words- 
Size 

Spacing of lines.-. 

Slant 

Alignment 

Heaviness 

Least important 



ryl 
















t- 




































2 












o 




fe 


O 




8 






O 






fe 




;s 






a 
o 




o 


o 




o 




a 


•^ 


w 


M 


em 




b< 




S 


03 






G 




a 




a 


4J 

03 


C8 




« 


03 


4.3 

a 


n 
to 




^ 




a 










P^ 


;? 


IB 


w 


cc 


32 


02 


< 


K. 


M.I. 


E. 


M.l! 


E. 


M.I. 


E. 


M.I. 


E. 


M.I. 


E. M.L 


E. 


M.I. 


E. 


M.I 


14 


44 






























8 


5 


8 


41 


























7 


a 


7 


1 


28 


32 






















2 


4 


3 


2 


3 


3 


2 


52 


















7 


3 


^ 


9 


n 


fi 


11 


10 


9 


41 














2 


3 


4 


7 


3 


1 


4 


5 


3 


5 


3 


49 










2 


5 


2 


2 


4 


1 


5 


2 


7 


6 


8 


8 


3 


43 






1 


3 


3 


7 


2 


2 


2 


3 


7 


13 


3 


2 


1 


10 


8 


44 









3 




1 




1 




2 




4 




9 




14 



Showing distribution of "equal to" and "more important" judgment upon the 
various points listed below. For key see Table 3. 









•o 












a> 












m 












o 


-d 






a 






S 


















o 


s 





1 


"O 




n < 


c 




•a 




..H 


J3 




o 


03 






























O 


+J 


"t^ 


■£ 
















a 


cc 




0^ 


&4 




'e. M.L 


'e. M.L 


'e. 


M.L 


E. 


M.L 


E. 


M.I. 


Smoothness -- - - --- — _ ._ 


13 56 
4 3 
2 3 
2 2 


6 41 

4 3 

5 9 


7 
7 


43 

1-^ 




















28 39 












1 


15 


6 


^. 


28 

















It is interesting to note the different ranks accorded tlio 
rubrics in Table 11 as compared with those in Table 2. "For- 
mation of letters" and "neatness" retain the same place, while 
"heaviness" becomes last and "slant" is only seventh as com- 
pared with third place in Table 2. It would be exceedingly 
interesting to know what there is in the experiences of the 
elementary school teacher with vrriting that leads to this rather 
marked difference of opinion, but any discussion would be 
largely speculative and so probably not very profitable. 



30 Bulletin of the University of Texas 

Again, by taking one-half of the "equal" judgments in each 
case and adding them to the "more important" judgments the 
results may be reduced to percentages. Tables 13 and 14 give 
these results. 

TABLE 13 

Alignment is judged more important than heaviness by 64% of judges 

Slant is judged more important than alignment by 50% of judges 

Spacing of lines is judged more important than slant by 67% of judges 

Size is judged more important than spacing of lines by 56% of judges 

Spacing of words is judged mor-^ iiniinrtant than size by 71% of judges 

Spacing of letters is judged more important than spacing of words by 61% of judges 

Neatness is judged more important than spacing of letters by 61% of judges 

Formation of letters is judged more important than neatness by 68% of judges 

TABLE 14 

Parts omitted is judged more important than parts added by 72% of judges 

Letters not closed is judged more important than parts omitted by 62% of judges 

Smoothness is judged more important than letters not closed by 59% of judges 

General form is judged more important than smoothness by 83% of judges 

Eef erring now to Thorndike's table (see foot note page 24) we 
get Tables 15 and 16. 

TABLE 15 

The difference between the values for alignment (b) and 

heaviness (a) = .532D 

The difference between the values for slant (c) and align- 
ment (b) = .337 D 

The difference between the values for spacing of lines (d) 

and slant (c) = .653 D 

The difference between the values for size (e) and spacing 

of lines (d) = .224 D 

The difference between the values for spacing of words (f) 

and size (e) = .821 D 

The difference between the values for spacing of letters (g) 

and spacing of words (f) = .414 D 

The difference between the values for neatness (h) and 

spacing of letters (g) = .414 D 

The difference between the values for formation of letters 

(i) and neatness (h) = .694 D 

TABLE 16 

The difference between the values for parts omitted (y) and 

parts added (z) = .865 D 

The difference between the values for letters not closed (x) 

and parts omitted (y) = .453 D 

The difference between the values for smoothness (w) and 

letters not closed (x) = .337 D 

The difference between the values for general form (v) and 

smoothness (w) =rl.412 D 



Score Card for Measurement of Handwriting 31 

Using the same value of "a" (3.33) as before and remember- 
ing that a-j-b+c+d+e +f-|-g+h+i^lOO, the following values 
are obtained : 



Expressed in 

whole numbers 

Heaviness (a) = 3.33 3 

Alignment (b) =: 5.42 5 

Slant (c) = 6.75 7 

Spacing of lines (d) = 9.32 9 

Size (e)= 10.20 10 

Spacing of words (f) ^ 13.44 14 

Spacing of letters (g) = 15.07 15 

Neatness (h) = 16.74 17 

Formation of letters (i) = 19.48 20 





Expressed 


in 


whole 


numbers 


1.44 




1 




3.04 




3 




3.87 




4 




4.50 




5 




7.11 




7 





Sum 99.95 100 

Using the same values of "z" and putting v-|-w-|-x+y-f-z= 
20, the following values are obtained : 



Parts added (z) = 

Parts omitted .• (y) ^ 

Letters not closed (x) = 

Smoothness (w) = 

General form (v) = 

Sum 19.96 20 

EBSULTS FROM TE A.CHERS AND STUDENTS OF 
EDUCATION. 

