Daniel Starch.

Educational psychology online

. (page 28 of 41)
Online LibraryDaniel StarchEducational psychology → online text (page 28 of 41)
Font size
QR-code for this ebook

prepared by C. T. Gray calls attention to a similar set of elements.
The methods of judging penmanship suggested by Palmer and by
Zaner aim likewise to center attention upon defects and excellencies
in various essential aspects of writing.


Processes or Steps Involved in Spelling

The child learns to spell by seeing or hearing the letters of a given
word, and by thinking, speaking, or writing them in the order in
which they are seen or heard. Stated in more minute detail, the
successive steps are substantially as follows:

(i) The reading of the word, that is, the sight, sound and pro-
nunciation of the word as a whole which involves all the elements
of the reading process and need not be enumerated here. (See
Chapter XVI.) These are presupposed as the child usually has
learned to read the word before he learns to spell it. At this point
the successive steps in the spelling process as such begin.

(2) Reception upon the retina (or the ear) of the visual (or
auditory) stimuli of the first letter of the word.

(3) Transmission of the visual (or auditory) impressions from the
retina (or ear) to the visual (or auditory) centers of the brain.

(4) Arousal thereby of mental images and other associations
of meaning.

(5) Transmission of the impulses from the visual (or auditory)
centers to the motor-speech centers or to the motor- writing centers.

(6) Transmission of motor impulses from the latter to the
speech-organs or to the writing-muscles. This occurs very prob-
ably even in the silent learning of spelling since silent reading is
accompanied by the so-called inner speech.

(7) Execution of the speaking or writing movements in pro-
nouncing or writing the letters.

(8) Return kinesthetic impulses from the speech or wTiting mus-
cles to the sensory centers and then to the motor speech or writing
centers. This series of steps from (2) to (8) is then repeated for
the second letter, for the third letter, and so on to the end of the

The steps here outlined are the ones involved in learning the
spelling of a word. In the perfected process, however, steps (2)
and (3) and possibly (4) drop out and step (5) is inaugurated



directly either through step (i) or through the idea or image of
the word to be spelled or written, and from then on the whole
process of writing or spelling the word consists of a circular series
of automatic connections between steps (6), (7), and (8) in which
(8) for the first letter of a word acts as stimulus to step (6) for the
second letter and so on for the succeeding letters of a given word.
Step (8) of each letter always acts in turn as the stimulus for the
series (6), (7), and (8) of the succeeding letter so that in the fin-
ished habit the mere pronunciation, sight, image or idea of the
word automatically brings about the succeeding links involved
in naming or writing the letters in correct order.

Economy in learning to spell consists largely in providing con-
ditions 4jnder which the half dozen links here outlined may be es-
tablished most easily, most quickly, and most permanently for
the words whose spelling a child should know.

Little is known directly concerning the manner of operation of
each of these factors. The most important step, if any one is more
important than any other, possibly is number (8). This link
determines what the next letter shall be in the automatic writing
of a word. In the original learning of the spelling of a word, steps
(2), (3), and (4), which together constitute the perception of the
letters, are highly important since the establishment of the other
links depends upon the accuracy with which the letters themselves
are perceived or observed. It seems probable, although not certain
in the absence of pertinent experimental data, that a considerable
part of the ditSculty of learning to spell, lies in the inaccurate
observation of successive letters of a word. The awakening of
mental images is probably very important, although our informa-
tion as to the types of imagery concerned, the extent to which
they are essential, and the methods of arousing them, is relatively

The Measurement of Efficiency in Spelling

(i) Methods of Measurement. On the face of it, it would
seem to be an easy matter to devise a definite and objective method
of testing attainment in spelling. All that would seem to be neces-
sary would be the selection of a series of words and the determina-
tion of the number or percentage of these words that a pupil or
class can spell correctly. But a closer study of the possibility of
measuring spelling ability reveals a number of complicated prob-


lems. What sort of words should be used as a spelling test? How
should they be presented to the pupils? How should they be
scored? Should any word be considered equal to any other word,
or should different values be assigned to different words?

