University of Southern California Dissertations and Theses

Reversal And Nonreversal Shift Performance Of Retardates Under Various Motivational And Stimulus Conditions

(USC Thesis Other)

Reversal And Nonreversal Shift Performance Of Retardates Under Various Motivational And Stimulus Conditions

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has b een
m icrofilm ed exactly as received
66-11,577
KRAM ER, R ichard A lb ert, 1925-
REVERSAL AND NONREVERSAL SHIFT PERFOR
MANCE OF RETARDATES UNDER VARIOUS
MOTIVATIONAL AND STIMULUS CONDITIONS.
U n iv ersity of Southern C aliforn ia, P h .D ., 1966
E ducation, psychology
U niversity M icrofilm s. Inc., A nn A rbor, M ichigan
REVERSAL AND NONREVERSAL SHIFT PERFORMANCE OF
RETARDATES UNDER VARIOUS MOTIVATIONAL
AND STIMULUS CONDITIONS
by
Richard Albert Kramer
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Education)
June 1966
UNIVERSITY O F SO U T H E R N CALIFORNIA
T H E G R A D U A TE S C H O O L
U N IV E R S IT Y PA R K
LO S A N G E L E S , C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
RICHARD ALBERT KRAMER
under the direction of hX.9....Dissertation C om
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R O F P H I L O S O P H Y
Dean
D ate 1.?66
DISSERTATION COMMITTEE
Chairma.
TABLE OP CONTENTS
Page
LIST OP TABLES...................................... iii
LIST OP ILLUSTRATIONS............................. iv
Chapter
I. PROBLEM.................................... 1
Introduction
Purpose of the Study
Scope and Limitations
Outline of Following Chapters
II. LITERATURE................................. 9
Review of Relevant Literature
III. PROCEDURE................................. 44-
Purpose
Subjects
Apparatus
Method
Design
IV. RESULTS................................... 55
First Discrimination Trials
Second Discrimination Trials
V. SUMMARY AND CONCLUSIONS................. 72
Summary
Discussion of Results and Conclusions
BIBLIOGRAPHY........................................ 87
APPENDIX...................................... 95
LIST OF TABLES
Table Page
1. Means, Standard Deviations, and Ranges
CA, MA, and I Q ............................... 4?
2• Means and Standard Deviations of Raw Scores—
Trials to Criterion in First Discrimination
and Second Discrimination Series........... 56
5* Ranges Under Various Transformations— Trials
to Criterion in First Discrimination and
Second Discrimination Series ............... 53
4* Logarithmic Transformation of Means and Standard
Deviations— Trials to Criterion in First Dis
crimination and Second Discrimination
Series....................................... 59
5* Analysis of Variance— Trials to Criterion in
First Discrimination Series ............... 61
6* Analysis of Variance— Trials to Criterion in
Second Discrimination Series ............... 65
7« Means and T-Tests— Logarithmic Transformation
Trials to Criterion in Second Discrimina
tion S e r i e s ................................. 64
8. Means and T-Tests— Logarithmic Transformation
Trials to Criterion in Second Discrimina
tion Series • . ............................ 64
9* Analysis of Covariance— Logarithmic Transfor
mation ....................................... 66
10* Adjusted Means and T-Tests— Logarithmic
Transformation Trials to Criterion in Second
Discrimination ............................... 68
11* Adjusted Means and T-Tests— Logarithmic
Transformation Trials to Criterion in Second
Discrimination............................... 68
' 1
LIST OF ILLUSTRATIONS
Figure Page
1. Schematic Representation of the Sequence of
Stimuli Presentations for a Typical Re
versal Shift G r o u p ........................ 13
2. Schematic Representation of the Sequence of
Stimuli Presentation for a Typical Non-
reversal Shift G r o u p ...................... 14
3. Schematic Representation of Single-unit and
Mediational Formulations in Typical
Reversal and Nonreversal Shifts ........... 16
4. Experimental Design........................... 53
iv
CHAPTER I
PROBLEM
Introduction. j
It lias recently been estimated (White House
Conference, I960) that five million people in the United
States are mentally retarded* Usually these unfortunate
I
people are relegated to living a marginal life in the
community or are removed from the community and placed in
institutions. If placed in institutions, only about a
quarter of those who have sufficient ability for inclusion
in classes for training or education are actually en
rolled in such classes* Although it is true that the
retarded have less ability than the average person, it
is also true that their condition partially prevents the
maximum utilization of the abilities which they do
possess* Teachers are apt to find these people frus
trating to deal with and consequently give them only
cursory attention. Teaching techniques used with re
tardates may simply be those used with normals of similar
mental age and may not take into consideration the special
needs of the retarded. The effect upon the retarded
individual is often that of discouraging effort rather
than encouraging it* To the extent that these conditions
exist the result is considerable waste in human resources,
added expense involved in caring for the retarded, as
well as an incalculable toll in human anguish and frus
tration*
Certain trends present in our society today do
not augur well for the future of the retarded. The trend !
toward automation appears irreversible. Its effect on
employment possibilities for the retarded can only be
detrimental. The very fact that the country's population
is increasing means that there will be inevitable short
ages of facilities with a consequent crowding of insti
tutions and classrooms for the retarded. A case in point
is the present plight of the moderately retarded, or
"trainable," child. It is only within the past ten years
that significant numbers of public school districts have
begun to provide instruction for this type of child.
Many school districts have no facilities for children of
this intellectual level. In this circumstance the child
is often institutionalized. His progress is then further
slowed by the lack of trained personnel and facilities in
the institution. Maintenance of state institutions for
the retarded is becoming an increasing burden for the
taxpayer. Efficiency in the care and training of the
retarded is imperative in order that the most benefit may
be derived from the available funds.
Procedures aimed at rehabilitating the retarded
have not been conspicuous for their success. An attitude
of hopelessness or resignation is often found in people
involved in the care or training of the retarded. Often
this attitude results in a slighting of the moderately
retarded in favor of concentration on the mildly retarded#
However, in terms of potential for improvement, the
moderately retarded child often surpasses the mildly re
tarded child.
Recently there has been a noticeable upswing in
research efforts in the field of mental retardation#
Whereas most of the past research in this field has been
of an applied nature, a newer trend is in the direction of
"pure” research. Although some workers in the field
appear opposed to this development (McPherson, 1958), the
majority seem to consider it a healthy sign. The history
of scientific discovery shows quite clearly that many
practical problems have been solved by applying the find- !
ings of "pure" research.
Another trend in current research is a shift in
the focus of attention from the mildly retarded, or
educable, child to the moderately retarded, or trainable,
child (Dunn and Capobianco, 1961). In 1956 it was re
ported that approximately 22,000 moderately retarded
children were being given training in day or residential
classes (Goldberg, 1957)* Until very recently studies
of learning and concept formation in the retarded have |
been relatively scarce and heterogeneous. Only fourteen
articles on the subject of learning in the retarded were
noted between the years of 1948 and 1958 (McPherson,
1958). A recent review of the literature on concept
formation in the retarded revealed a similar scarcity of
studies (Rosenberg, 1961). The tempo of research in
these areas has quickened somewhat since then, but much
information is still needed on these topics in order
that teachers of the retarded will be able to utilize
the most efficient techniques available.
The teaching-learning situation contains many
variables and many interactions between these variables.
Information is needed in such areas as: the develop
mental stages of the learning process in the retarded,
the most effective motivational techniques, the nature
of the learning process at various levels of mental age,
the type of materials most likely to facilitate learning,
and the nature of any interactions which may exist be
tween these variables. Hopefully this information can
be provided through research which aims at the basic
processes involved.
Purpose of the Study
One purpose of the study herein reported was to
explore some of the dimensions involved in the learning
process in moderately retarded individuals. More speci
fically, a reversal-nonreversal discrimination learning
task was employed in an effort to determine whether in
dividuals of this level of retardation use a direct-
association type of process or a mediational S-R type of
process in learning a simple discrimination. The vari
able of motivation wan explored by giving some subjects
abstract (verbal) reinforcement following their responses
and giving other subjects abstract (verbal) reinforcement
plus a concrete reward (candy). The dimension of con-
creteness-versus-abstractness of the stimulus objects was
explored by using drinking cups and cardboard squares for
the two aspects of this dimension. The use of a factor
ial design enabled a testing of the possible interactions
between these variables.
Another purpose of the study was to extend a
series of investigations of the developmental stages of
the learning process. Studies employing similar tech
niques have been done with rats, college students, and
kindergarten students. Evidence from these studies
indicates that more primitive organisms (rats) utilize a
direct association process in solving the problem, while
higher organisms (college students) use a mediational
process. Kindergarten students seem to he in a transi
tional stage between these two. In the present study an
effort was made to discover the type of process used by
the moderately retarded. Hopefully the retarded would be
found capable of using the mediational process. The
variables of motivation and nature of stimuli were ex
plored in an effort to find the conditions most condu
cive to the utilization of higher thought processes by
the retardates.
Scope and Limitations
For practical reasons of subject availability
institutionalized retardates were used. Undoubtedly
selective factors operate in the placement of a child in
a state hospital for the mentally retarded. To the
extent that this is true, the results of this study may
not necessarily be valid for non-institutionalized re
tardates.
Since male retardates are usually more disruptive
in the community (Farber, 1959) and are institutionalized
much more frequently than females, only males were used
in the study. Accordingly, the results may not be valid
for female retardates.
The chronological age range of the subjects was
from ten years through sixteen years. The findings may
not be applicable beyond these limits.
The mental age range of the subjects was from 3)6
years through 6)6 years, and consequently the findings do
not warrant generalization beyond these limits.
Strictly speaking, caution should be used in
interpreting the results beyond the type of learning
task and motivational techniques employed in the study.
Also, since the present study was done with individual
subjects in a quiet room relatively free from distrac
tions , the results of the study may not be applicable
directly to the typical classroom situation involving
group instruction and more distractions. However, it is
hoped that generalization beyond these particular tech
niques and situations will be found justifiable.
Outline of Following Chanters
A review of the relevant literature is presented
in Chapter II. At the beginning of this review is an
explanation of the reversal-nonreversal technique and the
rationals for its use. This section was included in
order to clarify the review of studies involving the use
of this technique. Following the review of the litera
ture is a section which summarizes the review and points
toward the need for the present study.
Chapter III contains a statement of the hypotheses
to be tested and a description of the procedures utilized
to test them.
Chapter IV presents the statistical analysis of
the data obtained from the study. The relationships
between the hypotheses and the results are then dis
cussed. Chapter V contains a brief restatement of the
entire study, including the need for the study, the pro
cedures utilized, and the results obtained. A final
section relates the findings of the study to educational
procedures•
CHAPTER XI
LITERATURE
Review of Relevant Literature
The review of the literature relevant to the
present'study involves consideration of several areas of
research* Since the study involves a procedure for
studying discrimination learning (reversal-nonreversal
shift) which is not widely known, a preliminary section
explaining this procedure is included as a basis for
understanding some of the studies later reviewed* This
section also included an explanation for the use of this
particular procedure. Following this introductory secticn
is a review of selected learning studies employing this
procedure in the exploration of areas related to the
present study* Since the subjects in the present study
are mentally retarded a section on the possible effeats
of etiological factors has been included* Since stimulus
characteristics are a crucial element in any discrimina
tion learning study, a review of the role of stimulus
characteristics follows. Since the present study was in
part concerned with manipulating the variable of motiva
tion, a section reviewing motivational procedures has
10
'been, provided. The next section presents various points
of view with regard to the dimension of abstract versus
j
concrete behavior, since this dimension has often been
used in describing the behavior of retardates. A final
section summarizes the research and points up the need
for further work in various areas. !
Briefly stated, then, the plan of the chapter is
the followingt
Description of the Reversal-Nonreversal Technique and
Rational for Its Use
Experimental and Theoretical Background of Present
Study
Etiological Factors
The Role of Stimulus Characteristics
Review of Typical Motivational Procedures
Abstract versus Concrete Behavior
Summary of Research
Description of the Reversal-
nonreversal Technique and
Rationale for Its use
In learning studies involving a reversal-nonrever-
sal shift the following basic type of design is used. A
group of subjects, designated the reversal group, is
given two successive discriminations to learn. In the
first of these discriminations one particular stimulus
dimension is correct, or relevant, and one aspect of this
11
dimension is correct. For example, the relevant dimen
sion might be "size" and the correct aspect of that
dimension might be "large."
When a subject has learned this discrimination
to some predetermined criterion, the opposite aspect,
"small," becomes the correct one and he is required to
learn this second discrimination to the same criterion.
For these subjects only one stimulus dimension is rele
vant during the entire experiment. The experimenter is
interested in the ease with which these subjects shift
their choices from one aspect to the other on this dimen
sion, as compared with the ease with which subjects in
nonreversal groups are able to shift from one relevant
dimension to another.
An example of this latter type of task (non
reversal shift) might involve having the subject learn to
respond to an aspect of the size dimension (large stimu
lus correct) during the first discrimination trials and