The data next procured was from teachers and students of 
education. Eighty-two persons are included in this group. 
Outside of those who give instruction in the above mentioned 
subject they were all either seniors or juniors in the Universit3'' 
of Texas. The matter was presented to each class either by the 
author or by one who was familiar with all the details of the 
investigation. 

Table 17 gives the distribution of the different judgments 
from this group. 



32 



Bulletin of the University of Texas 



TABLE 17 

Showing distribution of judgments by students upon the various points liuted 
below. Tor key see Table 1. 



Spacing of letters 

Spacing of words 

Spacing of lilies 

Slant 

Size 

Alignment 

Neatness 

Formation of letters 
Heaviness 

Parts omitted 

Letters not closed 

General form 

Parts added 

Smoothness 



1 


1.5 


2 


2.5 


3 


3.5 


4 


4.5 


5 


5.5 


6 


fi.5 


7 


7.5 8 


8 5 


c 


4 


18 


7 


18 


3 


n 


1 


7 





2 


1 


3 


1 








8 


8 


9 


10 


5 


18 


2 


9 


2 


5 


1 


7 


1 


1 











1 


3 


5 


3 


9 


2 


10 


6 


17 


2 


10 





s 





1 


1 


2 


1 


3 


1 


6 


1 


6 





9 


1 


14 





23 


1 


3 





12 





2 





9 


2 


9 





10 


2 


13 


1 


13 


1 


2 


3 


7 


2 


6 





7 





e 


2 


13 


3 


15 





15 





11 


4 


6 


1 


11 





6 


1 


14 


1 


7 


1 


5 


1 


6 


1 


44 


9 


8 


1 


4 





5 


3 


9 


1 


1 


2 


] 





1 


5 








1 





1 





3 





2 





5 


3 


7 


2 


9 


1 


10 


7 


12 


1 


8 





6 


4 


3 
















6 


3 


12 


2 


14 


1 


7 


1 


5 
















19 





7 





12 


2 


3 


3 


5 
















1 


4 


4 


1 


7 





17 


4 


l.S 
















9 





11 





5 


3 


6 


3 


14 

















After determining as before which point is most important, 
which second in importance, etc., the next step is to find the 
number of judges who rank each rubric above the next lower 
in order. Tables 18 and 19 give such results. 

TABLE 18 

Showing distribution of "equal to" and "more important" judgments of students 
upon the various points listed below. For key see Table 2. 





S 




















^^ 


t/j 






























































o. 




OJ 










° 


^ 


















a 
o 


o 
tuo 


o 
to 


« 


o 








la 




S 


c 




c 


c 




s 


+^ 


'? 






cS 


a 








M 




CS 




o 


02 


tK 


i 


K 


t» 


^ 


ao 






E. 


M.I. 


E. 


M.I. 


E. 


M.I. 


E. 


M.I. 


E. 


M.I. 


E. 


M.I. 


E. 


m.i.Ie. 


M.I. 


E. 


M.l. 


Spacing of letters 


6 


CO 


































Spacing of words 


1 


3 


16 


49 






























Neatness 


3 


3 





1 


9 


43 


























Spacing of lines. 


1 


4 




4 


7 


19 


3 


52 






















Size 


3 

2 


2 

1 


I 


2 


1 
2 


3 

5 




1 


2 
5 




1 


43 
10 


1 


40 














Alignment 




Slant . 


3 
2 


2 

1 




5 



1 

1 


1 
3 



1 


7 
6 






S 
9 


1 
1 


9 
12 


2 

i 


54 
10 


1 


60 






Heaviness 




Least important. 








1 


5 


5 


5 


2 


12 


48 



Score Card for Measurement of Handwriting 



33 



TABLE 19 
Showio. the distribution of Veaual to" and ^^ ore import aot" ^f^^^^^ot student, 
=r?u^d.^rnron^%lLSi'Si a^un'tflorthe ^numbers bein. smaller in thi. 
table than in the others. 



Parts omitted 

Letters not closed- 
Smoothness 

Parts added 



Least important- 



! . 1 










OJ 


















"O 










(U 






"O 




tj 






•c 




E 


c 


c 


■o 








x: 






«i 




o 


" 














^ 


s 


02 


53 


r 


E.M.I. 


E.M.I. 


E. 


M.I. 


E.jM.I. 


9,7 














9 


5 27 












5 


1 


3 K 










9 


4 8 


5 





29 


1 


5 


3 


€ 


19 





Reducing the "more important" and "equal judgments" lu 
percentages Tables 20 and 21 are derived. 