We shall not enter here into any critical discussion of the prin-
ciples involved in the construction of spelling tests, nor into a
consideration of the technique of administering and scoring them.
Some of the methods of measuring spelling ability will be men-
tioned briefly.^

Up to 1913, tests of spelling ability were made either by series of
arbitrarily selected words which were presented either as isolated
words or as parts of dictated sentences, or by determining the
percentage of misspelled words in spontaneously written composi-
tions. Since 1913, several more or less scientific methods of
measuring spelling ability have been devised.

The writer, ('15) prepared one test consisting of 6 lists of
100 words each by making a selection of words at certain inter-
vals from the dictionary and then discarding all technical and
obsolete words. The words in each list were then arranged in the
order of length. Each list as a whole was found to be practically
identical in difficulty with every other list. Average standards of
attainment were then prepared for the various grades as shown
in the following table, which gives the percentage of words of any
one of the 6 lists spelled correctly at the ends of the respective
years :


Attainment in spelling. After Starch ('15)
Grades 12 3 4 5 6 7 8

Percentage of words correct lo 30 40 51 61 71 78 85

The author has more recently prepared a different method of
testing spelling ability on the basis of the 2,626 most common words
in the English language. This list is useful as a study list as well
as a test list and will be distinctly more valuable because of this
double purpose. The plan by which these words were selected
and the method by which they are to be used will be described
later in this chapter.

Ayres prepared a list of words consisting of 10 words for each

' For a detailed discussion, see the orisinal monographs or tlie writer's Educational
Measurements, or Monroe's Educational Tests and Measurements.



grade so selected on the basis of experiments that on the average
70% of the words for any given grade would be spelled cor-
rectly by the pupils of that grade. Later, Ayres ('15) prepared
a very useful test consisting of the 1,000 most common words.
These words arc split up into 26 lists of varying length and so
arranged that the words in any one list are of approximately equal
difficulty and that the successive lists from i to 26 become harder
and harder. The scale gives the average percentage of the words in
the various lists that pupils can spell correctly in the different

Buckingham ('13) prepared, on the basis of experiments, a list
of 50 words carefully scaled in difficulty according to the percentage
of pupils who could spell the words correctly.

(2) Uses and Results. All these scales have been found useful
in measuring efficiency in spelling in different schools more ac-
curately than was formerly possible, in determining individual
differences in abilities among pupils, in ascertaining progress, and
in comparing the effects of various factors in the learning and
teaching of spelling. It is hoped that they will be still more useful
in the future in discovering the most effective methods for ac-
quiring proficiency in spelling.

The facts with regard to the enormous range of individual abili-
ties in spelling and the consequent overlapping of the abilities of
the pupils in the various grades are shown in Figure 18, Chapter
III. The facts, as in other subjects, are astounding. The best
pupil in the first grade spells as well as the poorest pupil in the
eighth grade. Certainly some adjustment of the pupils should be
made according to their capacities.

Spelling presents one of the more striking examples of mental
sex differences found in educational psychology. Investigators
uniformly report girls doing better than boys. Wallin found the
girls averaging nearly 2% better than the boys in terms of his
spelling lists. In a study previously quoted in Chapter XIV, Foster
found that 238 girls and 256 boys, all university freshmen who
were given a speUing test of 40 difficult words of Latin derivation,
made respectively 76.6% and 68% successes. Sackett gave 24
words of Buckingham's spelling list to over 7,000 school children
and found the girls about a half year of school progress ahead of
the boys. Sears reports that a test composed of 70 words from
the Ayres list given to nearly 13,000 children in Oakland, Cali-
fornia, showed the girls superior to the boys from 2 to 6%, esoe-


cially in the upper grades. He suggests that the average girl
might be graduated from a half to a whole year earlier than the
average boy.