then having them learn to respond to an aspect of the
color dimension (black stimulus correct) during the sec
ond discrimination. In a typical experiment involving
the stimulus dimensions of color and size then, the sub
jects in a reversal group would learn one of four kinds
of shifts: black-to-white, white-to-black, large-to-
small, or small-to-large. Subjects in a nonreversal
12
group would leara one of eight kinds of shifts: black-
to-small, black-to-large, white-to-small, white-to-large,
large-to-black, large-to-white, small-to-black, or small-
to-white.
Buring each of the first discrimination trials
all subjects are presented with either the combination
of a large, black stimulus and a small, white stimulus
or the combination of a large, white stimulus and a
small, black stimulus. Thus during these trials all
subjects are presented with the task of discriminating
among three possibly relevant dimensions: size, color,
and position. Each of these dimensions has two possibly
relevant aspects (size, large or small; color, black or
white; and position, left or right). Position is always
an irrelevant cue in these experiments and is varied
randomly. Figures 1 and 2 illustrate schematically
typical reversal and nonreversal tasks.
The reversal-nonreversal type of learning task
is particularly useful in studying the nature of the
learning process in various types of subjects. It re
sults in a simple comparison which can be used to deter
mine whether the subjects utilized a single-unit or a
mediational learning process. The single-unit formula
tion (Spence, 1956) posits a direct association between
stimulus and response. The mediational formulation
PIRST DISCRIMINATION
Subjects presented with:
B
V
m m
+
£T
W
B
—
+
Size is the relevant dimension.
Large is the relevant aspect of
tnat dimension.
Pig. 1.— Schematic representation
tions for a typical reversal shift group.
SECOND DISCRIMINATION
Subjects presented with:
m
+
or
Size is still the relevant dimension.
Small is now the relevant aspect of
of that dimension.
of the sequence of stimuli presenta-
H
FIRST DISCRIMINATION
Subjects presented with:
SECOND DISCRIMINATION
Subjects presented with:
B
w V
or or
V
Size is the relevant dimension. Color is now the relevant dimension.
Large is the relevant aspect of Black is the relevant aspect of
that dimension. that dimension.
Fig. 2.— Schematic representation of the sequence of stimuli presenta
tion for a typical nonreversal shift group.
H
i
i
. ._]
(Hull, 1930; Dollard and Miller, 1930; Guthrie, 1952;
Kendler and D'Amato, 1955) posits a covert S-R sequence
which is interposed "between the subject's initial re
action to the stimulus and his eventual overt response.
Figure 3 shows a schematic representation of these two
formulations applied to a reversal-nonreversal shift
discrimination learning task.
If the subjects accomplish the reversal shift
more rapidly than the nonreversal shift, this is con
sidered evidence that they used a mediational type of
process in their learning. The line of reasoning here is
that if the subjects used a mediational process they would
have available the relevant dimension at the end of the
first discrimination trials. They would then simply have
to switch from one aspect to the other on this same
dimension in order to learn the second discrimination.
For example, if the subjects in the reversal groups
learned that "size" was the relevant dimension and "large”!
the relevant aspect of that dimension during the first
discrimination trials, then they would only have to
switch from one aspect to the other of that dimension for
the second discrimination. Their task during the second
discrimination would thus be simplified, and they would
require fewer trials to reach the criterion. Phrased
differently if, as the mediational formulation posits,
16
SINGLE-UNIT FORMULATION
Reversal Shift Nonreversal Shift
S S
MEDIATIONAL FORMULATION
Reversal Shift
?Black
? rcoIor~scolor
Nonreversal Shift
rcoIors color RWhi-'te
r_,----- Aarge
size size
% mall
Fig. 3— Schematic representation of single-unit
and mediational formulations in typical reversal and non
reversal shifts.
the subject's behavior during a trial involves two
successive associations, then the stimulus of the first
association would be the test discriminanda and the
response would be the covert symbolization of the dis
criminanda. The stimulus for the second association
would be the cue derived from the response of the first
association, and the response for the second association
would be the subject's overt choice of one of the dis
criminanda. In these circumstances the reversal shift
problem is easier than the nonreversal shift problem for
the subject using a mediational process, since at the
termination of the first discrimination the proper
implicit cue for the second discrimination is already
available to the subject. All the subject needs to do
is associate it with the opposite overt response to the
dis criminanda.
If the subjects are found to execute the rever
sal shift more slowly this is considered evidence that
they utilized a single-unit type of process. The reason
ing here is that at the termination of the first dis
crimination in the reversal shift problem the subjects
do not have available the correct implicit cue for the
second discrimination, and, consequently, their shifting
behavior will require more trials. Their task is also
made more difficult by virtue of the fact they they must
18
extinguish, the previously correct response "before they
set about discovering the new correct response.
A1though the present experimental design permits
an evaluation of whether the subjects used a single-unit
or a mediational type of process in learning the dis
crimination, it should be noted that these two formula
tions do not constitute a complete catalogue of theories
in this area, nor do they explain sill aspects of behavior
in a discrimination learning experiment. For example,
since the present experiment deals with two successive
discriminations, the work of Harlow (194-9) on learning
set could be considered relevant but could not be di
rectly evaluated from the data obtained. The work of
Zeaman and House (1963) with regard to the role of atten
tion in the discrimination learning of retardates is
undoubtedly relevant but cannot be evaluated from the
present data. The same comment applies to the work of
Lawrence (1952) on the relative importance of reinforce
ment and the effect of the eventual isolation of relevant
cues by the subject. Along this line also, the proba
bility theories of Estes and Burke (1955)* Bush and
Hosteller (1951)* ^<1 Restle (1955) should be mentioned
as having possible relevance. Finally, the work of Amsel
(1958) on the motivational effect of the frustration .
attendant upon making a wrong choice may also have
relevance but cannot be evaluated from this data
Experimental and Theoretical
Background of Present study
The present study was designed as an extension of
a series of studies of the nature of the learning process
utilizing the reversal-nonreversal technique. A review
of some of these studies and some theoretical issues
concerning them will now be presented. Three reversal-
nonreversal studies, (involving rats, children and
adults) represent significant points along a dimension
of the response of simple-to-complex organisms on this
particular learning task. These studies will be reviewed
first and then some ancillary studies will be covered.
Kelleher (1956) employed a reversal-nonreversal
shift procedure with white rats in an effort to discover
whether they used a single-unit S—R process or a media
tional S-B process in learning a discrimination. The
relevant stimuli were black versus white alleys and
presence or absence of chain curtains in the alleys. All
subjects were run Just prior to their daily feeding. The
incentive was food pellets. The criterion was 18 correct
out of 20 consecutive trials with a maximum of 500 trials
permitted. Partial reinforcement effects during the test
trials were controlled by presenting one of the irrele
vant cues in both alleys on each trial. Subjects in the
20
reversal conditions were significantly slower in making
the test discriminations than subjects in the nonreversal
condition. Kelleher interprets these findings as support-*
ing Spence's (1936) single-unit S-R theory of discrimina
tion learning. It might be noted here, however, that
Spence was quite careful to specify that his theory <
applied only to the discrimination learning of nonverbal
subjects. Extension of Kelleher's findings to articulate
organisms is, therefore, not warranted.
Kendler and Kendler (1939) did a developmental
study aimed at discovering the nature of reversal and
nonreversal shift discrimination learning in children.
As subjects they used 122 kindergarten children with
mental ages ranging from 4—10 through 6-6. As discrim-
inanda they utilized black and white metal tumblers of
two different heights. The subjects were given a candy
reward for successful performance. The objective of the
experiment was to see whether a reversal or nonreversal
shift was easier for the children. If the reversal
shift was found to be easier this would suggest that the
subjects were employing mediational thought processes in
the task. If the nonreversal shift was found to be
easier, then the suggestion would be that the subjects
were using a direct association (or single-unit S-R)
approach to the task. The results indicated no signifi—
21
cant difference between reversal and nonreversal groups
in speed of learning the test task* However, when the
data were reanalyzed according to groups of slow and fast
I
learners on the basis of performance on the training
j
trials, a significant difference was found between the
reversal and nonreversal groups in speed of learning the
test task* The results were in the direction predicted,
with the fast learners accomplishing the reversal shift
more readily and the slow learners accomplishing the
nonreversal shift more readily. The author's conclusion
is that the kindergarten children were in a transitional
stage with regard to the development of their thinking
processes. The slower children were still using the
single-unit S-R process, while the faster children had
already begun using the mediational S-R process.
Kendler and D*Amato (1955) used a card sorting
task for studying the effects of reversal and nonreversal
shift learning in college students* Their study, which
consisted of a series of three separate experiments, !
aimed at answering some methodological questions raised
by an earlier experiment by Buss (1953)* described later,
as well as adducing additional evidence that reversal
shifts are learned more easily by human subjects, thereby
supporting the mediational formulation as an explanation
of the discrimination learning process in adults.
22
The first experiment of the series utilized 112
introductory psychology students in a task employing two
stimulus cards and a set of response cards which were to
be sorted appropriately under the stimulus cards. The
concepts to be learned were shape and color in all three
experiments. The experiment was designed in such a way ;
as to prevent any fortuitous partial reinforcement in the ;
nonreversal groups during the learning of the second
concept. This was accomplished by removing cards from
the deck which would provide such reinforcement during
the initial stage of learning the second concept and then
reinserting these cards during the second stage of learn
ing the second concept. This technique provided 100 per
cent nonreinforcement in both the reversal and nonrever
sal groups for the responses that were correct for the
first concept during the first stage of the learning of
the second concept. In this fashion the experimenters
hoped to negate the differential effect of partial re
inforcement of the initially correct responses in the
nonreversal groups during the learning of the second
concept. The results were in agreement with the authors'
prediction in that the reversal shift was learned signi
ficantly faster than the nonreversal shift. Accordingly
the authors concluded that the superiority of the rever
sal shift condition was independent of the inhibiting
23
effects of partial reinforcement in the nonreversal con
dition.
The design of the second experiment of the series
was similar to that of the first, with the exception that
no stimulus cards were used, so that any possible inter
action effect between the stimulus cards and the response !
cards was eliminated. The absence of the stimulus cards
resulted in slower learning of the initial concept in
both groups. Again, the findings was that reversal j
learning of the second concept was significantly faster.
In the third experiment of the series the same
procedure was used with the exception that no cards were
removed from the deck during the initial phase of the
test trials. The findings again showed significantly
faster learning of the reversal task. In all three
experiments comparison of experimental and control groups
consistently showed negative transfer in nonreversal
groups and positive transfer in reversal groups. The
results were interpreted as consistent with the media
tional S-R theory.
Kendler, Kendler, and Wells (I960) explored the
developmental dimension further by using nursery school
children in a reversal-nonreversal shift experiment.
They also wished to test the effect of having the subjects
verbalize the correct and incorrect stimulus dimensions
2 4 :
prior to shifting. The ages of the subjects ranged from
33 months to 63 months. Different forms, colors, and !
heights of cookie cutters were used as discriminanda.
The experimenters predicted that subjects of this age
would slow negative transfer on the reversal shift task
j
and that verbalizing the relevant dimension would make
the reversal shift significantly easier. The results
indicated significantly slower shifting in the reversal
groups. Having the subjects verbalize the relevant
stimulus dimension prior to the test trials produced no
significant effect. The authors speculate that the
reversal-nonreversal shift technique might be used as a
basis for evaluating developmental levels in concept
formation. They also suggest that their findings support
Luria's (1959) contention that in children and the brain-
injured adult, verbalization may be parallel to overt
response processes with no interaction taking place.
Kendler and Kendler (1956) used 128 preschool
children aged 34 to 59 months to test a conclusion of
Maier*s (1936) that inferential behavior is rarely found
in children under 6 years of age. For their experiment
they devised an apparatus which consisted of goal objects
inside a small box. The goal objects could be manipulated,
by pulling various strings which protruded from the box.
Inferential behavior was indicated if the subject was
25
able bo utilize experiences from initial trials with, the
apparatus and correctly guess which string to pull to
produce the goal, under a new arrangement. The results
indicated that inferential behavior is possible in
children of this age range if the experimental arrange
ments are made sufficiently simple.
Kendler, Glucksberg, and Keston (1961) used a
reversal-nonreversal shift design to demonstrate per
ceptual and symbolic responses within the behavioral
chain involved in the learning of the discrimination
tasks. Introductory psychology students served as sub
jects. The stimuli were projected tachistoscopically
and the subjects indicated their choice by pressing
telegraph keys. In addition to accomplishing the shifts
necessitated by the basic design, some subjects were
required to make an additional shift in their perceptual
orienting response, since the position of the stimuli was
changed for the test trials. This latter condition pro
duced significantly slower learning. The results are
interpreted as supporting a mediational S-R formulation.
In an early study using the reversal-nonreversal
shift design Buss (1953) found that both types of shift
produced negative transfer and that the effect was
greater with nonreversal shift. His subjects were intro