TABLE 20 

Slant is judged more important ^ha^ /;fi''''°f^^t^^; VSMVM'S of judges 

Alignment is Judged more important than slant by ^^ .^^^^^^ 

Size is judged more important ^han alignment by ^^^^^ . ^^ .^,^g^ 

Spacing of lines is judged more 'mPOf^'^"^,,^''„t"/of liMs" by 65% of judges 

Neatness is judged more ™Poi-t"nt "lan spac ng of hues oy ^^ .^^^^^ 

Spacing of words is .Judged more miportan^^ by— -70% of judges 

IrrZfiol l,f l^le'^^i^TuteXto^^Ko^tUftL^n^ spacing of letters by 70% of judges 

TABLE 21 

Smoothness is judged more important than P^^^s added^^^J,"/!;-::::::::^^ of jud|es 

Letters not closed is judged more -i^oj;^"* *^|^t°terT not Sd by 58% of judges 

Referring again to Thorndike's table (see footnote, page 
24) these differences may be stated in terms of probable 
error D 

TABLE 2 2 

Difference in values of slant (^^ ^^^VaJd'sfant^^^ ■■■■■.: '•= '.653 S 
Difference in values of alignment («,^ .^"^.J^^'^^^t - 

Difference in values of spacing of letters (h) and spacms oi^ ^^^^ ^ 
Differen'c'e'in values of' formation of" letteVs" (iV and'spacing^^^^^ ^ 
of letters 



34 



Bulletin of the University of Texas 



TABLE 23 

Difference in values of smoothness (y) and parts added (z)-= .262 D 

Difference in values of letters not closed (x) and smoothness= .367 D 
Difference in values of parts omitted (w) and letters not 

closed = .299 D 

Difference in values of general form (v) and parts omitted= .149 D 

Manipulating' these results as has been done before and 
putting a=3.33 and z=1.44, the following values are derived: 

Expressed in 

whole numbers 

Heaviness (a) = 3.33 3 

Slant (b) = 7.06 7 

Alignment . (c) = 9.61 10 

Size (d) = 9.61 10 

Spacing of lines (e) = 9.90 10 

Neatness (f ) = 12.15 12 

Spacing of words (g) r= 12.72 12 

Spacing of letters (h) = 15.76 16 

Formation of letters (i) = 19.85 20 

Sum 99.99 100 

Expressed in 
whole numbers 

Parts added (z) = 1.44 1 

Smoothness (y) = 2.58 3 

Letters not closed (x) := 4.23 4 

Parts omitted (w) = 5.55 • 6 

General form (v)= 6.19 6 

Sum 19.99 20 

The last step in the procedure is to get results by means of 
combining the judgments of the three groups. The distribu- 
tion of the judgments by such combination is given in Table 28. 



TABLE 24 

Showing distribution of jiuigment.'! by all jurlffes upon tlie relative importanre of 
the various i)oints listed below. For key see Table 1. 



Sparing- of letters, __ 
Spacing of words_-_ 

Spacing of lines 

Slant 

Size 

Alignment 

Neatness 

Formation of letters 
Heaviness 

Parts omitted 

Letters not elosed.. 

Oeneral form 

Parts added 

Smoothness 



1 


1.5 


2 


2.5 


3 


3.5 


4 


4.5 


5 


.5.5 


6 


6.5 


7 


7.5 


8 


8.5 


<i 


9 


33 


17 


39 


15 


34 


8 


18 


3 


22 


2 


15 


4 


4 


1 


i 


4 


11 


16 


26 


17 


38 


9 


29 


6 


23 


3 


19 


7 


16 


2 








3 


3 


17 


9 


19 


5 


26 


11 


34 


9 


36 


3 


28 


3 


10 


2 


12 


5 


9-> 


11 


19 


4 


18 


1 


24 


8 


32 


3 


35 


2 


5 


3 


24 


2 


18 


10 


31 


5 


28 


2 


24 


7 


26 


6 


25 


5 


4 


4 


12 


<) 


13 


2 


19 


5 


13 


6 


25 


11 


.36 


2 


35 


8 


3.3 


21 


28 


8 


26 


6 


14 


3 


26 


3 


13 


5 


11 


2 


20 


1 


112 


22 


26 


4 


20 


fi 


8 


6 


7 


1 


3 


3 


5 





6 


1 


4 


7 


9 


7 


7 


6 


9 


1 


8 


9, 


15 


5 


22 


5 


33 


7 


14 


7 


24 


8 


44 


16 


51 


24 


15 
















12 


7 


41 


10 


70 


.3 


.36 


5 


17 
















&0 


22 


43 


3 


21 


2 


8 


3 


10 
















1 


4 


6 


4 


17 


14 


m 


24 


81 
















52 


21 


51 


3 


17 


3 


17 


5 


32 

















Score Card for Measurement of Handwriting 35 

The calculation of the "equal to" and "more important" 
judgments gives Tables 25 and 26. 