The writer found the differences in one school as shown in Fig-
ure 71. The difference is in fav©r of the girls, particularly in the
upper grades.







I 40
-2 30



^^- ^^











Fig. 71.^ — Sex differences in spelling as measured by the author's test.

Economic Methods in the Learning and Teaching of Spelling

(i) The Words to be Learned. For a number of years the most
important problem in the economy of learning to spell has been
the question, What words should a child really learn to spell? The
words in the spelling-books have for decades been selected mainly
by the arm-chair method and have consisted to a large extent of
-rare and useless words. Swinton's Speller, published in 1872, states
in the preface that "It ornitsTHe alphabet and the 'abab's' on the
one hand, and on the other, quite a number of sesquipedalian
words common to all old-time spelling books." It further urges
as a vantage point, "The practical character of the work which
aims to set forth, not the tens of thousands of long-tailed words in
osily and aiion, but the actual vocabulary used in speaking and
writing." And yet the book contains in one lesson (page 143)
such words as, lethean, pharisee, pentagon, pneumatics, theoc-
racy, anathema, dysentery, etc. In another lesson (page 144) it
contains oleaginous, farinaceous, argillaceous, lachrymose, sacer-
dotal, animadversion.



In view of this situation, recent years have brought forth a
number of very extensive studies and tabulations to find the words
which are most commonly used in writing by various classes of
persons and to discover the frequency with which these words
occur. The following are the chief tabulations thus made:

The Eldridge List. Mr. Eldridge ('11), a business man in Buf-
falo, New York, made a tabulation of 43,989 running words from
four different newspapers in which he found 6,002 different words.

2,927 words occurred each, once


' S16




three times

four times

five times

six times

seven times

eight times

nine times

ten to nineteen times

twenty or more times

The Ayres List. Ayres ('13) of the Russell Sage Foundation
tabulated 23,629 words from 2,000 letters, chiefly business letters,
and found 2,001 different words.

The Jones List. Professor Jones ('13) of the University of South
Dakota tabulated 15 million running words from 75,000 themes
written by 1,050 pupils in grades two to eight and found 4,532
different words.

The Cook and O'Shea List. Cook ('14) tabulated 200,000
running words from the family correspondence of thirteen persons
and found 5,200 different words.

These four tabulations represent four distinct fields of writing,
each being the most extensive in its field, namely, journalistic,
business, juvenile and private domestic vocabulary. One im-
portant type of vocabulary has never been tabulated, namely, the
vocabulary of our best current literary writers. Children ought
not to be confined to the words which they naturally use (Jones
List), nor to adult business vocabulary (Ayres List), nor to news-
paper vocabulary (Eldridge List), nor to the vocabulary of ordi-
nary family correspondence (Cook List). An important point in
learning to spell is to learn the meaning of words, especially of
words whose use will enhance a person's vocabulary. Hence, the


writer made a tabulation (Starch List) of the vocabulary of the
best current literary authors. This tabulation is unpublished and
on file in the Library of the University of Wisconsin.

The Starch List. The writer ^ tabulated some 40,000 running
words, about 1,000 from each of forty authors in eleven current high
grade magazines. This yielded 5,903 different words as follows:

3,111 words occurred each, once

1,009 " " " twice

512 " " " three times

2S0 " " " four times

189 " " " five times

121 " " " six times

97 " " " seven times

82 " " " eight times

53 " " " nine times

225 " " " ten to nineteen times

224 " " " twenty or more times

From these five lists, words for spelling and testing purposes
were selected according to the following plan: Every word occur-
ring three or more times in the Starch List, every word occurring
three or more times in the Eldridge List, every word occurring
seven or more times in the Cook List, and every word in the Ayres
1,000 word list was selected if it also occurred in one other list
including the Jones List. A word occurring three or more times in
the Starch List or in the Eldridge List or seven or more times in
the Cook List or any word occurring in the Ayres List was not
included if it occurred only in the one list. To be included it had
to occur at least once in one other list. This safeguarded against
the inclusion of words confined to one type of vocabulary only.
For example, the word "cupfuls" occurred twenty-one times in
the Starch List but in no other list. Hence, it was excluded.

