Copyright
Daniel Starch.

Educational psychology online

. (page 21 of 41)
Online LibraryDaniel StarchEducational psychology → online text (page 21 of 41)
Font size
QR-code for this ebook


non-Latin pupils in their second year of English, as 'impediment,'
'advocate,' 'reference,' 'anticipate,' 'subside.'

"In the third measurement, the difference in the averages of the two
groups — 69.5% and Zi-z'^o — was so great that Miss Humphrey thought
that perhaps too difficult words had been selected, or at least words which
placed the non-Latin students at an unreasonable disadvantage. Cu-
riously enough, in this measurement the words were taken, not from the
notebooks of a Latin pupU, as in the first two tests, in which the difference
between the two groups was much less, but, as stated above, from The
Tale of Two Cities. Furthermore, in No. 3, the non-Latin pupils were
so far afield in giving accurate definitions, and so confused in classifying
the words as to parts of speech, that it was decided to give another test
in which they should be asked, not to define words, but to give their
meanings, with the parts of speech omitted entirely. The results in this
measurement — 57% and 27.5% — -were virtually the same as in No. 3.

" Since practically every second-year pupil could write at least passably
on such a subject as 'What do I like to do best' it was decided to make
the basis of comparison in No. 5, not the average of the two groups, but
the percentage of rating above the passing mark. Moreover, in this
vocabulary test, emphasis was laid, not merely upon words of Latin
origin, but upon any words out of the ordinary, from whatever source.
The wide difference in the results from the view-point of excellence in
vocabulary — 36.0% and 6.8% — shows clearly what I have always be-
lieved and maintained, namely, that the work in commercial Latin
necessarily gives the pupils the dictionary habit, the results of which
extend far beyond the Latin derivatives actually studied.

"Of all the measurements, No. 6, was perhaps the most convincing.
In this test, the Latin pupils, unlike those in Nos. 1-5, had had during
the last six months of the two years' course the benefit of drUl in a vocab-
ulary not in the commonest use and yet valuable and even necessary to
educated people. The list of words was taken entirely from Frankhn's
Autobiography and Silas Marner which the pupils had just read, and
was not of unusual difficulty, consisting of such words, for example, as
asperity, promiscuous, mortuary. Yet by referring to the results it wiU
be seen that to the non-Latin group of pupils such words were practically
meaningless.

"An examination of the marks on these tests may prove of interest.
Among the seventeen non-Latin students the highest grade was 30%, and
five zeros were recorded. In the Latin group, on the other hand, the
lowest mark was 30%, while one pupil received 100%, two 90%, two 80%,
five 70%, and only three had below 50%. The difference in averages of
the two groups was 53%." (Perkins, '14, pp. 11-14.)

This investigation is interesting and one of the few whose results
were carefully worked out to make a precise comparison after elimi-



TRANSFERENCE OF TRAINING 243

nating differences in original capacity. It is a question, however, to
what extent the Latin students may have been favored by the
manner in which the words for the various tests were selected. For
tests I and 2 they were taken from the lists compiled and studied
by the Latin pupils. Even when the words are selected from Eng-
lish sources such as Silas Marncr and Franklin's Autobiography
there is still the question as to the particular words chosen for the
test. It is obviously unfair to select words which are relatively rare
and whose meaning may readily be inferred from their origin. To
what extent the words were selected fairly cannot be judged since
Perkins does not give the lists of words used.

The writer ('17) undertook a study to determine as precisely
as possible, the relative shares contributed by language training
and by original ability toward proficiency in English composition.
A series of tests was carried out with a group of 177 university
students. These tests together with their findings are given in
Table 74.

No. 3 consisted in WTiting an extemporaneous composition within
a limited time. These compositions were rated by three judges by
the Hillegas Scale.

No. 4 gives the average number of words written by each group
of students.

No. 5 gives the average number of different words used in each
composition.

No. 6 was a test in speed of reading. The numbers refer to the
words read per second.

No. 7 gives the number of words written in reproducing the
thought of the passage read in No. 6.

No. 8 gives the number of A's canceled in one minute in the well-
known A-test.

No. 9 gives the scores made in canceling in one minute a certain
geometrical form on a page of similar forms.

No. 10 consisted in reading to the persons a series of ten words
to see how many they could recall immediately afterwards.

No. 1 1 was carried out by giving a stimulus word and having the
persons write as many associated words as they could in thirty
seconds.