TABLE 25 

Alignment (b) is judged more important than heaviness (a.) by -•_55% of judges 

Slant (e) is judged more important than alignment by 59% of judges 

Spacing of lines (d) is judged more important than slant by 52% of judges 

Size (e) is judged more important than spacing of lines by 59% of judges 

Spacing of words (f) is judged more important than size by 58% of judges 

Spacing of letters (g) is judged more important than spacing of words 

by - 68% of judges 

Neatness (h) is judged more important than spacing of letters by 51% of judges 

Formation of letters (i) is judged more important than iioatn<^'ss by 70% of judges 

TABLE 26 

Parts omitted (y) is judged more important than parts added by 75% of judges 

Letters not closed (x) is judged more important than parts omitted by--55% of judges 

Smoothness (w) is judged more important than letters not closed by 62% of judges 

General form (v) is judged more important than smoothness by 63% of judges 

INIanipulating these results as before the following values 

are found : 

Expressed in 

whole numbers 

Heaviness (a)=: 3.33 3 

Alignment (b) = 4.65 5 

Slant (c) = 7.05 7 

Spacing of lines (d) = 7.58 8 

Size (e) = 9.98 10 

Spacing of words (f) = 11.61 12 

Spacing of letters (?) = 16.55 16 

Neatness ( h) = 16.81 17 

Formation of letters (i) = 22.35 22 



Sum 



99.91 


100 






Expressed 


in 


whole numbers 


1.44 


1 




3.92 


4 




4.38 


4 




5.50 


6 




6.74 


7 





Parts added ( z) ^ 

Parts omitted ( y) ^ 

Letters not closed (x) ^ 

Smoothness ( w) ^ 

General form ( v) ^ 

Sum 21.98 22 

RESULTS BY MEANS OP THE REGRESSION EQUATION 

The first step in this procedure is to determine the correla- 
tion between "general merit" in handwriting and each of the 
nine points in the preceding list as well as the correlation of 
each point with each of the remaining eight. In order to get 
data for this part of the work, sixteen samples of writing were 
selected from the Thorndike seole. The reason for taking 



36 Bidletin of the University of Texas 

these samples from the scale is that they have been carefully 
graded on the basis of general merit. One sample was taken 
from each division of the scale except in the case of division 
15, from which two samples were selected. The samples were 
then ranked on the basis of each of the nine points in the pre- 
ceding list. The ranking was done by tifty persons, all of 
whom were university students. Many of them were mature 
persons with experience in teaching. The following directions 
were given each one : 

Dear Sir or Madam: 

Will you please rank the accompanying samples of hand- 
Avriting on the basis of the nine points at the top of the sheet 
attached to this. The numbers at the top of the sheet upon 
which the samples appear correspond to the numbers at the 
left of the attached sheet. 

To rank the samples on the basis of slant, decide which 
sample has the best "slant" and put the digit "1" in the slant 
column in the square opposite the number which corresponds 
to the number at the top of the sample. Then decide upon the 
second best sample as to. '"slant" and place the digit "2" in its 
appropriate rectangle as before. Proceed in this way until 
each of the samples have been ranked. When all have beoi 
ranked on the basis of "slant," proceed in the same manner Avith 
"alignment," etc., until the samples have been ranked with refer- 
ence to each of the nine points. If two or more of the samples 
should have the same rank, put down the digit representing 
this rank for each of them. 

The following table will give the points to be taken into 
consideration when considering the diiferent rubrics : 

Slant 

Uniformity 
Mixed 
Alignment 

To determine this a straight edge may be used as a line or 
the judge may turn the sample edgewise, close one eye, and 
look down the different lines. Consider also that lines should 
be perpendicular to the edge of the paper. 



Score Card for Measurement of Handwriting 3r 

Spacing- of words 
Uniformity 
Len^h of space 

Formation of letters 

Smoothness 

General form 

Parts omitted or added 
Spacing of lines 

Uniformity 

Too close 

Too far apart 

Heaviness / ^U (XJ ) 

Spacing of letters 
Uniformity 
Too close 
Too far apart 
Neatness 

Blotches or careless 
Size 

Uniformity 
Too large 
Too small 
In addition to the written directions the problem was ex- 
plained to each judge by the writer. The work was all done 
in the laboratory, and the time required was credited upon 
the regular work of the course which the student was taking, 
so that there was no occasion to rush. The time taken was 
usually one to two hours. Each person recorded his judgments 
in the following form : 



58 



Bulletin of the University of Texas 



10 



13 



C3 o 






16 



15 



15 



13 



Signed: J. C. I. 



This data was formulated as shown in Table 27. It will be 
noted that this table shows graphically the correlation between 
^'general merit" and "spacing of letters." Forty-five such 
tables are required : 



Score Card for Measurement of Handwriting 



39 



PS i 

< "S 



0)^ 









o J» 