ductory psychology students. However, his results are
26
j
questionable since a feature of his experimental design !
permitted partial reinforcement for subjects in the non
reversal groups during the test trials for responses j
which were correct during the training trials but in-
i
correct during the test trials.
In a card-sorting experiment Kendler and Mayzner
(1956) used two and four sorting categories during the
learning of the second discrimination instead of just two
categories as in a previous study (Kendler and D*Amato, j
1955). They predicted that this procedure would show
that the advantage of the reversal groups was partly due
to the use of two sorting categories. The subjects again
were introductory college students. The results revealed
a significant interaction between speed of shifting tinder
the two shift conditions and number of sorting categories.
The reversal shift took place significantly faster in the
two category situation but there was no significant
superiority for reversal groups in the four category
situation.
Harrow and Friedman (1958) compared reversal and
nonreversal shifts in a card sorting task with college
students as subjects. With partial reinforcement effects
eliminated by using different sets of cards for the two
tasks they found that the reversal discrimination was
still learned more rapidly than the nonreversal discrim-
27
ination. The authors concluded that the superiority of
the reversal shift is not due to partial reinforcement
effects in the nonreversal groups during the second die-
i
crimination. They concluded that a mediational S-B
formulation best accounted for their results. |
i
Plenderleigh (1956) compared feebleminded child- j
ren with normals on a discrimination reversal learning
task. She found no differences between the two groups
on errors in making the initial discrimination or in re
calling the discrimination after a six week interval*
However, the results indicated that the feebleminded
subjects had more difficulty learning the reversal dis
crimination following the six week interval since the
original learning*
Etiological Factors
In using mentally retarded subjects the factor of
etiology must always be considered. The two reference
points generally used on this dimension are organic
(brain-injured or exogenous) and familial (endogenous).
Some authorities consider all retardates as organic to
some degree (Spitz, 1965; Weatherwax and Benoit, 1957).
Other authorities prefer to classify the retarded as
organic only when there is reasonable evidence of neuro
logical damage (Werner, 1946; Strauss and Lehtinen, 194-7).
i
1
Other authorities prefer to classify the retarded as
- - - 1
28
organic only when there is reasonable evidence of neuro
logical damage (Werner, 1946; Strauss and Lehtlnen, 194-7)*
Some retardates show no signs of neurological damage and
therefore are considered familial types* A number of
studies have shown that, on a variety of tasks, there is
j
no significant difference between the two categories of i
organic and familial (Pascal, Stolurow, Zabarenko, and
Chambers, 1951* Barnett, Ellis, and Pryer, I960; Bensberg,
1955; Bunn and Capobianco, 1954; Capobianco and Miller,
1958; Capobianco and Funk, 1958)* Other studies indicate
that organics are inferior in the performance of various
types of tasks (Bensberg and Cantor, 1958; Halpin, 1955;
Barnett and Pryer, 1958).
If an organic has some obvious physical or sensory
handicap, his learning of a discrimination might well be
impaired. However, barring this circumstance, there
would appear to be no valid reason for considering the
organic as different from the familial with regard to the '
learning of this type of task. In the present study no j
distinction will be made on the basis of etiology.
The Role of Stimulus
Characteristics
Many stimulus variables have been used in dis
crimination experiments with retardates. Only those with
some relevance to the present experiment will be reviewed*!
An experiment by O’Connor and Hermelin (1959) !
indicates that retardates are capable of making simple
size discriminations. Stimuli used in the experiment
were black squares of two different sizes (ratio of 5*8)
presented on a red background. The subjects were a group
j
of retardates and a group of normals matched on MA* The
I
task was a simple discrimination followed by a reversal
of that discrimination. Candy served as the incentive.
Following the completion of the task the subjects were
asked to state the principle involved. Normals were
superior to retardates in ability to state the relevant
principle (size), although the retardates had accomplished
the reversal task in significantly fewer trials than the
normals. This later finding supports the thesis of
Luria (1959) that the retarded characteristically lack
coordination between their verbal and motor responses.
The effectiveness of color as a cue in a discrim
ination learning task has been explored by Zeaman and
House (1965)- They used five stimulus conditions: "junk"!
(three dimensional objects such as a plastic soap dish or
a tobacco can), color-form objects, form, color-form
patterns, and color. Form-color objects consisted of a
circle, a square, a triangle, a cross, and a "T" form.
These figures were cut from % inch Masonite, painted
various colors (red, green, yellow, white, and black),
30
and mounted on Masonite bases. They did not exceed two
inches in height or width. Using retardates (MA. four
years to six years) the authors found that the most
difficult discrimination was that in which only the color
of the stimulus object was varied (red square versus
green square). The easiest stimuli for the retardates
to discriminate were the "junk” stimuli. The effect of
varying the stimuli from three-dimensional color-forms
(green Masonite square) to two-dimensional representations
of these forms (color-form patterns) was investigated
also. It was found that this latter type of stimuli made
the discrimination much more difficult. The majority of
the subjects were unable to make this discrimination
although given 250 trials.
Another variable in a discrimination experiment
is the relative position of the stimuli. Usually this
involves only the possibilities of left or right. House
and Zeaman (1959) report that retardates with MAs from
two years to four years are generally able to learn this
position discrimination in fewer than three trials.
Typically the position variable is not one of the rele
vant variables in a discrimination experiment.
Another possible stimulus dimension is that of
concreteness versus abstractness. In the study by
Zeaman and House referred to previously, the "junk" dis-
criminanda and the two-dimensional color-form patterns
could be cited as examples of opposite points on this
I
dimension. It will be recalled that the "junk" (concrete)!
i
discriminanda gave rise to the fastest discrimination
learning. The authors speculate that the reasons for the
superiority of the concrete, three-dimensional object as
a stimulus may lie in the fact that it presents more
differential aspects to the subject than does a two-
dimensional stimulus. These results are in agreement
i
with an earlier finding by Heidbreder (194-7) that, in the .
development of concept formation ability, concepts in
volving concrete objects are the first to appear, while
abstract concepts of quantity and number are the last to
appear.
Review of Typical Motivational
Procedures
Various motivational procedures have been used
with the type of subject and task utilized in the present
experiment. An exhaustive review of this topic would be
beyond the needs of the present study. Therefore only a
few representative studies will be mentioned.
Considering motivational procedures employed in
studies involving a reversal-nonreversal shift type of
task, it is seen that when college students are used as
subjects no tangible reinforcement is given. Studies by
Kendler and Mayzner (1956), Buss (1955)« Buss (1956), and
Kendler and D'Amato (1955) serve to illustrate this
point. For this type of subject the need to achieve in
tasks with intellectual connotations is regarded as
sufficient motivation. With animal subjects food has
j
been used (Kelleher, 1956). With normal children as I
subjects reinforcement has taken the form of candy or
trinkets (Kendler, Kendler and Wells, I960; Kendler and
Kendler, 1959)* j
Consideration of various types of experiments
using mentally retarded individuals as subjects reveals
a wider variety of motivational techniques. Establishing
good rapport with the subjects was used in an experiment
by McCullough, Reswick, and Roy (1955) as a motivational
technique. The description of their technique is dis
tressingly vague, however. It would have been more
helpful if the authors could have specified the effective
factors operating in good rapport. Small charms of the
i
type found at novelty counters of variety stores were used!
by Plenderleith (1956). Cantor and Hottel (1955) used
two quantities of peanuts as reinforcement in an experi
ment with retardates. Some subjects were given one peanut
for each correct response, while other subjects received
four peanuts for each correct response. Contrary to the
authors* expectations no difference was found in the
35
subjects' performances as a result of these two levels
of rewards•
Heber (1959) was able to demonstrate that re
tardates have definite preferences with regard to tang
ible reinforcements. Furthermore, he was successful in
changing the level of their performance by manipulating
the desirability of their reward. The selection of
rewards available to the subjects in this study included
a pipe, a ring, peanuts, jelly beans, twenty-five cents
in coin, comic books, a pair of socks, gum, cigarettes,
a toy horn, a toy pistol, a chocolate bar, a billfold,
and an arm bracelet.
Ellis (1962) tried using one cigarette versus
three cigarettes as a motivational technique with re
tardates and found that no difference in performance
resulted. Considering the results of this study and the
Cantor and Hottel (1955) study against the background of
the Heber (1959) study one might consider the possibility
that cigarettes and peanuts may have rather low incentive
value for institutionalized retardates. Another possible
explanation might be that, while these rewards have a
slight motivating effect, this effect is minor compared
with some other motivational variable. Such a variable
might be the effect of personal attention from the exper
imenter, as will be referred to later in this review.
34-
Encouragement by the experimenter during the
performance of a formboard task was the motivational
technique used by Tizard and Loos (1954-) in an experiment
involving imbeciles. O'Connor and Tizard (1956) compared
the effectiveness of encouragement versus giving the sub
jects goals to attempt to beat. The subjects were re
tardates of the trainable level and the task was a leg-
perslstence test. The results indicated that giving the
subjects targets to beat provided stronger motivation
than giving them verbal encouragement.
Setting a goal for the subjects was tried also by
Gordon et. al. (1955)* This motivating condition was
found to be superior to conditions involving group compe
tition, competition with a single rival, and a control
condition involving no type of comparison. O'Connor and
Claridge (1955) tried setting a goal and giving encour
agement toward attainment of that goal versus simply
setting a goal for the subjects. They found that the
goal-plus-encouragement condition yielded significantly
superior performance to that obtained with the goal-only
condition.
Stevenson and Zigler (1957) have advanced a
motivational concept termed the "social deprivation hy
pothesis" to explain the differential effect of various
rewards with retardates and normals. Their hypothesis
contends, in part, that retarded subjects suffer from a
deprivation of social attention. As a result of this
deprivation they are particularly responsive to the
attention shown them by the experimenter in certain ex
perimental conditions. In fact, they contend that the
effect of this attention may well be more powerful than
that of such tangible rewards as candy, peanuts, cigar
ettes, or trinkets. This contention is supported by the
results of a number of studies (Zigler, Hodgden, and
Stevenson, 1958; Gerwirtz and Baer, 1958; Zigler, 1961;
Shepps and Zigler, 1962; Shallenberger and Zigler, 1961;
and Zigler, 1962).
Ellis and Distefano (1959) measured the perfor
mance of institutionalized retardates on a standard
laboratory device, the rotary-pursuit meter. One group
of subjects received praise from the experimenter while
the other received none. The performance of the praised
group was significantly superior to that of the control
group•
An interesting interaction effect was reported
by Stevenson and Knights (1962). They found that en
couragement from a male experimenter was more effective
with female subjects (average CA of 14.2 years) than with
male subjects. The subjects were institutionalized
retardates and the task was marble sorting.
" I
36 |
In an effort to separate the effects of institu- j
i
tionalization and retardation upon the subject's response
to praise, Stevenson and Fahel (1961) compared the per
formance of four groups of subjects: non-institution-
alized normals, institutionalized normals (orphange child-i
!
i
ren), non-institutionalized retardates, and institution- !
i
alized retardates. The task again involved marble sort
ing. Their results favored institutionalization as a
more powerful variable than retardation in production a
positive response to praise.
Ellis and Pryer (1958) sought to weigh the effects
of jelly beans (primary reinforcement) versus yellow paper
squares (secondary reinforcement) as motivational devices
with retardates. The paper squares proved more effective.
As Heber (1964) has pointed out, the procedures employed
raise a number of questions. For one thing, the reason
for designating the paper squares as secondary reinforce
ment is not made clear. Then there is the possibility
I
that the incentive value of jelly beans was lower than
that of the paper squares, regardless of whether the
jelly beans might be considered as a primary reward, and
thus more powerful, by the experimenter.
Abstract versus Concrete
fiefaavior
j
Behavior may be conceptualized along the dimension
of concrete versus abstract, or rigid versus flexible.
i
According to tbis formulation concrete or rigid behavior
is supposed to be more characteristic of young children, j
i
the brain injured, and the mentally retarded, while
normal adults are supposedly capable of both types of j
behavior.
Lewin (1936), Kounin (194-la, 194-lb), and Werner
(19^6, 1948) have described the behavior of the retarded
as rigid, while Goldstein and Scheerer (1941) preferred
i
the term concrete for this purpose. Rigidity, or con
creteness, implies a difficulty in shifting dimensions
and in being able to verbalize the relevant dimensions in
concept formation tasks. These formulations have not
been accepted without controversy, however. Zigler
(1962) has criticized Lewin for not making a distinction
between rigidity as a concept and as a behavior. Kounin
has been criticized by Goldstein (1943) and Werner (1946,
1948) for his supposed misconstruing of the concept of j
rigidity. Goldstein felt that the concept really con
sisted of primary and secondary aspects. In his view,
the primary type of rigidity is the result of subcortical
damage, while the secondary type is due to cortical
damage and is characteristic of the mentally retarded.
Heber (1964) views the controversy as based upon a failure
to grasp the difference between Kounin*s concept of j
38
rigidity and his description of rigid behaviors. Much of |
the disagreement between these authors appears to be
semantic in nature. There is no disagreement between
them with regard to the hypothesis that the behavior of
the mentally retarded tends to be more concrete or rigid
than that of normals. The results of a number of studies
have born out this contention.
Bolles (1937) compared retardates with normals on
some of the Goldstein-Scheerer tests and found the nor-
