Daniel Starch.

Educational psychology online

. (page 23 of 41)
Online LibraryDaniel StarchEducational psychology → online text (page 23 of 41)
Font size
QR-code for this ebook

subjects, failed almost completely. Gray, attempting by the same
methods to enlarge the range of visual impressions of two sixth-
•grade children for printed words, also failed completely. He re-
peated the experiment, however, with two fourth graders and
succeeded in approximately doubling it. He concludes that such


training, to be efifective, should take place not later than the
fourth grade.

(4) The movements of the eyes in reading. Closely related to
the scope of apprehension is the rapidity and the precision of the
movement of the eyes. More experimental work has been done
on these two elements in reading than on any others. It has been
known for a long time that the eyes in reading or in examining
any object do not move along smoothly and continuously but
that they move by jerks and pauses apparently in a very irregular
manner. The eyes take rapid glimpses or snap shots of successive
portions of a line of print and then piece them together in obtaining
the meaning. The reason for it is that the eyes cannot see objects
distinctly while they are in motion and consequently they per-
ceive little or nothing during that time. A person cannot see the
movement of his own eyes in a mirror. As soon as one can see them
distinctly, they are at rest again. Dearborn found that letters
fuse when passed in front of the eyes at the rate at which they
themselves move across a page. Dodge ('00) also found by experi-
ments that, while the eyes were in motion, no sensation resulted
even though the stimulus strikes the eyes.

A fairly good conception of the nature of eye movements, pauses,
and fixations in reading may be obtained by taking an open book,
placing a mirror on one page while some one reads the opposite
page, and then observing the action of the eyes as reflected in the
mirror. Javal ('79) counted the pauses of the eyes by means of a
sound attachment to the eyelids. Laudolt ('91) counted them
by direct observation. Erdmann and Dodge ('08) counted them
by observing the eyes of their subjects in a mirror. They found
more pauses with difficult than with easy reading material, and
also more pauses in reading a foreign language than in reading
one's native language.

The first successful attempt to record eye movements was made
by Huey ('98-'oo). He attached a plaster of Paris cup to the
cornea which was connected with a light aluminum pointer. This,
in turn, rested on the smoked drum of a kymograph on which the
movements were registered. In this manner he attempted to
study the nature and rapidity of eye movements and the nature
and length of successive pauses. Dodge then developed a falhng
plate camera which photographed a beam of light reflected from
the cornea of the eye. This method was also used by Dearborn.
It has the important advantage over Huey's method in that it


eliminates the attachment of anything to the eyes themselves,
which quite likely interferes with the natural movements of the

Readings by five subjects of the same newspaper passage

Sub- Number of Total

JKCTs Fixations Time Average

312 152 152 361

T ST. PE(TE]RSBUR|G, No[v.) 2.— Th|e Admiralty. . . 4 1007 251

490 III) 140 260

H ST. PE[TERSBURG, [N)ov. 2.— I— TheAd[m)iralty.. 4 1020 255

416 30a 136 136 310

S [ST. ]PETER|S(B]URG,Nov|. 2.— (Th]e Admiralty., s 1334 266

670 1 1 16 6tG

F ST. P|TERSBUR[G, Nov). 2.— The A[dm)iralty 3 2432 810

811 148 140 171 351 lia 273

M (S[T. IPE]TER|SBUR[G, )No[v.) 2.— [Th)e Ad|miralty 7 2057 293

T has telegraphe[d to the offi|cers of the |Baltic 3 493 164

330 220 160 210

H ha[s telegra)phed to th|e o(fficers] of the [Bal)tic 4 950 237

314 114 300 23S 233 263 130

S [) |has telegrap[he) (d] to th[e of)ficers| of th|e Baltic. . 7 1639 234

2a 1 179 S3 2 6S7 529

F |has telegrlaphe]d to the [o)fficers of the B|altic s 196S 393

21s I'Jo 231 226 171 179 7<

M |ha|s te'egraphed [to) the of|fice|rs of the lBa[lt)ic. ... 7 1301 185

Figure 58. The vertical lines and brackets indicate the locations of the eye
fixations. The numbers give the length of the fi.xations in loooth of a sec-
ond. After Dearborn ('06).