No. 12 consisted in giving a series of ten words and allowing
fifteen seconds to each word for writing as many synonyms as
possible.

No. 13 was a set of tests in imaging geometrical forms.



244



EDUCATIONAL PSYCHOLOGY



No. 14 consisted in presenting to the subjects seven words, one
at a time, spelled orally backwards by the experimenter. The
persons wrote down the words which they could thus recognize.



TABLE 74. After Starch ('17)



Percentage
9-15 Group
1-2 Group



1. Years of foreign

language

2. Number of persons .

3. Composition (Hille-
gas scale)

4. Words written

5. Different words used

6. Reading speed

7. Reading comprehen-

sion

8. Perception A-test. .

9. Perception form . . .

10. Memory words

11. Association free. . . .

12. Association

synonyms

13. Imagery forms

14. Imagery words. . . .

15. Years of English. . .

16. Grades, first year of

high school



14

67.

140.

82.

4-

60.

66.
7-
7-

23-

15-
7-

5-
5-

S3.



3-4
53



85.7



5-6
49

68.7

162. 1

96.6

5-5

70-5

66.2

7.8

7-3
22.6

iS-9
7.0

5-3

83.7



7-8
40

71

168

96

5

65
67

7

7
29

15
7
5
5



86.7



9-15
21



181. 4

III .6

6.0

68.2

66.1

8.0

7.2

28.7

14.2

7.8
6.1
5-5

88.0



15
28

35
33

13

o

6

— 2

26

-7



No

Latin

59



67
158



64
66

7

7
25

15

7
5
5

84-5



Latin
112



J!



A general inspection of the table shows that there is a steady in-
crease in the scores of practically every test from left to right with
increasing years of language study. Thus the 9-1 5 year group wrote
considerably better and longer compositions than the 1-2 year
group. The purpose of tests Nos. 6-14 was to ascertain to what
extent this superiority was due to original superiority of ability
or to the effect of language training. Tests 8-14 were selected par-
ticularly because the capacities to do them would probably be
affected very slightly if at all by language training. These show
on the whole a distinct superiority of inherent ability on the part
of the groups who studied languages for longer periods of years.

In order to make a crucial comparison as to how much of the
greater composition ability was due to the greater original capacity
of the pupils and how much was due to their greater training in



TRANSFERENCE OF TRAINING 245

language, the grades received by these students in all the subjects
carried during the first year of the high school were obtained from
the entrance records of the University. The amounts of difference
in original ability of the groups who later pursued varying amounts
of language work would be definitely indicated by this method,
since at that time none had had more than one year of foreign
language. The difference in the scholastic grades at the end of the
first year of the high school between those who later pursued lan-
guages for a total of 9-15 years and those who pursued languages
for a total of 1-2 years could certainly not be due to language train-
ing.

Row 16 gives for the different groups the average scholarship
grades during the first year of the high school. It will be noted that
there is a steady increase from group to group. The 9-15 year
group had an average grade of 88.0, or five points higher than the
1-2 year group.

.yThe next problem was to compare in common terms the five
points of difference in scholarship on the percentage scale with the
difference of 10.6 in quality of composition as measured by the Hille-
gas scalei To reduce these two types of measurements to commen-
surate units, fifty-eight compositions were rated by four persons
both by the percentage method and by the Hillegas scale. By a
process of equating values it was found that i.o point on the per-
centage scale equals 2.1 on the Hillegas scale. The difference of
five points, percentage scale, in original scholarship between the
1-2 year group and the 9-15 year gi'oup would be 10.5 in terms of
the Hillegas scale. The surprising result seems to be that the differ-
ence of 10.6, Hillegas scale, in quality of composition between
these two groups is approximately equaled by 10.5, the difference
in original scholarship when expressed in terms of the Hillegas
scale. The conclusion seems, therefore, unavoidable that the
difference in ability in English composition is due practically en-
tirely to a difference in original ability and only to a slight or no
extent to the training in foreign languages. [For the method of this
computation the reader is referred to the original report of this in-
vestigation. ('17.)]

The increase in length of composition and in speed of reading is
large and very probably in excess of the difference in original ability.
Training in foreign language seems to have produced a distinct
effect in greater fluency of words in writing and in more rapid per-
ception of words in reading.