i
mals more capable of abstract behavior. Prothro (194-3)
compared normal children with adult retardates matched
with regard to MA.. The tasks involved sorting of beads
of different shapes, sorting pencils by color and sorting
pasteboard figures by color and form. The results re
vealed that the normal children were superior in making
these abstractions.
Stacey and Portnoy (1951) compared retardates of
two different IQ levels on the Goldstein-Scheerer Object !
Sorting Test. The average IQ in one group was 59 and in |
the other 74-. The higher IQ subjects named the basis for I
the various sortings significantly more frequently than
the low IQ group. A similar study by Korstvedt, Stacey,
and Reynolds (1954-) employing the Goldstein-Scheered
Color-Form Sorting Test and comparing normals and retard
ates revealed that the normals were superior in naming j
the basis for the sortings, Halpin (1958), using the
same task (C-FST) with retardates, found that ability to
shift from one sorting basis to another increased as CA
and MA increased. Siegel (1957) compared retardates with
normals and found the retardates less capable of shifting
dimensions and naming the various dimensions involved in
a visual-verbal concept formation test. Plenderleith
(1956) tested normals and retardates, matched on MA, on
ability to reverse a discrimination once they attained
it. The reversal task was presented at two different
time intervals following the original learning: twenty-
four hours and six weeks. Under these conditions no
significant difference in shifting ability was found
between the two groups.
Stevenson and Zigler (1957) also did a discrimin
ation reversal experiment with retardates and normals
matched on MA. They also found no significant differences
in shifting behavior. The results of these latter two
studies run contrary to the hypothesis of more concrete
behavior in the retarded. An attempt to resolve this
discrepancy of findings eventually gave rise to a number
of studies which eventuated in the positing of the "social
deprivation hypothesis" mentioned in the review of moti
vational procedures.
To summarize the research in this area, some of
Rosenberg's (1963) conclusions would seem appropriate*
After reviewing the research he states that the retarded
are less capable than normals of comparable MA in using
abstractions for making discriminations involving non
verbal stimuli* He also points to some areas in which
research is needed* One of these areas involves studies
of abstraction with subjects of moderate and severe
levels of retardation, since most studies to date have
involved only the mildly retarded* Another area he
mentions is that of exploring the environmental conditions
that are conducive to abstract thinking in the retarded*
Summary of Research
The reversal-nonreversal shift technique is
useful in studying the nature of the learning process.
It provides a simple means by which it can be determined
whether the subjects utilized a single-unit S-R process
or a mediational S-R process in learning a discrimination*
Using this technique it has been found that nor
mal adults use a mediational process (Kendler and D'Amato,
1935) while rats use a single-unit process (Kelleher,
1956). Kindergarten children apparently are at an age
when they ere shifting from the single-unit process to
the mediational process (Kendler and Kendler, 1959)*
Further exploration of this developmental dimension is
needed. In utilizing this technique, however, the factor
of fortuitous reinforcement during the test trials for
subjects in the nonreversal groups must be controlled.
Having the subjects verbalize the correct and
incorrect dimensions in discrimination studies has not
proved helpful at the nursery school level (Kendler,
i
Kendler, and Wells, I960). Inferential behavior has,
however, been demonstrated in children aged 34- to 59
months when the problem is made sufficiently simple
(Kendler and Kendler, 1956).
Evidence has been adduced that normal adults
utilize a chain of covert responses in solving discrim
ination problems (Kendler, Glucksberg, and Keston, 1961).
With partial reinforcement effects in the nonreversal
groups controlled, normal adults still find the reversal
shift easier (Kendler and Mayzner, 1956).
Various studies have shown no significant differ
ence between organics and familials on a variety of tasks
(Pascal, Stolurow, Zabarenko, and Chambers, 1951* BarnettJ
Ellis, and Pryer, I960; Bensberg, 1955; Bunn and Capo- j
bianco, 1954-).
Retardates have been found capable of making
discriminations of relative size (O'Connor and Hermelin,
1959)* Three dimensional ("junk") stimuli are relatively
easier for retardates to discriminate while color dis
criminations are relatively more difficult for retardates |
42
(Zeaman and House, 1963)* Position discriminations are
quite easy for retardates (House and Zeaman, 1959) • The
more concrete the stimuli, the easier is the discrimina
tion for retardates. The more abstract the stimuli, the
more difficult is the discrimination for retardates
(Heidbreder, 1946; Zeaman and House, 1963)* !
With college students as subjects no tangible
reward is necessary as a motivational procedure. Their
need to achieve is sufficient (Kendler and Mayzner, 1956; ■
Buss, 1953s.; Kendler and DrAmato, 1955)* Pood is the
typical reward with rats (Kelleher, 1956). Retardates
have definite preferences among various rewards (Heber,
1959)* Candy and trinkets have been used successfully
with normal children and retardates (Kendler and Wells,
I960; Kendler and Kendler, 1959; Plenderlieth, 1956).
Peanuts, jelly beans, and cigarettes have been found to
have insignificant motivational value with retardates
(Cantor and Hottel, 1955; Ellis and Pryer, 1958; Ellis,
1962). !
Encouragement by the experimenter has been found
to have a motivating effect (Tizard and Loos, 1954) as
has the technique of giving the subjects a goal to attempt
to beat (O'Connor and Tizard, 1956; Gordon, et. al., 1955)*
Stevenson and Zigler (1957) have found that
'
attention from the examiner is motivating for institu-
4-3
tionalized children. Praise was used effectively by
Ellis and Distefano (1959) with retardates. Stevenson
and Knights (1962) found that female retardates were
more strongly motivated by encouragement from a male
examiner than were male retardates.
The behavior of retardates has been found to be
more rigid or concrete than that of normals (Lewin, 1936;
Kounin, 1941a, 1941b; Werner, 1946, 1948; Goldstein and
Scheerer, 1941). Retardates have been found to be in
ferior to normals in naming the relevant dimension in
sorting tasks (Korstvedt, Stacey, and Reynolds, 1954).
Halpin (1958) found that ability to shift from one
sorting dimension to another increases as MA. increases.
However, Plenderleith (1956) and Stevenson and Zigler
(1957) found no significant difference in shifting be
havior between normals and retardates. This type of
finding led to the positing of a "social deprivation
hypothesis" to explain these findings. In brief, this
hypothesis states that institutionalized subjects will
be highly motivated by attention from the experimenter.
Prom this review of the literature it appears
evident that more information is needed in the areas of
the development of the learning process in the retarded,
the effect of stimulus characteristics, and the effect
of various motivational procedures•
CHAPTER III
PROCEDURE
i
j
Purpose
The present study constitutes a further explora-
i
tion of the developmental dimension of the concept forma-j
tion process utilizing a reversal-nonreversal discrimin
ation learning task# Various points along this dimension ;
have been located by other studies# Kendler and D"Amato
(1955) have demonstrated that college students respond to
this type of task with a mediational S-R type of process#
Kelleher (1956) has shown that a single-unit S-R formula
tion best explains the behavior of rats with this type of
experimental task. Kendler and Kendler (1959) located a
point between these two extremes in their study of kin
dergarten children, demonstrating that these subjects are s
i
in a transitional stage from one type of process to the i
other#
The present study constitutes an attempt to
determine the nature of the concept formation process in
mentally retarded subjects of younger mental age than the
subjects in the Kendler and Kendler study. The results
of the Kendler and Kendler study indicate that using
44
45
subjects of lower mental age should result in the finding
of a single-unit, rather than a mediational, process*
However, with proper stimulus and motivational condi
tions , it might be possible to demonstrate a mediational
process in the concept formation behavior of retardates
of lower mental age. In the present study concrete and
abstract reinforcements were used as motivational devices,
and concrete and abstract discriminanda were used as
stimuli. It was hypothesized that the subjects would be
better able to utilize a mediational process with con
crete, three dimensional stimuli than with the more
abstract, two dimensional stimuli. In addition, it was
hypothesized that the subjects would be better able to
utilize a mediational process with concrete reinforcement
in the form of candy than with abstract reinforcement in
the form of the statement "Eight" or "Wrong" by the ex
perimenter.
Summarizing the foregoing, then, the present
study attempts to test the following hypotheses:
I. The relative speed of learning in the rever
sal groups will vary significantly depending
upon the nature of the discriminada.
la. Reversal groups will learn faster than
reversal groups with concrete discrim
inanda.
46
lb* Nonreversal groups will learn faster
than reversal groups with abstract
discriminda*
II* The relative speed of learning in the rever
sal and nonreversal groups will vary signi
ficantly depending upon the nature of the
reinf orc ement•
2a. Heversal groups will learn faster than
nonreversal groups with concrete re
inforcement.
2b* Nonreversal groups will learn faster
than reversal groups with abstract
reinforcement.
Sub.jects
The subjects were 128 mentally retarded males
ranging in age from 8.25 years through 16.0 years (Table
1). The mean age of the subjects was 15*0 years. The
mental ages of the subjects ranged from 5*5 years through
6.5 years with a mean of 5*3 years. The IQs of the sub
jects ranged from 29 through 62 with a mean of 45*6.
Fifty-nine of the subjects were patients at Fair-
view State Hospital, Costa Mesa, California and sixty-
nine of the subjects were patients at Pacific State
Hospital, Pomona, California. Within the above limits of
CA, MA, and IQ, the subjects were chosen without regard
to diagnostic category. Subjects with, known gross de
fects in vision or hearing were excluded, as were subjects)
; I
1 with disabilities of the upper extramities of such nature|
as to preclude adequate participation in the experimental
procedures.
i
i
TABLE 1
Means, Standard Deviations, and Ranges
CA, MA, and IQ
Variable Mean Standard Deviation Ranges
CA 13*0 years
1.97
years
8.5
thru 16.0 years
MA 5*3 years 0.81 years
3.5
thru 6.5 years
IQ
45.6 7.22
29
thru 62
The MAs and IQs of the subjects were determined
by means of the Stanford-Binet, LM (Terman and Merrill,
I960). This test was administered by the experimenter to
each subject not more than thirty days prior to the sub
jects participation in the experiment. Twenty-two
I
subjects who qualified for the experiment on the basis of
their CA and MA were, however, not included. Eighteen of
these were disqualified for failure to meet the learning
criterion during the training series and four were elim
inated because of errors of the experimenter in conduct
ing the experimental procedures. The subjects were
I
assigned at random to the eight experimental conditions.
Apparatus
In order to ascertain the workability of the
apparatus and procedures a pilot study involving eight
subjects was performed prior to the study herein re
ported. As a result of this pilot study the following
apparatus and procedures were adopted.
The discrimination apparatus was placed on a
table between the experimenter and the subject. This
apparatus consisted of two pieces of quarter-inch ply
wood each measuring two feet by three feet and joined in
such a way that the vertical member bisected the two
feet dimension of the horizontal, or base, member. The
result was a screen two feet and three feet long, mounted
on a base three feet long and two feet broad.
The purpose of the screen was to conceal from the
subject the experimenter's movements in preparing the
discriminanda and recording the results of the trials.
In the middle of the lower edge of the screen (the ver
tical member) was a ten inch by ten inch opening. A
small door, hinged at the top of the opening, was placed
on the experimenter's side of the screen. A handle was
attached to this door, which permitted the experimenter
to raise the door sufficiently to slide the discriminanda
through the opening on an eight inch by ten inch piece of
cardboard.
I
49
The discriminanda consisted of a pair of large
coffee cups, a pair of small (demitasse) cups, a pair of
large cardboard squares with, a bottle cap glued to the
middle of their underside, and a similar, but smaller,
pair of cardboard squares. One member of each of these
pairs of discriminanda was painted black and the other !
one was painted white. The cardboard squares were mounted
on the bottle caps for ease of handling. The squares
extended far enough beyond the bottle caps so that the
bottle caps were not visible to the subjects* The large
cardboard squares measured four inches by four inches*
The small cardboard squares were two and half inches in
i
diameter at the brim. The cups were presented over
turned, that is, resting on their brims. The cups were
considered concrete discriminanda (CD) and the squares
were considered abstract discriminanda (AD).
Half of the subjects were reinforced by hearing
the experimenter say "Right" or "Wrong" immediately
following their choices. This was considered abstract !
reinforcement (AR). The other half of the subjects, in
addition to hearing the experimenter say "Right" or
"Wrong" following their choices, were promised ten pieces
of candy known commercially as "M and M"s. This candy
was considered a concrete reinforcement (CR).
During the first discrimination all subjects
50
responded to one of two pairs of discriminanda. The two
pairs were designated LB— SW and SB— LW. In groups
presented with abstract discriminanda this meant that
they were presented with a large black square and a
small white square or a small black square and a large
white square. Subjects in groups presented with concrete
discriminanda were presented with a large black cup and
a small white cup or a small black cup and a large white
cup. During the second discrimination trials the stim
ulus cue which was varied randomly during the first
discrimination trials was now held constant by being
present or absent for both discriminanda during the pre
sentation of either pair. Bor example, if "large” were
the relevant cue during the second discrimination trials,
then the pairs of discriminanda presented to a subject
would be either SB— LB or SW— LW, thus neutralizing the
cue of "color."
Method
All testing was done in a room that was reason
ably quiet and free from distractions. Upon entering the
testing room the subject was shown both sides of the
discrimination apparatus as well as the discriminanda
that he would be working with. The purpose of this pro
cedure was to eliminate the mystery of what the experi
menter might be doing behind the screen and thus reduce
!
51
the subject's motivation for peeking around the screen
during the experiment. The subject was then seated on '
i
his side of the screen and the experimental instructions j
were given, as follows:
Here is what we sire going to do. I will push two
of these things (experimenter indicates the discrim
inanda to be used with the subject) through this
opening on this piece of cardboard. You are to guess ;
which one is rig^it. Tap your finger on the one you