55 



o 



to I 1 1 1 I 1 1 1 1 rH 1 1 1<M <0 O 
t-( ' 1 1 1 1 1 1 1 ! Ill "* 


in 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 
in III 1 1 1 1 1 1 1 

rH 1 1 1 I 1 1 1 1 1 1 1 1 1 1 [ j 


\a 1 1 1 C 1 rH C> r-H O 

■^ 1 1 1 1 1 1 1 1 1 1 1 i 


lO 1 1 1 1 1 1 I 1 1 r-1 1 1 1-1 « R 1 
M* 1 t 1 1 1 1 1 1 1 II 1 
rH 1 1 1 ' 


tX 1 rH 1 IrHrHinSiCC-l 

rH 1 1 1 j 1 1 1 11 rH rH 


lO 1 1 1 1 1 t rH C-l C-1 rH 1 

CO 1 1 1 1 1 1 1 1 1 1 1 1 
rH 1 1 1 1 1 1 1 1 j 1 1 1 


CO 1 1 1 1 1 1 1 1 C<1 M i-l ^ CO « rH i-H 
i-H 1 1 1 1 1 1 1 1 '^ 


lO 1 1 1 1 1 1 1 1 1 rH 1 rH (M 1 1 1 
C-i ' ' ' 1 1 1 1 1 1 1 III 


s i 1 r i i ! i I'^sg'^^ 1 1 


in llllllllrHi<Mr-.rHlii 
T-{ 1 1 1 1 1 i t 1 j III 


rH 1 1 1 1 IrH 1 1I>>-hO0^IM 1 TH 


O ' 1 1 1 ! 1 1 1 1 1 III! 


O 1 1 ICl IrH lr-(^K>OCr:i-i 1 1 1 


lO 1 1 1 rH 1 N 1 1-1 C<5 rH M rH 1 1 1 1 
■^ i i i 1 1 1 1 1 1 


Ci 1 1 irH icorHa-iccajinrHiM | i j 


in ' 1 1 1 '^ 1 '"' 1 1 1 1 1 1 1 1 ' 

00 i i 1 i 1 1 1 1 1 1 1 1 1 1 


00 1 1 r C<1 (M 1-1 to tiJ 00 <N r-l rH ] | | | 


in lllC>li.^(NCC(M1l"'"'i 

^ III 1 ! 1 1 1 1 1,1 


t~ 1 C-l rH rH !M O c» O) c^ t-< rH rH 1 1 C-l 1 


m "^^ ' '^ 1 1 1 1 1 1 1 1 

<a 1 1 1 1 1 1 1 1 1 1 1 1 1 1 


iO ' 1 rH CO .-^r -* CC in ' ' rH ' j I j j 


in '-'''^'"''"'''""^llllllll 
in 1 ] 1 1 1 1 1 1 1 1 


in rHMeOOC5(>4C.« ' ' ' ' |rH 1 1 

^ '[III II 


in IfMrHlOinrHrH IrH 1 1 1 " [ [ j 
"^ ' ' 1 1 1 1 1 t 1 


^ rHl'Mnt-OC^linrHrH 1 1 | | ] [ | 
i-l i ! 1 1 1 I 1 


m i-i"-i'-i'-''-'i?^''-''l|||||| 

CO 1 1 1 1 1 1 1 1 1 


CO COOOC9CO 1 <N rH 1 ! | [ | | | | 


in C^ ^ t^ rH .* IrH ' ' 1 j 1 1 1 1 1 
Ol j 1 1 1 1 1 1 1 1 1 


CM .-t O OO .* IC rH • ' ! ! ! 1 I 1 I I 
"M . I I 1 1 1 1 1 t 1 1 


in " "^ '"' '^ 1 1 I 1 1 1 1 1 1 1 1 1 


r^ ^'MnHlrHljIIIIII]] 
^ ' 1 1 1 1 I 1 1 1 ' ' 


rHC<lC0iniOt01>Q0»OrHIMC0M.inj0 



40 



Bulletin of the University of Texas 



6 2 D- 
Applying the formula 1*^1 the following table of 



correlations is obtained 



n(n--l. 



TABLE 28 



03 ~ 4) 



a. 
































































s 




4> 


«H 




^ 






o 




O 


O 




a 






a 
o 




O 


o 




o 







■M 


tn 


bl 


M 




ko 




6' 




QJ 


a 


a 




O 
















a 

03 





fc- 


03 


03 


ca 


S 


C3 




P 


<U 


ft 


ft 


ft 




\^ 


K 


02 


CB 


CO 


!» 


CO 


< 


95 


94 


94 


91 


82 


91 


90 


94 




91 


92 


91 


80 


91 


89 


91 






90 


88 


78 


87 


89 


91 








89 


80 


89 


90 


90 










76 


88 
75 


88 
76 
87 


90 
77 
89 
90 



General merit 

Formation of letters 

Neatness 

Spacing of letters 

Spacing of words 

Size 

Spacing of lines 

Slant 

Alignment 



The probable error of these correlations is found by the 

formula: P.E.=.706 -^-^ 

1 u 
Applying this formula to the lowest correlation (70), the 

probable error is found to be .091. Since "„" remains constant, 
the other P.E.'s will not be greater than this. This value of 
P.E., one-eighth of the coefficient of correlation, is quite satis- 
factory. 

Attention is called to the order accorded the different rubrics 
on the basis of the correlations with general merit. The most 
interesting point is that "alignment" has the same rank as 
' ' neatness ' ' and ' ' spacing of letters, ' ' which is very much higher 
than accorded it in any of the previous results. The average 
of the intercorrelations for "alignment" gives it a rank of 
three. Attention is also called to the fact that "size" is ranked 
eighth either by its correlation with "general merit" or by its 
intercorrelations which is lower than the rank accorded it in 
the other results. 