Figure 58 shows the locations of successive pauses, their lengths
in 1 000th of a second, and the distances between them as deter-
mined by Dearborn. The brackets and vertical lines indicate the
locations of these fixations. Sometimes there is a slight movement
after the fixation. The direction and distance of this movement
is indicated by the parentheses. These regressive or corrective
movements occur when the eyes have moved too far to the right.
They also occur when the eyes make the sweep back to the begin-
ning of the next line but have not gone far enough. These correc-
tive movements come more often in long lines than in short ones.
They are probably caused by the fact that the peripheral percep-
tion of the beginning of the line is not accurate. The exact location
from the beginning and from the end of the line is gotten as a
habit after several lines have been read if the lines are of uni-
form length. For this reason the lines should be of the same


length, especially for beginners. The opposite condition, however,
is often found in reading texts in which each sentence is printed
as a paragraph and in which pictures are inserted in the margins.
The latter is still more objectionable when the pictures are colored,
since the reflex action of colors influences the position of the fixa-
tions. The numbers above the vertical lines, Figure 58, indicate
in loooth of a second, the length of each pause. It will be noticed
that these vary considerably from time to time and that on the
average they are approximately one-fifth of a second in length.

Schmidt found on the average nearly twice as many refixation
movements in oral as in silent reading. These movements grow
progressively fewer and shorter from year to year as children
progress through the grades. Gray found frequent refixati,Qii_
movements very characteristic of slow readers. They were also
much more numerous with all readers in extremely careful read-

Dearborn found that, with newspaper lines, five persons made
an average of 4.76 fixations per line, that the number of fixations
tends to decrease and with it the range of apprehension tends to
increase as a passage is read over and over again, and that the eyes
finally develop considerable precision and accuracy in the mmiber
and the position of the fixations. He also found evidence for the
belief that the eyes form a short-lived motor habit which deter-
mines the number and position of the fixations for each line in
reading. The fermation of this habit develops during the reading
of the first three or four lines of print and seems to come much
sooner for some persons than for others. The fast readers seem
to form the motor habit more cjuickly than the slow ones. The
decrease in the time is due to a shortening of the pause and a
widening of the scope of attention, especially in the latter parts
of the line.

The fixations may come at any point in the line, between words,
at the beginning, at the end or in the middle of a word. "The
first and last fixations generally fall within the edges of the line,
that is, a little distance from the beginning or the end words."
The first fixation in a line is usually longer than the other fixations.

There are fewer pauses in the re-reading of a given material,
even after one month, and the number decreases somewhat as the
selection is read repeatedly. Prepositional phrases, connectives,
and substantives make the greatest demand upon p erfioption due
to the fact that a slight change in them affects the whole meaning


of a sentence. Dashes, punctuation marks, and capitals in the
middle of a line change the points of fixation, and disturb some
readers more than others. Reading a foreign language or reading
aloud requires longer fixation than reading one's native tongue or
reading silently. Dearborn obtained records of three boys aged
9, 10, and II years, in the third, fourth, and fifth grades respec-
tively, and found that the oldest one approached very closely to
the conditions of fixation of adults, while the younger ones made
many more fixations or pauses. Fatigue resulting from the use of
the eyes during an entire day decreases the rate of reading by
about one-tenth.

Schmidt ('17) made an investigation of the eye movements, in
oral and silent reading, of eighty individuals including elementary,
high school, and university students. He found that they made
1.6 more pauses per line in oral than in silent reading, that the
average duration per pause was from 20 to 27% longer, and that
the perception time was from 44 to 64% longer.