think is right. If you tap the right one I will say
"right." If you tap the wrong one I will say
"wrong." Your Job is to figure out some way to guess :
which one is right every time.
Approximately every ten trials the experimenter would
repeat the phrase, "Try to figure out some way to get it
right every time." This phrase was used following in
correct choices only. For example, if on trials ten and
eleven the subject made correct choices, the experimenter
said nothing. If trial twelve yielded an incorrect
choice, then the phrase would be repeated. The occasional
repetition of this phrase was felt necessary, because of
the limited attention span of retardates.
The above instructions were used with subjects !
receiving abstract reinforcement (AR groups). Subjects
receiving concrete reinforcement (CR groups) were given
these additional instructions. "If you can figure out
how to guess which one is right every time, I will give
you ten of these "M and M's." The experimenter then held
the bag of candies open for the subject to see the reward.1
Design
A 2x2x2 factorial design was used. The three
dimensions were: reversal and nonreversal shift, abstract
and concrete reinforcement and abstract and concrete
discriminanda. The resulting eight-celled design is
i
shown in Figure 4-. Each cell represents a group of six
teen subjects. Subjects in the reversal groups were
subdivided into four orders of stimulus presentation,
namely black-to-white (black stimulus correct during
training trials then white stimulus correct during test
trials), white to black and small to large (as explained
above). Subjects in the nonreversal groups were sub
divided randomly into eight subgroups according to the
order of stimulus presentation as follows: black to
small (black stimulus correct for training trials and
small stimulus correct for test trials), black-to-large,
white-to-small, white-to-large, large-to-black, large-to-
white, small-to-black, and small-to-white (as explained
above)* For example, size (large to small) would be
irrelevant if color (white) were the significant cue.
Therefore, the table of random numbers was used to deter
mine for a particular trial whether the small, white or
large, white discriminanda should be presented, and thus,
whether the LB-SW or LW-SB sequence should be used. The
criteria for successful learning during the training
TYPES OF REINFORCEMENT
Abstract Reinforcement (AR) Concrete Reinforcement (CBJ
B -* V
V -»B
L -* S
S -* L
/-\
2
R AD— AR AD— CR
v_^
C C S
s
-p 4 Ss in each condition N-16 N-16
O
a J
P i
-*S
B -» L
W -» S
W ->L
L -9 B
L 3 V
S -»B
s w
d
•H
a
£ j
■P
O Q NR
AD— AR AD— CR
<
O
0 3
2 Ss in each condition N-16 N-16
•H
fi
s~\
O
0 3
0 )
e
o
'w'
R
CD— AR CD— CR
< D
-p
N-16
f n
O
d
o
o
NR
CD— AR CD— CR
N-16
SU128 |
Fig* 4— Experimental Design
54
trials was nine correct choices out of ten consecutive
trials. When this criterion was achieved, the test
trials were run with no break in the presentation of
i
discriminanda and no additional instructions to the sub
jects. The criterion for successful learning of the test j
trials was the same as that for the training trials. j
CHAPTER IV
RESULTS
The measure of the dependent variable was the
number of trials to reach the criterion of nine correct
responses out of ten consecutive responses. The criter
ion trials were excluded from the calculations.
Table 2 shows the means and standard deviations
of the trials to criterion for the eight experimental
groups in the first discrimination series and the second
discrimination series. This table shows that, on the
average, the subjects learned both the first discrimina
tion and the second discrimination in relatively few
trials. The mean for the first discrimination was 9*9
and for the second discrimination 9.6. This indicates
that the task was not very difficult for most of the
subjects, despite their level of retardation. Table 2
also reveals that the distributions were skewed and the
variances heterogeneous. This finding is in line with
findings from previous studies in this area (Kendler and
D'Amato, 1955; Kelleher, 1956; Kendler and Kendler, 1959)
In order to reduce the heterogeneity of the
variance an appropriate transformation was sought.
55
TABLE 2
MEANS AND STANDARD DEVIATIONS OF RAW SCORES— TRIALS TO CRITERION IN
FIRST DISCRIMINATION AND SECOND DISCRIMINATION SERIES
First Discrimination Series Second Discrimination Series
Group Mean s s ................................................ Mean """■ ' SB ....................
ARAD-R
11.9
18.45 9.6 12.04
ARAD-NR
4.9 6.29 5.7 5.37
ARCD-R 6*6 11.45 2.7 2.59
ARCD-NR
6.7 12.03
18.8
29.87
CRAD-R
17.5
23.18
14.1
20.22
CRAD-NR
8.3 16.75
5.2 5.10
CRCD-R
9.9 15*97 8.3
9.16
CRCD-NR 13.6
23.39
14.1
17.95
ALL SUBJECTS
9.9 15.94 9.6
12.79
Coder
AR (abstract reinforcement)
AD (abstract discriminanda)
CR (concrete reinforcement;
CD (concrete discriminanda)
R (reversal)
NR (nonreversal)
57
Following Winer (1962), ranges were used for this pur
pose. Table 5 shows the ranges of the eight groups under
the square-root and logarithmic, i.e. log+1 transforma
tions. Inspection of this table reveals that both of
these transformations were effective in reducing skew- j
ness and heterogeneity of variance. The logarithmic
transformation was the most effective of the two, how
ever, and thus was applied to the raw data preparatory to
!
performing the analysis of variance.
Table 4 srhows the means and standard deviations
of the eight groups for the first discrimination trials
and second discrimination trials following the applica
tion of the logarithmic transformation. Inspection of
this table reveals that skewness and heterogeneity of
variance are still present, but are considerably reduced
from what they were in the raw data. In view of the
findings of Lindquist (1955) regarding the use of tests
of homogeneity of variance, such a test was deemed un
necessary prior to the analysis of variance.
The data were subjected to a three factor analy
sis of variance (reinforcements X discriminanda X shifts).
Appreciation for assistance in this computation is ex
pressed to the Health Sciences Computing Facility,
Department of Preventative Medicine and Public Health,
School of Medicine, TJCLA.
TABLE 3
RANGES UNDER VARIOUS TRANSFORMATIONS— TRIALS TO CRITERION IN
FIRST DISCRIMINATION AND SECOND DISCRIMINATION SERIES
First Dis crimination ISeries Second Discrimination Series
Group Original Square-Root Logarithmic Original Sauare-Root Logarithmic
ARAD-R
69 8.307 1.85 56 5.083 1.28
ARAD-NR 26
5.099
1.4-3
17 4.123 1.26
ARCD-R 43 6.557
1.64 8 2.000 0.70
ARCD-NR
50 7.071
1.71
86 9.274- 1.94-
CRAD-R 90 9.487 1.96
87 9.527
1.94
CRAD-NR 61 7.810
1.79 15 5.875
1.20
CRCD-R 58 7.616
1.77
50 4.568 1.21
CRCD-NR
83
9.110 1.92
75
8.660 1.88
Code:
AR (abstract reinforcement)
AD (abstract discriminanda)
CR (concrete reinforcement)
CD (concrete discriminanda)
R (reversal)
NR (nonreversal) vj i
TABLE A
LOGARITHMIC TRANSFORMATION OF MEANS AND STANDARD DEVIATIONS— TRIALS TO CRITERION IN
FIRST DISCRIMINATION AND SECOND DISCRIMINATION SERIES
First Discrimination Series Second Discrimination Series
Group Mean ' ' Sti Mean 31)
ARAD-R
0.69 0.59
0.74 0.40
ARAD-NR 0.58 0.40 0.46 0.40
ARCD-R 0.52 0.52 0.50
0.25
ARCD-NR 0.60 0.46
0.71 0.72
CRAD-R 0.92
0.57
0.96 0.42
CRAD-NR
0.53
0.54 0.58 0.44
CRCD-R 0.72 0.46 0.80
0.37
CRCD-NR 0.66 0.65
0.87 0.58
MEAN
0.65 0.53
0.70
0.45
Code:
AR (abstract reinforcement)
AD (abstract discriminanda)
CR (concrete reinforcement)
CD (concrete discriminanda)
R (reversal)
NR (nonreversal)
ui
vO
_i
1
60
First Discrimination Trials |
The analysis of variance of the transformed data
for the first discrimination trials is presented in Table
5* The analysis revealed no significant differences for
the main effects or the interactions. The F ratio re- j
quired for rejection of the null hypothesis at the .05
i
level of confidence with df=120 is 3*92. The largest F j
ratio obtained from this analysis was 1.78f which does
not approach significance. This indicates that there i
i
were no significant differences in performance among the
eight groups of subjects immediately prior to their be
ginning the second discrimination trials. It implies
that during the first discrimination trials no differ
ential effect was created by either the nature of the
discriminanda (abstract vs. concrete) or the nature of
the reinforcement (abstract vs. concrete). It also
implies that there was no significant difference between
the first discrimination trial performance of subjects
who were later to undergo the reversal shift and the
nonreversal shift conditions.
Second Discrimination Trials
The within-cells mean square was used as the
error term for evaluating the significance of all of the
main effects and interactions. The analysis of variance
of the transformed data for the second discrimination
61
TABLE 5
ANALYSIS OF VARIANCE— TRIALS TO CRITERION IN
FIRST DISCRIMINATION SERIES
(Logarithmic Transformation)
Source Sums of Squares ■a f
"f*--
P
A Reinforcement 0.5916 1 1.30 ns
B Discriminanda
0.1047 1
0.35
ns
C Reversal 0.4442 1 1.47 ns
AB 0.0162 1
0.05
ns
AC 0.3528 1
1.17
ns
BC 0.5582 1 1.78 ns
ABC 0.0406 1
0.13
ns
Within Cells 56.2428 120
Total
38.1311 127
*F at .05 level - 5*92
62
trials is shown in Table 6. Inspection of this table
reveals a significant interaction between the nature of
the discriminanda and the type of shift. It will be
recalled that hypothesis 1 predicted a significant inter-
i
action between these variables. The data confirm this
prediction at the .01 level of confidence. Table 7 shows J
the direction and significance of the specific effects
involved in that interaction. Prom these data it can be
seen that the interaction was of the nature predicted by
i
i
hypotheses la and lb. That is, reversal groups reached
the criterion faster with concrete discriminanda and non
reversal groups reached the criterion faster with abstract
discriminanda. However, the t-tests reveal that only the
effect of the abstract discriminanda can be considered
significant. The effect of the concrete discriminda,
while in the direction predicted by hypothesis la, failed
to reach conventional levels of significance.
The other significant finding revealed in Table 6
is that of a significant effect exerted on the subjects' j
performance during the second discrimination trials by
the nature of the reinforcement. Table 8 shows the
direction and significance of this effect. Inspection of
this table reveals that the difference between reinforce
ment conditions was in the direction that would be
inferred from hypotheses 2, 2a and 2b, although no
TABLE 6
ANALYSIS OP VARIANCE— TRIALS TO CRITERION IN
SECOND DISCRIMINATION SERIES
(Logarithmic Transformation)
Source Sums of Squares df
- j*- ■ ■
£
A Reinforcement
1.3654-
1
5.89 .05
B Discriminanda 0.0332 1 0.14 ns
C Shift 0.2538 1
1.09
ns
AB
0.0167
1 0.07
ns
AC 0.0810 1
0.35
ns
BC 1.64-26 1 7.08 .01
ABC
0.0075
1
0.03
ns
Within Cells
27.824-3 120
Total 31.224-5
127
*P at .05 level 3.92
*F at .01 level 6.85
64
i
I
I
TABLE 7
MEANS AND T-TESTS— LOGARITHMIC TRANSFORMATION TRIALS
TO CRITERION IN SECOND DISCRIMINATION SERIES
Reversal Nonreversal t
- -
Abstract Dis criminanda 0.85 0.55
2.62 <.01
Concrete Dls criminanda
0.65 0.79
1.14 ns
TABLE 8
MEANS AND T-TESTS— LOGARITHMIC TRANSFORMATION TRIALS
TO CRITERION IN SECOND DISCRIMINATION SERIES
f AR Group
0*60
£ CR Group
0.80 2.43 <.05
65
specific hypothesis had heen advanced with regard to the
general effect of concrete reinforcement. That is, from
hypotheses 2, 2a and 2b it would follow that concrete
reinforcement would be assumed to have a facilitating,
rather than an inhibiting, effect upon the subjects' per
formance. This is, in fact, what was found*
Returning to Table 6, it is seen that hypotheses
2, 2a, and 2b (concerning a significant interaction be
tween type of shift and nature of reinforcement) were not
confirmed. The main effects of type of discriminanda and
type of shift did not reach significance. The same was
true for the interaction between type of reinforcement
and type of discriminanda and for the second order inter
action. Appendix Tables 1 and 2 show sums and sums of
squares for the raw and transformed data respectively*
In view of the findings of Kendler and Kendler
(1959) to the effect that the performance of subjects
during the second discrimination was related to their
performance during the first discrimination, the data was
subjected to an analysis of covariance. Using this tech
nique the subjects scores for the second discrimination
were adjusted in terms of their scores for the first
dis criminat ion*
Table 9 presents the results of this analysis*
Inspection of this table shows that the two significant
1
66 !
F
TABLE 9
ANALYSIS OF COVARIANCE— LOGARITHMIC TRANSFORMATION
Source Sums of Squares . T" P
A Reinforcement
1*2079
1 5.24
<.05
B Discriminanda .0486 1 .21 ns
C Shift
.1599
1
.69
ns
AB .0124 l
.05
ns
AC .0449 l
.19
ns
BC
1.4239
l 6.18 <.05
ABC .0060 1
.03
ns
Within Cells 27.4244
112
Total 30.3270 126
67
findings of the analysis of variance retain their signi
ficance, with the only change being a drop in the level
of significance from .01 to .05 in the case of the dis
criminanda X shift interaction. Tables 10 and 11 show
the direction and significance of the specific effects
i
involved in the significant main effect and interaction, i
The fact that the analysis of covariance failed
to reveal results substantially different from the
results of the analysis of variance is not surprising in
view of the non-significant within-cells correlation of
.12. The fact that this correlation was not significant
is probably an indication that the level of task complex
ity changed with the introduction of the second dis
crimination. The finding from the analysis of variance
that performance was effected by type of reinforcement
during the second discrimination but not during the first
discrimination could also be explained in this fashion.
Clinical Observation
Observation of the subjects' behavior during the
experimental sessions yielded several findings which,
although not quantified, may nevertheless provide useful
information for future investigations of similar nature
or those using this type of subject.
The purpose of the pilot study, performed prior to!
the study herein reported, was to check on the appropri-
68
TABLE 10
ADJUSTED MEANS AND T-TESTS— LOGARITHMIC TRANSFORMATION
TRIALS TO CRITERION IN SECOND DISCRIMINATION
Reversal ftonreversal b " P
Abstract Dls criminanda
0.85 0.53
2.48 <.05
Concrete Discriminanda 0.66
0.79
1.12 ns
TABLE 11
ADJUSTED MEANS AND T-TESTS— LOGARITHMIC TRANSFORMATION
TRIALS TO CRITERION IN SECOND DISCRIMINATION
0.61
~ f ITroup
0.80
2.29 <.05
69
ateness of the problem with subjects of this level of
retardation and to determine if the apparatus and pro
cedures were workable. In the pilot study a mixed
assortment of three types of candy ("M and M"s, jelly
beans, and after-dinner mints) was used as concrete
reinforcement. The subjects were permitted to choose
which type they wanted. They showed a definite prefer
ence for the MM and MMs and the after-dinner mints and
rarely selected jelly beans. For reasons of economy,
the "M and M"s were selected for use in the experiment.
Following the subject’s indication of his pre
ference of candy and prior to each trial, the chosen
type of candy was placed under the correct cup (for the
concrete discriminanda groups) or the bottle caps glued
to the underside of the correct cardboard squares (for
the abstract discriminanda groups). If the subject in
dicated the correct member of the pair of discriminanda,
the experimenter lifted the discriminanda and allowed
the subject to have the candy. The instructions to the
subject stated that he could eat the candy at that time
or save it, as he wished.
This procedure was found to have several dis
advantages. For one thing, some subjects, in their
eagerness to get the candy, would lift both of the dis
criminanda simultaneously, thus ignoring the problem
70
posed by tbe discrimination task. Another difficulty
was that the subjects soon accumulated a pile of "M and
MMs. Often they would then become so interested in their