One objection which may be raised to the judgments upon 
which these coefficients are calculated is that the opinions upon 
any one point are not independent of the influence of the other 



Score Card for Measurement of Handwriting 41 

elements, i. e,, the judgments upon "formation of letters" are 
not independent of "slant," "spacing of letters," etc. 

The regression equation gives the value which may he at- 
tached to any one point, independent of any relation it may 
have with any other element. The equation will be of this 
general form : 

General merit=(bi)slant-(-(ho)size+(b3)formation of letters 
H etc. This equation would require that nine different co- 
efficients be determined. The work for determining such co- 
efficients is so involved that an equation with more than four 
variables is seldom used. This necessitates the grouping of the 
elements in the following manner on the basis of the inter- 
correlations. 

Group one is composed of "formation of letters," "spacing 
of letters, " " spacing of words, ' ' and ' ' spacing of lines. ' ' Group 
two contains "slant," "alignment," and "neatness," and group 
three has in it "heaviness" and "size." The regression equa- 
tion now becomes: General merit (Xi)=bi (formation of let- 
ters) (x2)-|-b2 (slant) (x3)--l-b3 (heaviness) (xj or in a more 
simple form it is, Xi=biXo-}-b2X3-|-b3X4. 

The first step in determining the coefficients b^, bo, and b., is 
to find the correlation coefficients of the first order in terms of 
the coefficients of zero order found in Table 28. The equations 
for this liave the following form : 

r^„ — ri3 . r„3 



Other equations of the same type must be used for r,, ,, r„.i etc. 
These calculations are to be followed by the second step whicli 
is the determination of the correlation coefficients of the second 
order in terms of the coefficient of the first order. The equa- 
tions used here have this form : 



"^^ l/l-rL.3 -/l-rL^ 
Again, other equations must be used for r,, „^. r„3.,, etc 
The third step is to determine the standard deviation of 
higher orders. The equations used here have this form : 

^.=.3.=<^x i/i=?; T^i— r:3, i/i— r:,,3 



42 



Bulletin of the University of Texas 



Other equations of the same general type must be used for 
0-, j^^, 0-3 J, ^ etc. 

Since the desired results should not be proportional to 
cTi, o-o and 0-3 these are all made equal to unity. This data^ can 
now be used for calculating the coefficients bib2b3. The equa- 
tions used here take the form 



Others similar to this must be used for b,3 „, and b„ „. Th* 
following table gives the results obtained : 

TABLE 29 



Showing values of the different coefficients required in the 
solution of the first regression equation. 



Zero order 


First order 


Second order| 




coefRcients 


coefficients 


coefficients 1 










CT, 


=.279 


12 


.95 


12.3 


.77 


12.34 


.73 


a 


=.291 


13 


.90 


13.2 


.39 


13.24 


.26 


a 


=.408 


23 


.89 


23.1 


.26 






a 


=.661 


14 


.77 


14.2 


.20 


14.23 


.13 


b, 


=.693 


24 


.77 


24.1 


.20 


24.13 


.16 b, 


=.244 


14 


.77 


14.3 


.36 






b, 


,3=.058 


34 


.74 


34.1 


.17 


34.12 


.12 






24 


.77 
.74 


24.3 
34.2 


.36 
. .19 










34 

















The equation then becomes Xi=.693x2-|- .24-1x3-}- .057x, 

If 100 points are divided among the three groups before men- 
tioned in the ratio indicated by these coefficients, 69.6 is the 
value given group one, 24.5 is the value for group two, and 5.0 
is given group three. 

In order to determine how the 5.9 points are to be distributed 
between "size" and "heaviness" another regression equation 
of three variables is used. The solution of this gives "lieavi- 
ness" equal to 2.5 anad "size" equal to 3.4. 

The distribution of the 24.5 points in group two is determined 



.■}. Tlie regression equation has been worked out with the product moment eo- 
efBcient, but for all iiractirnl purposes it is thought that the rank coeffifients are 
sufficiently accurate. 



Score Card for Measurement of Ilandivriiing 



43 



by another regression equation of four variables. The solution 
of this gives "neatness" equal to 9.1, "alignment" equal to 
11.4 and "slant" equal to 4. 

The distribution of the 9.5 points in group one is also de- 
termined by a regression equation. In this group "spacing of 
words" and "spacing of lines" are considered of equal value on 
account of their intercorrelations. The solution of the equation 
gives "formation of letters" equal to 29.3, "spacing of letters" 
equal to 20.2, "spacing of words" and "spacing of lines" each 
to 10. 

The results of the investigation can now be summarized in 
the following table : 

TABLE 30 

Showing values and ranks obtained for the difierent rubrics by the different 
methods. 





BaokB 


Values 




< 


C 

-a 

3 


■SO.C 

Hi 


tux: 

~ ni 
~ a; 


a 
_o 

+^ 
a 

St 
|g 


a 
o 

"3 
a) o 


O 

So 

OJ z 


Si 


a 


S o 2 
vxiri 


ins: 

■BJ 


> 
C 

o 

ll 

£; 3 

>. <s 

33 


Heaviness _ ._. .-- 


9 


9 


9 j 5 





9 


9 


3 


& 


3 


12 


3 






Alignment — . . _ . .- 


8 


7 


S 9 


9 


.", 


.'? 