By an ingenious combination of camera and phonograph. Gray
determined accurately the relation of the eye to the voice in oral
reading. He found that the eye always precedes the voice, with
some subjects as much as four words and with others much less.
As a general rule a wide eye- voice span was associated with fluent

(5) Transmission of visual nerve impulses to the visual area of
the brain. Little can be said about this process. All that we know
about it is that the velocity of nerve impulses varies in different
persons, and may be slow or rapid in slow or rapid readers, but we
do not know definitely. Experimental work is necessary to de-
termine to what extent it may be true.

(6) The arousal of association processes. A word gradually
acquires meaning in the life of a child by its recurrence in numerous
situations and by the connection of the particular significance or
exi^erience with the word. Thus, the child hears the word " chair"
in connection with a certain object on which he sits. He hears it
also after a while in connection with other objects which look
different from the one he is accustomed to using. But new associ-
ative bonds are formed and he knows in general what is meant
when he hears the word "chair." Later on he is shown a certain
combination of visual characters and is told the word "chair."
A new bond is then formed between the sound and the visual
stimulus of the word "chair." An important part of the act of


reading consists, therefore, in the arousal of associative bonds as
soon as the visual nerve impulses are brought in from the eye to
the visual brain areas. The quickness of reading depends, no
doubt, upon the rapidity with which these incoming stimuli arouse
associated bonds and thereby give meaning to the word.

That the rate of reading depends upon the rapidity with which
the visual stimuli are interpreted is shown by such experiments as
Huey made in which he showed that it takes about twice as long
to read nonsense as sense material. He also found that words in
context give fewer associations than words out of context. When
the words were shown in context the associations would sometimes
go ahead of the amount seen and would anticipate what was com-

Hamilton ('07) attempted to determine the part played by
context in reading by ascertaining the time required to perceive
isolated words and words in various relations, such as in para-
graphs, in miscellaneous sentences, and in miscellaneous phrases.
His results, expressed in terms of seconds per word required for
the reading, are as follows:

TABLE 80. After Hamilton

Paragraphs Sentences Phrases Words

Mean 429 .456 .466 .66»

M. V 047 .047 .041 .078

Per cent using paragraph

as base 100% 94% 92% 65%

Isolated words apparently require a distinctly longer time for
perception than words in context.

Hamilton also found that the first part of a word is more im-
~^ortant in giving the meaning of a word than the latter part is,
that the marginal impression at the right aids in the perception of
the words seen in the next eye fixation, and that the upper
and lower marginal impressions probably do not aid in per-
ception. He also thinks it probable that the interpretation
takes place during the rest period, that is, during the movement
of the eye.

The central neural and mental activity of putting meaning into
the incoming impressions is probably the most important step in the
whole reading process. The important part ^f reading really is the
reading of meaning into words. Judd, McAllister, and Steele be-
lieve that the essential factors controlling the rate of reading are



central rather than peripheral and that it is a matter of assimilation
rather than a matter of getting material into the brain. Ruediger
has a similar opinion:

"In reading, a similar reinstatement of experience takes place as in
thought or in oral communication. The printed symbol arouses the
meaning that has through education and experience become connected
with it. It is to the rapidity with which this meaning is aroused that we
have to look for the cause of the differences in reading rate.

"Reading rate may then be taken to depend chiefly upon the rapidity
with which meaning is aroused in the mind after the symbol is seen.
This, in turn, is in the main dependent upon the person's native brain

(7), (8), and (9) Transmission of nerve impulses from the visual
center to the motor speech centers and thence to the speech organs.
The motor speech centers concerned in the control of the speech
organs are highly specialized. In right-handed persons, they are
located in the left hemisphere of the brain in the region of the fis-
sure of Rolando. These processes obviously will be involved in
oral reading and in speech, but they are also active in silent reading
in the form of incipient speech movements, particularly in the
tongue and, to a less extent, in other speech parts. They are
active in the same manner as in speaking, only on a much smaller
scale. To what extent this inner speech is an important part in
the reading process is somewhat uncertain. It is believed by
some investigators to be an important agency for maintaining
the continuity of the thought aroused by the successive visual
glimpses of the printed line.