246



EDUCATIONAL PSYCHOLOGY



Miss M. Theresa Dallam ('17), a teacher of English in the
Western High School of Baltimore, made a series of experiments
on her pupils to test the truth of her conviction "that Latin classes
do much better work in English than the classes that have not
studied Latin." Out of 114 students she selected 34 fourth-year
students, 17 Latin and 17 no-Latin or modern language pupils,
by pairing them on the basis of their general scholarship records,
so that the two groups would be equal in mental ability. The Latin
group had an average of 78 and the no-Latin group 79. Kelley's
Silent Reading test was also made with them. The Latin group
made an average score of 31.1 and the no-Latin group 33.1. The
two groups were, therefore, very nearly equal in general ability.
The Latin group had studied the language for four years.

Miss Dallam then made seven tests: spelling, reproduction, dic-
tation, Latin derivations, definitions, compositions, and English
grammar. The results were as follows:



TABLE 75. After Dallam ('17)
Average percentages attained



Sp.



DiCT.



Deriv.



Defin.



Com p. Gram, i j



Mod. Lan. Group.
Latin Group



89.2
90.7



63.0
659



96.0
950



293
52.0



73-0

75-2



71.8
75-4



S9-f
67.6



Thus the Latin group made a distinctly higher average in deriva-
tions and grammar, an appreciably higher grade in compositions;
a slightly higher average in spelling, reproduction, and definitions;
and a slightly lower average in dictation.

Miss Dallam then computed coeflftcients of correlation for the
Latin group between their grades in Latin and each of the seven
tests, and for the modern language group between their grades in
modern languages and each of the seven tests. These correlations
were as follows:



TABLE 76. After Dallam {^n)
Coefficients of correlation



Mod. Lan. Group.
Latin Group



Sp.



+ .09

+ .OS



Repro.



+ .19

+ .15



DiCT.



.04
.16



Deriv. Defin.



— .02

+ •13



+ •23

+ .15



CoMP. Gram.



+ .11

+ .28



+ .28
+ .46



TRANSFERENCE OF TRAINING 247

These correlations are so low that, with the exception of the ones
for grammar, no significance can be attached to them except to say
that there is practically no correlation between the various com-
parisons made and that the Latin group shows no superiority over
the modern language group in spelling, reproduction, definitions,
and a doubtful superiority in dictation, derivations, compositions,
and grammar. The differences that are shown are non-committal
and so small that they would have to be substantiated with other
groups to be conclusive.

(/. Science. Claims of general training to be derived from the
pursuit of the sciences are practically as extensive and confident
as those made for the languages. The only difference is that the
claims for the sciences have not been questioned or investigated
as much as those made for the languages. Bagley has summarized
them in the following manner :

"i. The formation of some useful specific habits, — through training,
routine, rationalized practice.

"2. The acquisition of useful information, — through methodical
study, instruction, and drill.

"3. The adoption of valuable ideals, or 'emotionalized standards,' —
inculcated through the inspiration to be gained from the teacher, from
the lives of great scientists, and from experiences of intimate contact
with nature.

"4. The acquisition of facility in the use of facts, ideas, and methodical
thought processes, for the solution of problems, the overcoming of
difficulties, and the accomplishment of worthy purposes, — through the
mental discipline afforded by properly graded practice in the solving of
scientific problems.

"5. The development of taste, and power of appreciation, — to be
gained through a clear apprehension of unity, adaptation, economy,
order, and system in nature as interpreted by science.

"6. The development of scientific or philosophic insights, perspectives,
and attitudes of mind that serve as safeguards to the intelligent inter-
pretation of contemporary life, — through acquaintance with systems of
organized knowledge." (Bagley, '11.)

One of the important values claimed for the sciences is the general
training of accuracy and fidelity in observation and the transfer-
ence of this particular type of training to other types of observation.
Miss Hewins ('16) made an attempt to measure the extent to which
this improvement is general or carries over to other types of ob-
servation. She divided each of three classes in botany, composed
of 34 boys and 50 girls, in the first year of a New York high school,



248



EDUCATIONAL PSYCHOLOGY



into two groups, and gave them a series of tests in various kinds of
observation of a biological and non-biological nature as follows:



Series i



No.