accumulating hoard that they gave less and less atten
tion to the discrimination task. A further difficulty
with this arrangement was that the subjects who took
longer to learn the discrimination generally received
more candies than subjects who learned the discrimination
quickly. These difficulties were avoided by giving the
candy to the subjects following the learning of the test
discrimination.
During the experimental sessions the following
behaviors occurred, which appear worthy of noting:
1. Some subjects seemed uncomfortable when tin-
able to see the experimenter. They would push open the
door in the screen and peer at the experimenter until
instructed not to do so.
2. Upon entering the experimental room the
subjects generally appeared interested in the experi
mental apparatus. Allowing them to see both sides of
the screen and showing them how it worked reduced this
curiosity.
3* Only a very few subjects (about six) made
comments concerning the various stimulus dimensions. Not
infrequently, those who made such comments did not seem
to be guided by this information in their choices. For
example, they might mention black, but then choose the
j
i
white discriminandum.
4-. The subjects generally seemed eager to be
taken to the experimental room. Many of them asked to be j
tested again. Being removed from the routine of the day- |
hall or yard may have been perceived as rewarding by
them. They may also have found it rewarding to be given
individual attention by a member of the staff. In this
connection, it should be noted that most of the subjects
were tested on Saturdays. Their presence on the ward
that day generally meant that they had not been home for
a visit within the past five days at least. Consequently,
they probably had not had much individual attention from
an adult during that period.
CHAPTER V
SUMMARY AND CONCLUSIONS
Summary
Purpose
The purpose of the present study was to investi
gate the development of discrimination learning in the
moderately retarded. Use of a reversal-nonreversal type
of discrimination learning task permitted an assessment
of the nature of the thinking process in the subjects.
Studies of a similar nature have located other points on
this developmental dimension. One particular study
(Kendler and Kendler, 1959) demonstrated that normal
children of kindergarten age are in the process of de
veloping mediational thought processes with this type of
task. The present study was designed to determine
whether retarded subjects of lower mental age than those
in the Kendler and Kendler study could use a mediational
thought process with this type of task under certain
motivational and stimulus conditions.
Procedure
The subjects were 128 institutionalized mentally
retarded males with an average MA of 5*3 years. Their
73
average age was 13*0 years and their average IQ was
4-5.6. The experimental design involved two reinforcement
conditions (abstract and concrete), two types of dis-
criminanda (abstract and concrete), and two types of
learning shift (reversal and nonreversal)* Abstract
reinforcement consisted of the experimenter saying
"Right" or "Wrong" following the subjects response.
Concrete reinforcement consisted of ten pieces of candy
given to the subjects following their completion of the
task. Abstract discriminanda consisted of large and
small cardboard squares painted black or white. Concrete
discriminanda consisted of large and small drinking cups
painted black or white. In the reversal shift condition
the subjects learned one discrimination during the first
discrimination trials (black is right) and then learned
the reverse of this discrimination during the second
discrimination trials ("White is right"). In the non
reversal shift condition the subjects learned a discrim
ination in one dimension during the first discrimination
trials ("large is right") and then shifted to another
dimension during the second discrimination trials ("black
is right").
A 2x2x2 factorial design was used. Use of a
mediational thought process by the subjects would be
inferred if the reversal shift groups learned faster than
74 -
the nonreversal shift groups. Use of a direct associa
tion thought process hy the subjects would be inferred
if the nonreversal groups learned faster than the re
versal groups. It was predicted that faster learning
would result from the use of concrete discriminanda than
from the use of abstract discriminanda. It was also
predicted that faster learning would result from the use
of concrete reinforcement than from the use of abstract
reinforcement.
Results
The experiment yielded the following results:
1. A significant interaction was found between
the type of discriminanda and the type of shift. This
finding is in accord with hypothesis 1 regarding the
effect of the nature of the discriminanda upon speed of
learning in the reversal and nonreversal conditions.
2. Reversal shift groups learned faster with
concrete discriminanda and nonreversal groups learned
faster with abstract discriminanda, although only the
latter of these two findings1 reached conventional levels
of significance. Thus, hypothesis la failed to be con
firmed by the data, but hypothesis lb was confirmed.
3. No significant interaction was found between
the nature of the reinforcement and the type of shift.
This finding indicates that hypotheses 2, 2a, and 2b were
75
not supported "by the data*
4* Concrete reinforcement was found to produce
significantly faster learning than abstract reinforcement
over all conditions in the second discrimination series*
5* The subjects learned the first and second
discriminations rather quickly, requiring between nine
and ten trials on the average.
6* No significant differences in performance
were found among the eight groups at the termination of
the training trials.
The following incidental observations were made
during the experimental sessions:
1. The subjects seemed eager to be chosen to
participate in the experiment. Apparently it was re
warding for them to be removed from the ward routine and
to be given individual attention by a staff member.
2* Some of the subjects seemed to dislike being
denied face-to-face contact with the experimenter during
the experimental sessions.
3. The subjects generally seemed interested in
the task and the apparatus*
4. Occasionally a subject would mention a
stimulus characteristic of the discriminanda, but his
next response often would not appear to be guided by his
statement*
76 l
A pilot study preceded the experiment and yielded ;
the following observations:
1. "M and M^s appear to be a preferred type of ;
candy with this type of subject.
2. Giving the subjects their candy reinforcement
immediately following each of their correct responses
appeared to distract them from attending to the experi
mental task. This procedure also involved the disadvant
ages of providing the slower learning subjects with more
candy than the faster learning subjects.
Discussion of Results and Conclusions
Nature of the Discriminanda
The present study indicated that the nature of
the discriminanda is an-important variable in discrimina
tion learning. Abstract discriminanda produce performance
that is consistent with a single-unit theory of learning,
as shown by significantly slower learning (.01 level) of
i
the reversal task by subjects in the abstract discrimin-
1
anda condition. This finding is in agreement with that of1
Zeaman and House (1963) that abstract stimuli yield slower!
i
i
learning. It is in accord with the earlier finding of
Heidbreder (19^6) that concepts involving abstractions
develop later than those involving concrete objects. It
also agrees with the findings of Kendler and Kendler
|
(1956) that the nature of the task is an important vari—
able in determining the level of inferential behavior in
the subject* The reason why abstract discriminanda are
more difficult for these subjects to react to in a learn
ing situation is open to speculation. One might reason,
with Zeaman and House (1963) that the abstract discrim
inanda are more difficult because they present fewer cues*!
Or one might guess that they are less familiar to the
subject, and thus make him cast about for the proper
aspect to attend to* However, if this were the case, one
must still explain the lack of an opposite finding with
the concrete discriminanda*
The concrete discriminanda used did not produce
significantly faster learning in the reversal shift con
ditions , although the results tended in that direction.
The suggestion from this finding is that further explora
tion of the dimension of concrete discriminanda might
result in behavior that would be consistent with a media
tional theory of discrimination learning. The issues
involved are whether learning is mainly a perceptual or a
motivational matter. Unfortunately, the findings of this
study do not permit a final resolution of this problem.
With regard to the educational implications of
these findings, it would appear appropriate to conclude
that abstract stimuli should be avoided in the instruction
of mentally retarded students of this MA level, since
78
this type of stimuli will tend to make mediational think-j
ing more difficult, for whatever reason. j
Nature of Reinforcement
The results indicate a significant (.01 level)
|
facilitating effect over all conditions of learning from
the use of concrete reinforcement. It should be pointed
out, however, that the candy reinforcement was only one
of the reinforcements for these subjects. They also were
reinforced by the experimenter's statement of "Right" or
"Wrong" following their responses. In addition, they
appeared to derive positive reinforcement from the indi
vidual attention they received while participating in the
experiment. It is assumed that these latter two types of
reinforcement were equated for all subjects. The finding
of a facilitating effect with the use of candy is in
agreement with the results of studies by Kendler and
Kendler (1959)* Heber (1959)* and Kendler and Wells
(I960). The implication for educational practice is to
utilize candy as a motivational device with the mentally
retarded in the classroom. The results of the present
study and the pilot study which preoeded it would indi
cate that the children's preference in candies should
first be determined. The candy then should be given at
the completion of the task rather than at intervals duringj
the task. Of course, good Judgment would have to be j
exercised in dispensing the candy. For example, spacing
of the reinforcements would he necessary in order to
avoid satiation effects* It would also appear advisable
to hear in mind that candy should not he given shortly
before meal-time, lest it reduce the pupil's appetite.
The "social deprivation hypothesis" of Stevenson
and Zigler (1957) could not he evaluated quantitatively
from the present study, hut the subjective impression of
the examiner was in the direction of agreement with the
hypothesis. The majority of the subjects seemed to want
to participate in the experiment and appeared to enjoy
receiving individual attention from a staff member.
Sneed of Learning
The fact that most of the subjects were able to
learn the discriminations rather quickly is in agreement
with the findings of Prothro (194-3) and O'Connor and
Hermelin (1959). The implication for educational prac
tice is that moderately retarded children can learn
differential reactions to instructional materials varying
grossly in color and size, if the discriminations are
taken two at a time.
Effect of Varying Reinforcement
during Training Trials
The findings of no differences between the groups
receiving concrete and abstract reward at the termination
80
of the training trials is rather surprising in view of
the differences found between these groups during the
test trials. Apparently the differential effect of the
two types of rewards became operative only when the task
became complicated by the shifts required. The promise
of candy for successful performance apparently induced j
!
subjects under this condition to exert significantly more'
effort. The implication for educational practice would
be that candy rewards should not be necessary as rein
forcements for all tasks, but should be reserved for the
more difficult tasks.
Interaction between Reinforce
ment and Shi^E
The finding of no significant interaction between
type of reinforcement and type of shift runs contrary to
expectation. The motivational effect of the candy was of
such nature as to increase the learning rate generally
but was ineffective in causing the subjects to change
i
from a direct association to a mediational approach to
the task. Whether the subjects used a single-unit or a
mediational process apparently was more a function of
their cognitive level than a function of the particular
motivational conditions utilized.
Considering the effects of the manipulation of
both discriminanda and reinforcement variables, then, it
81
t
is seen that neither of these proved successful in pro
ducing a significant amount of mediational thinking in
the subjects in connection with this particular task.
However, the finding of a significant discriminanda-
shift interaction suggests that further exploration of
this variable might be rewarding.
Overall versus Interaction Effect
of Varying Bis crimlnanda
The finding of no significant overall effect from
varying the type of discriminanda, but a significant
interaction between shift and discriminanda, suggests
that the task must have been somehow different for the
subjects following the shift in relevant stimuli. Pre
sumably the shift complicated the subject's task, since
it added a new factor to the proceedings. This new
complication of the task, together with the use of
abstract discriminanda, produced significantly faster
learning by the subjects in the non-reversal groups.
During the training trials, the task was less complex
and the nature of the discriminanda had no differential
effect. The implication for educational practice with
the moderately retarded would be to attempt to keep the
learning tasks simple in order to encourage mediational
thinking. For example, learning a rule and its ex
ceptions in one lesson may well prove so complicated a
task for the moderately retarded that they would use !
direct associative thinking. Spacing this learning might :
j
enable them to utilize mediational thought processes. |
Research which will supply more information about such
matters is needed.
The finding of no significant overall effect from
varying the type of shift probably indicates that the
subjects were not using one or the other type of pro
cess consistently. It may well be that the subjects
could use either a direct association process or a media
tional process, depending upon the situational variables
present at the time. A further exploration of the effect
of such variables should provide valuable information on
the thought processes of the moderately retarded.
Alternative Theoretical
Explanations
Although the present experiment has been des
cribed within the framework of reinforcement theory, it
should be noted that the results could be explained by a
number of other theories also. I1 or example, the frustra
tion theory of Amsel (1958)* although primarily based
upon studies of rats in double runway mazes, may be
applied to the present experiment. Amsel's findings in
dicate that when the subject has been frequently rein
forced for making a certain response, a fractional anti-
cipitory goal response results* This leads the subject
to "expect" a reinforcement. At this point, a lack of
reinforcement following a response will lead to frustra
tion.
Repitition of this sequence of events leads to
the build-up of a fractional anticipitory frustration
response. This response consists of a mixture of moti
vating and inhibiting qualities. Applying this theory to
the present experiment, then, one would conclude that
abstract discriminanda are associated with a significant
increase in the inhibition factor and concrete reinforce
ment with a significant increase in the frustration-
induced motivation factor*
Zeaman and House (1963) theorize that learning in
retardates is as rapid as in normals, once the retardates