10 


5 


3 


T< 




1 




Slant — --. _-...-_ 


7 1 S 


7 3 7 


1 7 


■, 


7 


7 


15 4 















Spacing of linos __ .. . . . 


6 


7 


C 


8 


7 1 E 8 


10 


i) 


1 


10 












Size - 


5 





5 


4 


8 
5 


8 1 8 


10 


10 


10 


13 


4 








Spacing of words - .. .- . 


4 


3 


4 


7 


5 


5 :-2 


12 


14 


8 


in 








Spacing of letters 


3 


2 


3 


6 


2 


2 


2 IC 


16 1 15 


12 


20 




2 


t 


2 


1 


2 


4 


6 


n 


12 


17 


15 


9 








1 


1 


1 


1 


1 


1 


22 


20 


20 


18 


•?9 








5 


5 


5 


5 








1 


1 


1 


1 














Parts omitted 


4 


9 


4 


4 








4 


6 


3 


3 
















3 


3 


3 


3 








4 


4 


4 


3 
















1 


1 


1 


2 








7 


6 


7 


5 
















9 


4 


9 


1 








6 


3 


5 


6 








' 













Final value may now be gotten by taking the average of the 
value given by the regression equation and "all judges." Such 
values expressed in whole numbers are as follows: 



44 Bulletin of the University of Texas 

Heaviness 3 

Slant 5 

Size 7 

Alignment 8 

Spacing of lines 9 

Spacing of words 11 

Neatness 13 

Spacing of letters 18 

Formation of letters 26 

Sum 100 

The value given to ''formation of letters" is 1.17 times as 
great as that given to the same rubric by "all judges." To get 
final values for the points subsidiary to "formation of letters" 
the values given for these various elements on page 43 may be 
multiplied by 1.17. Such a procedure gives the following re- 
sults : 

Parts added 2 

Parts omitted 5 

Letters not closed • • • • 5 

Smoothness 6 

General form 8 

Sum 26 

USES AND FORMS OF THE SCORE CARD 

In addition to the scientific uses ^yhich have already been 
suggested there are many other uses for the score card. Teach- 
ers of writing can use it to stimulate interest by scoring the 
writing of each pupil or by having the pupils score their own 
writing or the writing of other pupils. This would call the at- 
tention of the pupils to the elements of good writing and the 
relative merits of each point. It would also give the pupils a 
definite idea of the progress which they are making. Super- 
visors might use it for comparison of teachers and grades as 
well as for judging different methods of teaching. For any one 
of these purposes the author suggests the following form : 



Score Card for Measurement of Handwriting 45 

STANDARD SCORE CARD FOR THE MEASUREMENT OF HANDWRITING 



Sample 
■ s 3 

2. Slant 5 

Uniformity 
Mixed 

3. Size •. 

Uniformity 
Too large 
Too small 

1. Alignment 8 

5. Spacing of lines 9 

Uniformity 
Too close 
Too far apart 

6. Spacing of words 11 

Uniformity 
Too close 
Too far apart 

7. Spacing of letters 18 

Uniformity 
Too close 
Too far apart 

8. Neatness 13 

Blotches 

Carelessness 

9. Formation of letters (26) 

General form 8 

Smoothness 6 

Letters not closed 5 

Parts omitted 5 

Parts added 2 

TOTAL SCORE 


1 


3 


3 


4 


5 


6 


7 


8 


9 


10 


11 


12 


13 


14 


15 








It can also be used as a permanent record for the grades of 
pupils in writing'. For such a purpose it would probably be 
best to make it the same size as the white card recommended 
by the United States Bureau of Education. For such a pur- 
pose the following abbreviated form is recommended : 



46 Bulletin of the University of Texas 

Pupil Age Grade. 



3 
03 

> 

1. Heaviness i 

2. Slant 5 

3. Size 7 

4. Alignment 8 

5. Neatness 13 

fi. Spacing of 

Letters IS 

Words 11 

Lines 9 

7. Formation of letters (-^e) 

Smoothness 6 

General form 8 

Parts added : 2 

Parts omitted 5 

Letters not closed 5 

TOTAL SCORE . 


d 

a 

CO 

l-H 


6 

S 
•a 
d 

CO 


6 

a 


d 

a 

a 


6 

a 


6 

a 

CO 


d 

a 

i3 


d 

a 

CO 


i 

-3 


d 

a 
1 







Another important use which suggests itself for the score 
card is in making a careful and detailed study of the writing 
of a single individual. Such a clinical study is often highly 
desirable when a pupil does not make proper progress in 
writing. 

The scoring of a sample of handwriting by means of the 
score card is a simple process. If the score for slant is desired 
from the value shown in the last mentioned form, it is only 
necessary to examine the sample carefully for uniformity and 
degree of slant and then to assign a value in accordance with 
this examination. The scoring for any other point is similar 
to this. 

OBJECTIONS TO THE SCORE CARD. 

In addition to the objections wdiich have already been sug- 
gested no doubt others will be urged. Some will argue that 
the time required for grading by it is a decided disadvantage. 