Disturbances in the motor speech centers are known under the
names of various types of aphasia which will not be considered here.
But even in normal persons the rapidity and facility with which
the neural centers act may have an effect upon the efficiency of

Much valuable information could undoubtedly be obtained
from a careful laboratory study of individuals with language de-
fects, such as persons who have difficulty in learning to read or
difficulty in the speech functions proper. The following case may
be of interest in this connection. A boy, 17 years of age, in the
second year of high school, had always had considerable difficulty
in reading. His reading was extremely slow. He was finally sent
to a private teacher of public speaking with the hope that his


reading ability might be improved. After several months of work
with this teacher, apparently no progress was noticeable. He was
then brought to the psychological laboratory for examination to
see, if possible, wherein his difficulty lay. As careful tests as possi-
ble were made to discover in which of the various steps in the read-
ing process the difficulty might be. These successive processes
were tested with such means as were available or could be devised,
beginning with the first, the reception of the stimuli upon the
retina. Inquiry showed that his eyes had been carefully examined
and found to be only very slightly defective. Glasses had been
supplied as his father was a physician. Tests were next made to
determine the range of distinct vision by observing how large an
area he could see distinctly at one time. Next, his span of atten-
tion in apprehending visual stimuli was measured. This was
done by rapid exposures with a tachistoscope to determine the
number of words or pictures of objects he could apprehend at one
time. In the next place, tests were made to determine the con-
trol of his speech organs. It was thought that possibly his diffi-
culty in reading might be in an inability to control rapidly his
speech organs. Tests were made by having him repeat statements
from memory as rapidly as he could and also by having him repeat
short sentences from dictation. All these tests indicated that
these various functions were normal. His eyes had only slight
visual defects, his field of distinct vision was apparently of normal
size, his span of attention likewise was as large as that of a normal
individual, and the control of his speech organs was rapid and
accurate. The inference by a process of elimination was that the
chief difficulty in his reading ability lay in the central assimilation
or association processes. It seemed that visual impressions were
brought into the brain centers with normal speed and facility,
but that there was, for some reason, extreme slowness in the
mental interpretation of these stimuli. A test in silent reading
and in oral reading showed him to be equally slow in both cases.
His rate of reading was approximately that of a child at the end
of the first grade, namely, 1.5 words per second. His compre-
hension of what he read was good. His general intelligence, as
shown by other tests, was normal for his age. His work in other
school subjects was satisfactory. It seemed, therefore, probable,
although not absolutely certain in the absence of further tests, that
his difficulty lay in the central interpretation processes. This case,
similar to others, shows in an interesting manner the great intricacy




of the reading functions, and the difficulty in determining pre-
cisely what may be the trouble in a child who is backward in a
given school subject but normal or even superior in all others.

In order to ascertain more fully the part played by the various
factors enumerated at the beginning of this chapter, a series of
tests was carried out by A. D. Mueller under the direction of the
author,^ with 36 high school pupils. Speed and comprehension of
reading were carefully measured by three different selections
according to the writer's method. Then each of the capacities
mentioned in Table 81 was measured, and the coefficients of cor-
relation computed. These show that the two elements upon
which rate and comprehension of reading probably depend most
are the visual apprehension span for related and unrelated words
and quickness of association, correlation Nos. 2, 3, 6 and 12. Ap-
parently the amount apprehended and the quickness of giving
meaning to the visual symbols are the most important factors in
reading. Correlation No. 6 corroborates Nos. 2 and 3 since the
number of eye fixations per line of print is probably inversely
proportional to the amount apprehended at each fixation.


Correlations between reading ability and various elements entering into reading







I. Visual attention span — letters




2. " " " — unrelated

words. . . .




3. " " " • — rclatedwords




4. Auditory attention s[)an




5. Rapidity of voluntary eye move-





6. No. of eye fixations in reading ....




7. Association with auditory stimuli . .




8. " " visual " . .




9. Quickness of articulation — alphabet




10. " " " — rhyme. .