The reason for selecting words that occurred three or more
times in the Eldridge List or in the Starch List was that the words
found less frequently are so rare that they constitute a very small
part of the running words of ordinary writing. This point may be
shown most emphatically by the accompanying graph, Figure 72,
on which the relative number of words of different frequencies is
indicated. A remarkably close parallel exists between the Eldridge
and the Starch Lists. The particular point to note in the graphs

1 In cooperation with L. C. De Bruin. Reported in a thesis in the library of the Uni-
versity of Wisconsin, 1916.



is the fact that the sharp bend in both curves occurs between
words whose frequency is between two and three. After three,
the curve shoots up very rapidly. Tliis same breaking point occurs
in the Cook List between seven and eight. It is higher in this list
because Cook tabulated a larger amount of writing. Words
occurring three or more times in the Starch and Eldridge Lists
constitute over nine-tenths of all running words.

This process of selection yielded 2,626 words. This number may

seem small compared with the number of words in former spelling





'E 2100


^ 1800

fe 1500


j| 1200



£0 10 9 8 7 G 5 4 3 2 1

Number of Times Words Occurred
Fig. 72. — Number of words occurring with various frequencies in the El-
dridge and Starch Hsts.

books and even in some contemporary spellers. But it is obvious
that it is not only useless but wasteful of a pupil's time to learn
words which he will never use in writing in all his years after school
and at the same time neglect to master thoroughly the words
which he will actually use. Spelling texts commonly contain from
ten to fifteen thousand words. In fact, the number of words here
presented, however, is even larger than that found in the special
spelling lists prepared by many cities which often do not have
more than from 1,500 to 2,000 words.

It seems, however, that a spelling list ought to include all words
of reasonably common occurrence but not words of extremely




Starch L








uncommon occurrence. This number seems to be approximately
2,500 or 2,600. Another reason for hmiting the spelling list to this
number is the fact that the writing vocabulary of the average
adult is not over 2,500 words and probably less for a great many
people. Cook found that three out of his thirteen persons who had
each furnished 40,000 running words (the largest number obtained
from any one individual) of family correspondence had used a
vocabulary of 2,575, 1,546, and 2,330 different words respectively.
This would mean that a person could write 100 letters each 400
words in length, making a total of 40,000 running words, and not
use more than 2,500 different words. If we may assume that the
average man or woman, exclusive of persons whose occupation
involves considerable writing, such as novelists, teachers, and
journalists, writes one such letter a month, his entire correspondence
for ten years would not involve more than 2,500 different words.

Incidentally we may also point out that this number furnishes
enough words to supply two new words for each school day in
grades two to eight, or 360 a year, counting iSo days to the school
year, and 106 words for grade one. /This is in accord with the
practice of many school systems of teaching not more than two
new words per day.

It is possible that such a list as that described above may need
to be supplemented to meet the needs of certain sections of the
population. Houser tabulated the words used by farmers in
California in corresponding with the state department of agricul-
ture. He found certain radical differences between this list thus
obtained and that published by Ayres.

The Placement of the Words into the Various Grades. Can this
be done on any scientific basis? Which of these 2,626 words
should a pupil learn in each grade? There are three possible ways
in which the words might be distributed into various grades:

(i) We might distribute the words according to their frequencies,
putting the most frequent words in the lower grades and the less
frequent words in the upper grades.

(2) We might put each word into the grade in which children
first begin to use it rather frequently in their writing.

(3) We might put each word into the grade in which, according
to the consensus of competent judges, such as teachers, it ought
to be taught.

The placement of the words in the present list was made partly
according to all three principles, but where discrepancies existed



with reference to any given word final decision was made according
to the third plan.