Score Card for Measurement of Handwriting 47 

Tliis will doubtless be true when a teacher begins the use of the 
method, but some preliminary experiments by the author show 
clearly that a very little practice reduces the time required for 
the use of the score card very much. It is possible that the 
extra time required will give more accurate results. If sucli 
should prove to be true, its advantage would be established. 

Others may argue that such grading will be formal in every 
respect, and so no better than the usual grading. If such is 
the case, it cannot be used as an objection to the method, but 
rather against the person who uses it. 

The last objection to be noticed is that the grading of any 
single point such as slant harks back to the old percentile 
method. The only answer which need be made to this is lo 
call attention to the very accurate and scientific work done by 
the agriculturalists in work very similar to this. 

EXPERIMENTAL WORK WITH THE SCORE CARD. 

The problems which it is possible to attack by means of the 
score card are many, and only a few can be mentioned in addi- 
tion to those suggested in the preceding pages. 

One of the most interesting problems in connection with the 
score card is the effect it will have upon the grading of teach- 
ers. This problem has been studied by Kelly (10) for other 
methods of grading, and it is hoped that the present method 
can be compared with all other methods in the near future. An- 
other problem might deal with the effect of practice in the use 
of the score card. This problem has been dealt with hj the 
author (7) in connection with the Ayres Scale, and the same 
general plan should be used with the score card. 

Still another problem is suggested by certain training given 
in agricultural courses known as judging. Briefly, this means 
that after a student has been trained for a considerable period 
of time in the use of the score card he is then given the problem 
of judging. In judging he evaluates the product without the 
aid of the score card. In the same manner it would be very 
interesting to train, say, six judges in the use of the score 
card, and then allow three of them to judge the writing while 
the other three continued the use of the score card. If such 



48 Bulletin of ike University of Texas 

training will produce experts in the judging of handwriting it 
would be worth while. 

The table of correlations on page 40 suggests another very- 
interesting problem. This table indicates clearly that several 
of the elements correlate with general merit in about the same 
degree. From this it may be argued that all the elements here 
proposed are not needed in order to get an accurate measure- 
ment of handwriting. The first regression equation nsed gives 
values which could be put in score card form. A comparison 
of grades gotten by means of such an al)breviated form and the 
forms already proposed gives a basis for some very interesting 
work. 



BIBLIOGRAPHY 

1. Ayres, Leonard P., "A Scale for Measuring the Quality of 

Handwriting of School Children"; Russel Sage Foun- 
dation, New York. 

2. Ballou, F. W., "Scales for the Measurement of English 

Composition"; The Harvard Newton Bulletins, No. 
II, Sept., 1914. 

3. Boyce, A. C, "A Method for Guiding and Controlling the 

Judging of Teaching Efficiency"; The School Review 
Monographs No. 6, pp. 71-82. 

4. Buckingham, B. R., "Spelling Ability: Its Measurement 

and Distribution ' ' ; Columbia Contributions to Edu- 
cation, T. C. S., No. 59. 

5. Courtis, S. A., "Manual of Instructions for Giving and 

Scoring the Courtis Standard Tests in the Three R's." 
Dept. of Cooperative Research, Detroit, Mich. 

6. Gray, C. Truman," Variations in the Grades of High-School 

Pupils"; Warwick and York, Baltimore. 

7. Gray. C. Truman, "The Training of Judgment in the Use 

of The Ayres Scale for Handwriting"; J. E. P., Feb., 
1915. 

8. Gray, W. S., "A Tentative Scale for the Measurement of 

Oral Reading, in 'The Measurement of Ability in 
Reading' "; bv Thorndike, Teach. Col. Rec, 1-17, S. 
1914. 

9. Hillegas, M. B., "A Scale for the Measurement of Quality 

in English Composition" ; Teach. Col. Rec, Sept., 1912. 

10. Kelly. F. C, "Teachers' Marks"; Columbia University 

Contributions to Education, T. C. S., No. 66. 

11. Kelley, Truman L., "Educational Guidance"; Columbia 

Contributions to Education, T. C. S., No. 71. 

12. Reudiger, W. C, and Strayer, G. D., "The Qualities of 

Merit in Teachers"; J. E. P., 1. 

13. Starch, Daniel and Elliott, E. C, "Reliability of Grading 

Work in IMathematics" ; School Review 21, 254-259. 

14. Stowe, A. Monroe, "A Method of Recording and Report- 

ing Critical Observations of Classroom Instruction"; 
Pub. by tlie author, DePauw Un., Greencastle, Ind., 
1913. 

15. Thompson, Thomas E., "Minimum Essentials"; Sheets of 

Graded Questions in Arithmetic and Language, Ginn 
and Co. 

16. Thorndike, E. L., "Handwriting"; Teach. Col., Rec. Vol. 

11. 



50 Bulletin of ihe University of Texas 

17. Thorndikc, E. L., ''The Measurement of Achievement in 
DraMang"; Teach. Col. Kec, Vol. 14. 

-8. Witham, C. W., "School and Teacher Measurement"- J 
E. P. 5, 207-278. ' ' 

19. Ynle,_Udney G., "An Introduction to the Theory of Sta- 
tistics"; Charles Griffin and Co., London, Chapter 12. 





LIBRARY OF CONGRESS 



021 775 748 4 *