II. " " " — dictation

■ 33



12. Continuous association




Reported in an unpublished thesis in the hbrary of the University of Wisconsin, 1918.


These results are supported completely by a very similar ex-
periment of Gray's, although he computed no correlations. As
in the author's experiment there was a close relation between
reading ability and the range of visual apprehension for meaningful
material but, "The difference between the span of attention of
the good and poor readers disappears in a very large measure when
nonsense syllables, digits, groups of the same digit or the aussage
test are given." ^

The Measurement of Efficiency in Reading

(i) Essential Elements to be Measured. In order to be able
to determine the effectiveness of different methods and procedures
in learning and teaching a given school subject, it is necessary
to be able to measure with some degree of precision and objectivity
the general ability in that function. In order to measure such
abilities it is necessary, furthermore, to determine what the essen-
tial elements in the process are which ought to be measured. These
would seem to be in the case of reading as follows:

(a) In silent reading, speed and comprehension should be
measured. The prime purpose of reading is the comprehension
of the thoughts presented upon the printed page. The second
element is necessarily the speed at which the thought processes
may be obtained.

(b) In oral reading the chief additional element to be measured,
besides speed and comprehension, is the correctness of the pro-

(2) Method of Measurement. Several types of tests have
been developed for measuring efficiency in the various aspects of
reading. These methods will simply be mentioned here without a
detailed, critical discussion of their relative values or of their
various techniques of administration and evaluation. Tests for
speed and comprehension of silent reading, have been devised by
Gray, Kelly, Starch ('15), Courtis, Brown, Fordyce, and others^
The writer's test consists of a series of eight passages, one for each
grade. It is administered by having the pupils read for 30 seconds
and by having them write out as full an account as possible of the

^ Gray draws the following conclusion which is at least worth consideration: — "This
seems to prove that the inherent diflerences in mental capacity which exist between
members of this group are not the causes of diflerences in the span of attention. It
appears rather that differences in training — that is, acquired abilities to deal with mean-
ings — are the source of differences in perceptual span,"


portion read. Speed is expressed in terms of the number of words
read per second; comprehension is expressed in terms of the number
of words written after those words have been discarded which
represent incorrect statements of thought, additions of ideas not
found in the original passage, or repetitions of ideas previously
stated. The tests by Gray, Courtis, Brown, and Fordyce measure
rate and comprehension of reading substantially in the same manner
except that in some of them comprehension is determined either
partly or entirely by answers to questions, and in the fact that
most of them employ fewer than eight test passages. For ex-
ample, Gray uses three, Fordyce two, and Courtis one passage for
all grades.

Thorndike's ('14) tests are designed to measure comprehension
in reading, either of isolated words or of paragraphs. They are
constructed on the scale principle. Similar tests have been pre-
pared by Haggerty ('17). The Kansas silent reading test (Kelly)
is designed to measure comprehension primarily and speed second-

Gray ('17) has devised a test for oral reading consisting of a
series of twelve graded passages which are used to determine the
rate of oral reading and the number of errors in the pronuncia-

(3) Uses and Results of Measurements, (a) The first obvious
use of exact measurements of reading ability is a determination
of the actual abilities of the individual pupils, classes, or schools.
The first logical step in the management of any activity is a diagno-
sis of the conditions as they exist. Then, upon the basis of this
diagnosis, it is possible to prescribe more intelligently what should
be done. In order to be able to make definite comparisons of the
abilities of the pupils, it is necessary to know the standard norms,
or averages, of achievement in the various grades. For the writer's
test these are as follows:




2 3

Speed of reading

(words per second) . . .

(words written)

• 1-5

i.S 2.1
20 24

2.4 2.8 3.2 3.6 4.0

28 33 38 45 50

^ Further discussion of these tests may he found in the original sources in which these
tests were reported, or in the writer's Educalional Measurements, or in Monroe's Edu-

Online LibraryDaniel StarchEducational psychology → online text (page 23 of 41)