Date


Test Series


Exposure


I


\pri]


22


Horse chestnut stem




2




23


Picture


30 seconds


3




23


Forsythia flower




4




24


10 syllables


30 "


5




24


Lilac leaves




6




25


Nonsense figure


30 "


7




25


Geometrical figure


30 "


8




26


ID 2-place figures


3° "


9




29


Scouring rush


I minute


lO




29


Maple seedling


I minute


II




30


Pea chart


30 seconds


12




30


Figure in air




13


May


I


Potato chart


30 seconds



Time for
Recording

10 minutes



1 May 15

2 16

3 17



5


21


6


22


7


23


8


24


9


27





28



I June



3
4
S
6

7
8

9
10



13



Practice series
Material
Description of lilac flower

" " box-elder leaf

" " the stem, leaf, and flower

of gill-run-over-the-ground
Description of flower stalk and flowers

of the lily-of-the-valley
Description of the horse-chestnut flower
" " " buttercup flower

" " " mustard flower

" " " dogwood flower

" " " deutzia flower

'• " " columbine flower



Series 2



Test Series



Sassafras stem
Picture

Syringa flower
10 syllables
Forsythia leaves
Nonsense figure
Geometrical figures
10 2-place figures
Moss plant
Pumpkin seed chart
Grape chart
Figure in air
Wild carrot



30 seconds

30 "

30 "
30 "

30 ;

I minute

I minute

30 seconds

3° "



Recording



10 minutes
for each
test



Time tor
Recording



TRANSFERENCE OF TRAINING



HO



The practice series was continued for a period of ten days with
one section of each of the three classes. The other section of each
class answered questions from books on the material of the lessons.
At the end of the period, she gave the tests in Series 2 to both
sections of each class and obtained the following gains in scores
over Series i:



TABLE 77. After Hewins ('16)
Gross gains in scores



Biological Tests



Non-biological Tests



Practiced groups:

Boys

Girls



Average

Unpracticed groups:

Boys

Girls



Average

Residual difference between practiced
and unpracticed groups

Percentage gains of residual differ-
ences over scores in first test series



8


06


6


41


7


23


3


03


— I


24




S9


6


34


33


9%



8,97
6.20


7.58

5-37
5.60


5 48

2. ID

5-4%



(Median score in biological tests in Series i before practice 1S.7)
( " " " non- " " " " " " " 39.0)

Miss Hewins infers a considerable gain in both the biological and
non-biological types of observation.

"Table 77 shows that in the biological tests, the average gain of each
practised boy was S.06 per test for the 5 tests while the unpractised
showed a gain of 3.03. The practised girls averaged 6.41 gain per test,
while the unpractised lost 1.24 per test. In the non-biological, the prac-
tised girls gained 6.2 per test for the 6 tests while the unpractised gained
5.6; the practised boys gained 8.97 per test and the unpractised boys
gained 5.27 per test."

"Feeling that the balance of arguments and scientific proofs were
against formal discipline when this investigation was begun, I am forced
by the results obtained to admit that in this experiment the proof seems
to be on the affirmative side.

"A valuable lesson, I think, can be drawn from one phase of this
investigation. By consulting the tables and summaries, it will be seen



250 EDUCATIONAL PSYCHOLOGY



that sometimes one division does not fall in line with the general trend
but that a larger number outweighs the negative and shows positive re-
sults." (Hewins, pp. iii and 113.)

In order to make a relative comparison of these gains on the
basis of the original scores in the first series of end tests, I have
computed the percentages of gain in these end tests as shown at the
bottom of the above table. When such a relative comparison is
made, the net gain in the non-biological observations is very small,
being only 5.4% as compared with 33.9% in the biological obser-
vations, or about one-sixth as much in the non-biological as in
the biological types of observation. The improvement in the
biological end tests can hardly be counted as evidence of spread
of training since these tests were so similar to the training series
that they were all but identical with it. The spread of training is
apparently not as great as Miss Hewins believes it to be.

e. Geometry. Geometry, particularly in the high school, is urged
to possess a considerable amount of disciplinary value:

Plato (Republic, Book 7) emphasized this opinion thus:

" Moreover, the science (Geometry) has indirect effects which are not
smaU.

" ' Of what kind? ' he said.

" ' There are the military advantages of which you spoke,' I said;
and in all departments of knowledge, as experience proves, any one who
has studied geometry is infinitely quicker of apprehension than one who
has not.

" * Yes, indeed,' he said, ' there is an infinite difference between them."