The 2,^26 words were first arranged according to their frequencies
of occurrence and all those words in the list which were also in the
Jones List were then placed into the grades in which the Jones
List shows them to be first used by children. A considerable num-
ber of the 2,626 words were not found in the Jones List. These
words were placed into higher and higher grades according as they
occurred less and less frequently.

After this task was completed the entire list of words was re-
checked according to the third plan. Fifteen different lists of
words which had been prepared by various cities or school systems
for their own uses such as the Boston Minimum List, the Stockton
List, the Santa Cruz List, the Chicago Speller, etc., were used.
After each of the 2,626 words occurring in one or more of the lists,
was written the number of the grade into which the word was
placed by each list. An average of these placements was then
obtained and accordingly the word was finally placed into its
grade. For example, the word "flower" was placed by seven
lists into different grades as follows: 4, 4, 2, 3, 2, 2, 4. This gives
an average grade placement of 3.00. " Cough" was assigned by six
lists to grades 5, 5, 3, 3, 2, and 4, with an average grade placement
of 3.66. All of the 2,626 words were thus assigned with the excep-
tion of 126 words which did not occur in any of the lists employed
and 1 78 words which were found in only one list. In order to make
the grade placement of these 304 words with equal confidence, a
group of seven experienced elementary teachers or supervisors
were asked to assign each of these words to some one grade accord-
ing to their best judgment. An average of these judgments was
then obtained and the words were placed accordingly.

All of the 2,626 words were then assigned to the various grades
according to the average grade placement as follows:

All Wcirds WnosE
Average Grade Placements were

In Grade One From i . 00 to 2 . 00

" Two

" Three

" Four

" Five

" Six

" Seven. . .

" Eight



2.76 '

' 3-28


' 3-66

3-67 '

' 4-71

4.72 '

' 5-66

5-67 '

' 6.70


' 8.00


This process gave 360 words in each grade or two for each school
day with the exception of grade one into which the remaining 106
easiest words fell.

Measuring the Attainment of the Pupils by Means of These Words.
The problem that was next attacked was this: How many of the
words of a given grade may we reasonably expect the pupils at
the end of the year to be able to spell? In pursuit of an answer
to this question, six lists of 6© words were selected from each
grade list by taking, for a given list, every sixth word through
the entire 360 words of a given grade. The words of the first grade
were split up into two lists of 53 words each. These lists were then
given as a special test at the end of the school year to approximately
7,000 pupils in 28 schools in 15 cities ranging in size from very small
towns to a city as large as Seattle. The percentage of words of each
grade spelled correctly by the pupils of that grade was as follows:

Grade 12 3 4 5 6 7 8

Av. percentage of words

spelled correctly 56.2 63.6 77.0 80.4 S3. 9 S5.0 82.9 80.7

Thus a very important advantage of these spelling lists ^ is that a
school or teacher can at any time test the pupils and determine
their efficiency by comparing them with the above standard aver-
ages. This can be done by selecting a list of 60 words from the
360 words for a given grade and giving them as a test. For ex-
ample, if at any time the teacher of the fourth grade desires to
compare the achievement of her pupils with the standard averages
of other fourth grades in schools generally, all she needs to do is to
turn to the words for the fourth grade and begin with any one of
the first six words and then pick out every sixth word through
the list. This will give a total of 60 words. At another time a
different list of 60 words may be chosen in like manner. The
same procedure may be follow^ed in any other grade by using
the words for that respective grade.

The important advantage of this plan is the fact that the same
words which have been used as a study list may be used at any
moment as a test list and comparisons may thus be made with
the standard averages.

The averages here presented indicate that the words for all
grades above the second are of approximately equal difficulty for
the respective grades since they all are within a few points of So%.

1 PublLshcd in the author's Spelling Book, 1919, by Silver, Bunlett &: Co.



That the words in the second grade are not too difficult is shown

Online LibraryDaniel StarchEducational psychology → online text (page 28 of 41)