attend to the relevant stimulus cue. They feel that the
attention factor is the crucial element. Accordingly,
they would explain the present findings as an indication
that abstract discriminanda are less adequate in atten
tion value, while a candy reinforcement would tend to
enhance the subject's attention to the stimuli.
According to Estes and Burke (1953) "tHe subjects
in the present experiment would form a correct S-R asso
ciation by contiguity at full strength in one trial.
Achieving the criterion would then be a matter of in-
84-
creasing the number of correct stimulus elements asso
ciated with the response choice and decreasing the in
correct stimulus elements so associated* According to
the present findings, this process would be inhibited by
the use of abstract discriminanda and facilitated by the
use of concrete reinforcement.
Bush and Mosteller (1951) theorize that a sub
ject in a discrimination learning experiment attends to
only a portion of the entire stimulus complex. Event
ually, through the operation of positive and negative
reinforcement following responses to various of the stim
ulus elements, the discrimination is learned. Abstract
discriminanda would be seen by these authors as slowing
this process, while a concrete form of reinforcement
would be seen as accelerating it.
Lawrence (194-9* 1950) speaks of "relevant" and
"irrelevant" stimuli. Consistent reinforcement following
a response to a relevant stimulus gives rise to a
mediating response to that cue. This adds distinctive
ness to the cue and facilitates the learning. Lawrence
would interpret the present findings as indicating that
abstract discriminanda tend to retard this process, while
concrete reinforcement tends to facilitate it.
Discussion of Observations
The observations that some subjects appeared to
85
dislike the lack of face-to-face contact with the experi
menter during the experimental sessions, that they seemed
interested in the task, and that they seemed eager to he
chosen to participate are all in agreement with the
"social deprivation hypothesis" of Stevenson and Zigler
(1957)* These observations all suggest that the subjects
anticipated positive reinforcement when in a one-to-one
relationship with a staff member. The implication of
this finding for educational practice would be that in
dividual attention can be used as an effective motiva
tional device in teaching the moderately retarded.
The tendency for some subjects to mention a
stimulus characteristic and then appear to ignore it in
their next response is in agreement with the findings of
Lurie (1959) regarding the lack of verbal-motor coordina
tion in the retarded. It also agrees with the findings
of O'Connor and Hermelin (1959) that retardates can learn
a discrimination task but be unable to verbalize the
principle they used* The educational implication of
these findings would appear to be that teaching the mod
erately retarded to verbalize the principles guiding
their behavior in learning situations would probably not
be a very rewarding approach.
As a general conclusion, it can be said that the
results of this study indicate that the nature of the
discriminanda and tlie nature of the reinforcement in
learning situations are significant variables and deserve
further exploration. The nature of the discriminanda can
influence the type of thinking process used by the ment
ally retarded. More information is needed regarding the
effect of such variables in order that eventually the
level of functioning of the mentally retarded may be
raised.
B I B L I O G R A P H Y
BIBLIOGRAPHY
Amsel, A. "The Role of Frustrative Nonreward in Non-
continuous Reward Situations.” Psychol. Bull., 1958,
55, 102-119.
i
Barnett, C. D., Ellis, N. R., and Pryer, M. W. "Learning'
in Familial and Brain-injured Defectives." Amer. J.
Ment. Defic., I960, 64, 894-901.
Barnett, C. D., and Pryer, M. W. "A Note on Depth Per
ception in Defectives." Percept. Motor Skills, 1958,
8, 159.
Bensberg, G. J., Jr. "Concept Learning in Mental De
fectives as a Function of Appropriate and Inappro
priate 'Attention Sets'." J. Educ. Psychol., 1958,
49, 157-143.
Bensberg, G. J., Jr., and Cantor, G. N. "Reaction Time
in Mental Defectives with Organic and Familial
Etiology." Amer. J. Ment. Defic.. 1958, 62, 534-557.
Bolles, M. M. "The Basis of Pertinence." Arch. Psychol.,
N.Y., 1937, No, 212.
Bush, R. R., and Mosteller, F. "A Mathematical Model for
Simple Learning." Psychol. Rev.. 1951, 58, 313-323*
Bush, R. R., and Mosteller F. "A Model for Stimulus
Generalization and Discrimination." Psychol. Rev.,
1951, 58, 413-423.
I
Buss, A. H. "Rigidity as a Function of Reversal and Non-
reversal Shifts in the Learning of Successive Dis
criminations." J. Exp. Psychol.. 1958, 45, 75-81.
Buss, A. H. "Reversal and Nonreversal Shifts in Concept
Formation with Partial Reinforcement Eliminated."
J. Exp. Psychol.. 1956, 52, 162-166.
Cantor, G. N., and Hottel, J. V. "Discrimination Learning
in Mental Defectives as a Function of Magnitude of
Food Reward and Intelligence Level." Amer. J. Ment.
Defic.. 1955, 60, 380-384.
88
89
Capobianco, B. J*, and Funk, E. A* A Comparative Study
of Intellectual, Neurological, and Perceptual Pro
cesses . as Kelajefl ho Heading Achievement of jfeogenous
and Endogenous Retarded Beaders. Syracuse N.Y.t
Syracuse University Research Institute, 1958* 102pp.
Capobianco, E. J., and Miller, D. Y. Quantitative and
Qualitative Analyses of Exogenous and Endogenous
Children in Some Reading Processes. Syracuse, fr.Y. t
Syracuse University Research Institute, 1958* 61 p.
Dollard, J*, and Miller, N. E. Personality and Psycho- '
therapy. New York: McGraw-Hill• l95o.
Dunn, L. M., and Capobianco, E. J. "Studies of Beading
and Arithmetic in Mentally Eetarded Boys." Monogr.
Soc. Bes.. Child Develop.. 1954-* 19* No. 1:7-99*
Dunn, L. M., and Capobianco, E. J. "Mental Betardationt
A Beview of Besearch. In: Eothstein, J. H. Mental
Betardation: Beadings and Besources. New YorSX
Holh, Binehart and Winston, 1961.
Ellis, N. B. "Amount of Beward and Operant Behavior in
Mental Defectives." Amer. J. Ment. Defic., 1962. 66.
613-617*
Ellis, N. E., and Distefano, M. K. "Effects of Verbal
Urging and Praise Upon Botary Pursuit Performance in
Mental Defectives." Amer. J. Ment. Defic., 1959, 64-,
486-490.
Ellis, N. B., and Pryer, M. W. "Primary Versus Secondary
Beinforcement in Simple Discrimination Learning of
Mental Defectives." Psychol. Beports, 1958, 4, 67-70.
Estes, W. K., and Burke, C. J. "A Theory of Stimulus
Variability in Learning." Psychol. Bev., 1953, 60.
276-286.
Farber, B. "Effects of a Severely Mentally Betarded Child;
on Family Integration." Monogr. Soc., Bes. Child
Developm.. 1959* 24, No. IT.
Gewirtz, J., and Baer, D. "The Effect of a Brief Social
Deprivation on Behaviors for a Social Beinforcer."
J. Abnorm. Soc. Psychol.. 1958, 56, 49-58.
Goldberg, I, I. "Some Aspects of the Current Status of
Education and Training in the United States for
Trainable Mentally Retarded Children," Exceptional
Children. 1957, 24, 226-227-
Goldstein, K. "Concerning Rigidity." Char* Pers., 1943,
11, 209-226,
Goldstein, K., and Scheerer, M. "Abstract and Concrete
Behavior," An experimental study with special tests.
Psychol. Monogr.. 1941, 53, No. 2.
Gordon, S., O’Connor, N., and Tizard, J. "Some Effects
of Incentives on the Performance of Imbeciles." Brit,
J. Psychol., 1954, 45, 277-287. _
Guthrie, E. R. The Psychology of Learning. Revised.
New Yorki Harper, 1952*
Halpin, V. C. "Rotation Errors Made by Brain-injured and
Familial Children on Two Visual-motor Tests." Amer,
J. Ment. Defic.. 1954-55, 58, 485-489-
Halpin, V. C. "The Performance of Mentally Retarded
Children on the Weigl-Goldstein-Scheerer Color Form
Sorting Test." Amer. J. Ment. Defic. 1958, 62, 916-
919.
Harlow, H. F. "The Formation of Learning Sets." Psychol.
Rev.. 1949, 56, 51-65.
Harrow, M., and Friedman, G. B. "Comparing Reversal and
Nonreversal Shifts in Concept Formation with Partial
Reinforcement Controlled." J. Exp. Psychol., 1958,
55, 592-597.
Heber, R. "Motor Task Performance of High Grade Mentally
Retarded Males as a Function of the Magnitude of In
centive." Amer. J. Ment. Defic.. 1959, 63, 667-671.
Heber, R. "Personality." Int Mental Retardation.
Stevens, H. A., and Heber, R. Chicagot The tfniver-
sity of Chicago Press, 1964.
Heidbreder, E. "The Attainment of Concepts. III. The
Process." J. Psychol., 1947, 24, 93-158.
House, B. J., and Zeaman, D. "Position Discrimination
and Reversals in Low-grade Retardates." J. Comp.
Physiol. Psychol., 1959, 52, 564-565.
91
Hull, C. L. "Simple Trial and Error Learning: A Study
in Psychological Theory." Psychol. Rev., 1930, 37,
241-256.
Kelleher, R. T. "Discrimination Learning as a Function
of Reversal and Nonreversal Shifts." J. Exp. Psy
chol., 1956, 51, 379-384.
Kendler, H. H., and D'Amato, M. F. "A Comparison of
Reversal Shifts and Nonreversal Shifts in Human
Concept Formation Behavior." J. Exp. Psychol., 1955,
49, 165-174.
Kendler, H. H*, Glucksberg, S., and Keston, R. "Per
ception and Mediation in Concept Learning." J. Ex d.
Psychol., 1961, 61, 186-191.
Kendler, H. H., and Mayzner, M. S., Jr. "Reversal and
Nonreversal Shifts in Card Sorting Tests with Two or
Four Sorting Categories." J. Exp. Psychol., 1936,
51, 244-248.
Kendler, T. S., and Kendler, H. H. "Reversal and Non
reversal Shifts in Kindergarten Children." J. Exp.
Psychol.. 1959, 58, 56-60.
Kendler, T. S., Kendler, H. H., and Wells, D. "Reversal
and Nonreversal Shifts in Nursery School Children."
J. Comp. Physiol. Psychol., I960, 53, 83-88.
Korstedt, A., Stacey, C., and Reynolds, W. "Concept For
mations of Normal and Subnormal Adolescents on a
Modification of the Weigl-Goldstein-Scheerer Color
Form Sorting Test." J. Clin, Psychol., 1954, 10, 88-
90.
Kounin, J. "Experimental Studies of Rigidity. I. The
Measurement of Rigidity in Normal and Feebleminded
Persons." Char. Pers., 1941 a, 9, 251-273.
Lawrence, D. H. "Acquired Distinctiveness of Cues: I.
Transfer Between Discriminations on the Basis of
Familiarity with the Stimulus." J. Exp. Psychol.,
1949, 39, 770-784.
Lawrence, D. H. "Acquired Distinctiveness of Cues: II.
Selective Association in a Constant Stimulus Situa
tion. J. Exp. Psychol.. 1950, 40, 175-188.
92
j
Lawrence, D. H. "The Transfer of a Pis crimination Along j
a Continuum." J. Como. Physiol. Psychol.. 1952. 45.
511-516.
Lewin, K. Principles of Topological Psychology. New
York: McGraw-Hill, 19^6.
Lindquist, E. F. Design and Analysis of Experiments in
Psychology and Education. Boston: Houghton hifflin, |
Luria, A. R. "Experimental Study of the Higher Nervous
. Activity of the Abnormal Child." J. Ment. Def. Res..
1959, 3, 1-22.
McCulloch, T. L., Reswick, J., and Roy, K. "Studies of
Word Learning in Mental Defectives. I. Effects of
Mental Level and Age." Amer. J. Ment. Defic.. 1955.
60, 140-145.
McPherson, M. W., "Learning and Mental Deficiency."
Amer. J. Ment. Defic.. 1958, 62, 870-877.
O'Connor, N., and Claridge, G. S. "The Effect of Goal
Setting and Encouragement on the Performance of
Imbecile Men." Quart. J. Exp. Psychol.. 1955, 7, 37-
45.
O'Connor, N., and Hermelin, B. "Discrimination and Re
versal Learning in Imbeciles." J. Abnorm. Soc.
Psychol.. 1959, 59, 409-415.
O'Connor, N., and Tizard, J. The Social Problem of Mental
Deficiency. New York: Pergamon, 1956.
Pascal, G. R., Stolurow, L. M., Zaborenko, R. N., and
Chambers, K. S. "The Delayed Reaction in Mental
Defectives." Amer. J. Ment. Defic., 1951. 56. 152-
160.
i
Plenderleith, M. "Discrimination Learning and Discrim
ination Reversal Learning in Normal and Feebleminded
Children." J. Genet. Psychol.. 1956, 88, 107-112.
Prothro, E. T. "Egocentricity and Abstraction in Children
and Adult Aments." Amer. J. Psychol., 1943. 56, 66-
77-
Restle, F. "A Theory of Discrimination Learning."
Psychol. Rev.. 1955, 62, 11-20. i
93 j
t
Rosenberg, S. "Probability Reversal Learning as a |
Function of Prior Training in Brain Injured Retard
ates." Unpublished study, Vineland Training School,
N.J., 1961.
Rosenberg, S. "Problem-solving and Conceptual Behavior." j
In: Handbook of Mental Deficiency. N. R. Ellis, (EdJ)
New York: McGraw-Hill, 1965*
i
Shallenberger, P., and Zigler, E. "Rigidity, Negative
Reaction Tendencies, and Cosatiation Effects in
Normal and Feebleminded Children." J. Abnorm. Soc.
Psychol.. 1961, 63, 20-26.
Shepps, R., and Zigler, E. "Social Deprivation and
Rigidity in the Performance of Organic and Familial
Retardates." Amer, J. Ment., Defic., 1962, 67 (2),
262-268.
Siegal, S. M. "Discrimination Among Mental Defective,
Normal, Schizophrenic and Brain-damaged Subjects on
the Visual-Verbal Concept Formation Test. Amer. J.
Ment. Defic.. 1957, 62, 338-34-3.
Spence, K. V. "The Nature of Discrimination Learning in
Animals." Psychol. Rev., 1936, 4-3, 4-27-44-9.
Spitz, H. H. "Field Theory in Mental Deficiency." In:
Handbook of Mental Deficiency. N. R. Ellis, (Ed.,)
New York:— HcSrawlHill,” 1963*
Stacey, C. L., and Portnoy, B. "A Study of Concept For
mation by Means of the Object Sorting Tests with Sub
normals." Amer. J. Ment. Defic., 1951, 56, 169-173.
_——_ _ _ _ |
Stevenson, H. W., and Fahel, L. S. "The Effect of Social I
Reinforcemen on the Performance of Institutionalized j
and Noninstitutionalized Normal and Feebleminded
Children." J. Pers. , 1961, 29, 136-14-7.
Stevenson, H. W., and Knoghts, R. M. "The Effect of
Visual Reinforcement on the Performance of Normal and
Retarded Children." Percept. Motor Skills, 1961,
13, 119-126.
Stevenson, H. W., and Zigler, E. F. "Discrimination
Learning and Rigidity in Normal and Feebleminded In
dividuals ," J^_Fers., 1957, 25, 699-711*
Strauss, A. A., and Lehtinen, L. E. Psychopathology and
Education of the Braln-in.jured ChTld. New YorEt
Grune and Stratton, 194?.
Terman, L. M., and Merrill, M. A. Measuring Intelligence.
Boston: Houghton Mifflin, I960*
Tizard, J*, and Loos, F. M. "The Learning of a Spatial
Relations Test by Adult Imbeciles." Amer. J * Ment.
Defic* * 1954-» 59* 85-90.
Weatherwax, J., and Benoit, E. P* "Concrete and Abstract
Thinking in Organic and Non-organic Mentally Retarded
Children." Amer. J. Ment. Defic., 1958, 62, 548-
553.
Werner, J* "Abnormal and Subnormal Rigidity." J*
Abnorm. Soc* Psychol* * 1946* 41, 15-24. I
Werner, J. "The Concept of Rigidity: A Critical Review."
Psychol. Rev*, 1948, 53* 43-52.
White House Conference on Children and Youth. I960,
Information Sheets on Children and Youth. (Washing-
ton, D.C.: Vrh.ite House Conference on Children and
Youth, I960).
Winer, B. J. Statistical Principles in Experimental
Design. New York: McGraw-Hill, 1962•
Zeaman, D., and House, B. J. "The Role of Attention in
Retardate Discrimination Learning." In: Handbook
of Mental Deficiency. N. R. Ellis, (Ed.). New York;
McGraw-Hill, T963.
Zigler, E. "Social Deprivation and Rigidity in the Per- j
formance of Feebleminded Children." J. Abnorm. Soc. S
Psychol., 1961, 62, 413-421.
Zigler, E. "Rigidity in the Feebleminded." In: Read
ings on the Exceptional Child. E. P. Trapp and P.
Himelstein (Eds. ) • tfew York: Appleton-Century-
Crofts, 1961. 141-162 pp.
Zigler, E., Hodgden, L., and Stevenson, H. W. "The j
Effect of Support and Non-support on the Performance
of Normal and Feebleminded Children." J. Pers.. 1950*
26, 106-122. I
A P P E N D I X
APPENDIX TABLE 1
SUMS, SUMS OF SQUARES, AND SUMS OF CROSS-PRODUCTS—
TRIALS TO CRITERION IN TRAINING AND TEST SERIES
Group
Training Series Test Series
X X2 X
ARADR 190
7,711
154 3,796
ARADNR
79 1,019 59 683
ARCDR
105 2,795
44 224
ARCDNR 108 3,036 301
19,929
CRADR
277 13,383
226 9,720
CRADNR
133 5,595
84 850
CRCDR
159 5,647 133 2,445
CRCDNR
217 11,723 225
8,339
Total 1,368
50,909
1,226 45,986
97
APPENDIX TABLE 2
SUMS, SUMS OF SQUARES, AND SUMS OF CROSS-PRODUCTS—
TRIALS TO CRITERION IN TRAINING AND TEST SERIES
(Logarithmic Transformation)
Group
Training Series Test Series
X X2 X X^
ARADR
19.34-
24.7438
19.35 24.2527
ARADNR
18,03 20.7051 . 17.55 19.5927
ARCDR
18.23 21.5151
17.24 18.6878
ARCDNR 18.28 21.5748 20.21 27.7809
CRADR 21.11
29.2669
20.74 27.7656
CRADNR
18.43 22.3253
17.78 20.0814
CRCDR 19.24 24.1130
19.35 23.9471
CRCDNR
.19.50
25.3810
20,5.7
27.5271
Total 132.18 189.6250 152.79 189.6653