Rugg ('16) made a study of the spread of training in the case of
learning descriptive geometry. He performed three groups of
tests at the beginning and at the end of the course in descriptive
geometry, in February and in June respectively, with 326 students
in the College of Engineering of the University of Illinois. The
same tests were made with 78 students in other colleges as a control
group who did not pursue the course in descriptive geometry.
The three groups of tests were of a non-geometrical, Tests i and 2,
quasi-geometrical, Test 3, and strictly geometrical nature, Tests
4 and 5. Partial illustrations of the various tests follow:



1

^d, I!



Test I.



Divide eighty-one by seven.
Divide seventy-eight by four, etc.



TRANSFERENCE OF TRAINING



251



Test 2.

Test 3.
Test 4.



Test 5.



1. Divide eight sixty-two by three.

2. Divide seven ninety-five by four, etc.

A test in imaging letters.

The Painted Cube Test.

A three-inch cube, painted on all sides, is cut into one-inch cubes.
How many one-inch cubes have paint on three sides?

a « u u u u j^Q u p

3- " " " " " " " " one side?
4. " " " " " " " " no side?



Geometrical Objects' Test.

The problem: Form a mental picture of each object and count the

number of straight lines which it would take to construct each

one in space.

1. A wedge.

2. Four triangles attached to a square, bases coinciding with the

sides of the square, etc.

The main results are represented in the following table:

TABLE 78. Representing the results for " Rights." Adapted from Rugg's
table, p. 123 ('16)

Test 1 Test 2 | Test 3 Test 4 Test 5

Train- Con- Train- Con- Train- Con- Train- Con- Train- Con-
ing TROL INC TROL ' ING TROL ING TROL ING TROL

Group Group Group Group! Group Group Group Group Group Group

Feb.scores 20.0 17.38 20.00 19.3 ,4.67 4.50 0.59 0.28 2.32 1.92

Junescores 22.81 20.06 23.00 19.15 5.66 4.58 1.50 0.68 i.ii 1.81

Gross gain 2.81 2.68 3.00 -.15/ .99 0.083 .91 0.40 l.Ol -0.10

%gain 14.5 15.6 15.0 -.78 121.2 1.8 157.0 143.0 43.5 -5.0

Residual 1

diSerence -1.10 15.78 19.4 14.0 48.5

Averages 7.34 19.4 31.7

Rugg concludes that the residual gain indicates a considerable
transfer of improvement to all three types of abilities tested.

"The study of descriptive geometry (under ordinary class room condi-
tions throughout a semester of 1 5 weeks) in which such natural and not
undue consideration is given to practice in geometrical visualization as is
necessary for the solution of descriptive geometry problems operates:

" (i) Substantially to increase the students' ability in solving problems
requiring the mental manipulation of a geometrical nature, the content
of which is distinctly different from the visual content of descriptive
geometry itself.

" (2) Substantially to increase the students' ability in solving problems



252 EDUCATIONAL PSYCHOLOGY

requiring the mental manipulation of spatial elements of a slightly-
geometrical character, i. e., problems utilizing the fundamental elements
of geometry (the point, line, and plane), but apart from a geometrical
setting, and in such form as to offer no geometrical aids in solution.

" (3) Substantially to increase the students' abihty in solving problems
requiring the mental manipulation of spatial elements of a completely
non-geometrical nature, i. e., problems in which the straight line and
plane do not appear in any way whatever.

" (4) The training effect of such study in descriptive geometry operates
more efficiently in those prol^lems whose visual content more closely
resembles that of the training course itself, i. e., in those problems whose
imagery content is composed of combination of points, lines, and planes,
and in which the continuity of the manipulating movements approaches
the continuity of those in the training course. (Rugg, pp. 114-115.)

These results are not very different from those surveyed in the
preceding chapter. Rugg states the results perhaps as favorably as
the data permit, perhaps too favorably. At any rate we ought to
note that the gain in the non-geometrical tests, Nos. i and 2, is
only about one-fourth or one-fifth as much as in the strictly geomet-
rical tests, Nos. 4 and 5. We also should note, as Rugg himself
points out, that only about two-thirds of the persons in the train-
ing group gained; the remaining one-third did not gain or show
transfer.

General Interpretation. The transference of training of the capa-
cities involved in the learning of school material is very small so far
as present partial data indicate. This seems to be equally true
of all school subjects, the sciences as well as the languages and
mathematics. If we represent the possible transfer effect as ranging
from 0%, or none, to 100%, or an improvement in other capacities



Online LibraryDaniel StarchEducational psychology → online text (page 21 of 41)