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Frustration Versus Secondary Reinforcement In Children Under Continuous And Partial Primary Reinforcement

PDF

Performance Of Retardates On Piagetian Tasks As A Function Of Ethnicity

PDF

Positive And Negative Verbal Reinforcement In The Conditioning Of The Verbal Behavior Of Adolescent Underachieving Delinquents With Dependent, Neutral, And Aggressive Stimuli

PDF

Ethnic Group Differences In Certain Personal, Intellectual, Achievement, And Motivational Characteristics

PDF

Validity Concomitants Of Various Scoring Procedures Which Attenuate The Effects Of Response Sets And Chance

PDF

The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

PDF

Variables Differentiating Mexican-American College And High School Graduates

PDF

A Semantic Differential Investigation Of Critical Factors Related To Achievement And Underachievement Of High School Students

PDF

Color Vision Deficiency In Third And Sixth Grade Boys In Association To Academic Achievement And Descriptive Behavioral Patterns

PDF

Educational Group Counseling Within A Remedial Reading Program

PDF

The Effects Of Distracting Stimuli Upon Mentally Retarded Minors In The Performance Of Different Kinds Of Tests

PDF

Investigation Of A Self-Report Method In The Study Of Underachievement Among Elementary School Children

PDF

The influence of social intelligence and social teaching methods upon academic achievement

PDF

Rigidity Factors And Value Choices

PDF

Development Of A Mid Primary Screening Battery For Schools Of A High Socioeconomic Community

PDF

An Examination Of Positive And Negative Reinforcement In Classical And Operant Conditioning Paradigms In The Primary Psychopath

PDF

An Individual Testing Approach To Assess Creative Abilities

PDF

A Study Of Several Linguistic Functions Of Mexican-American Children In A Two-Language Environment

PDF

An Experimental Analysis Of The Relationship Between The Reliability Of Amultiple-Choice Examination And Various Test-Scoring Procedures

PDF

Inhibitory-excitatory interaction of conditioned autonomic responses through interoceptive and exteroceptive conditioning

Asset Metadata

Creator Kramer, Richard Albert (author)

Core Title Reversal And Nonreversal Shift Performance Of Retardates Under Various Motivational And Stimulus Conditions

Degree Doctor of Philosophy

Degree Program Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, educational psychology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Meyers, Charles Edward (committee chair), Longstreth, Langdon E. (committee member), Metfessel, Newton S. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-220261

Unique identifier UC11360106

Identifier 6611577.pdf (filename),usctheses-c18-220261 (legacy record id)

Legacy Identifier 6611577.pdf

Dmrecord 220261

Document Type Dissertation

Rights Kramer, Richard Albert

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA