h

THE EFFECTS OF FOUR TRAINING PROCEDURES ON THE _
MATCH-To-SAMPLE PERFORMANCE OF SEVERELY AND *
PROFOUNDLY MENTALLY IMPAIRED STUDENTS-

DIssertation for the Degree 0f 'PhL'D.'1 " E
MICHIGAN STATE UNIVERSITY ‘ '
' 7 PATRICIA I. HECHT~ '
1977‘ ’

- —‘ 0A —.-m_. .‘-_

"‘Y

i.
LIPR/T RY ‘

Adidligfi; State- I!
Umvcrszty 4‘

   
 
 

This is to certify that the

thesis entitled

THE EFFECTS OF FOUR TRAINING PROCEDURES
ON THE MATCH-TO-SAMPLE PERFORMANCE OF
SEVERELY AND PROFOUNDLY MENTALLY IMPAIRED
STUDENTS
presented by

Patricia J. Hecht

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Special Education

Maw W

Major professor

DateJM

0-7639

 

 

éfiflé‘

 

ABSTRACT
THE EFFECTS OF FOUR TRAINING PROCEDURES ON THE

MATCH-TO-SAMPLE PERFORMANCE OF SEVERELY AND
PROFOUNDLY MENTALLY IMPAIRED STUDENTS

By
Patricia J. Hecht

One mode of instructional presentation which has
demonstrated results with young normal children and
moderately retarded individuals is vicarious learning,
derived from Bandura's (1971) social learning theory.

The effects of adjunctive verbal coding are equivocal,
apparently as a function of the conceptual level of the
subjects and the complexity of the tasks.

The purpose of this study was to compare the effects
of four training procedures, all involving vicarious
learning, on the match-to-sample performance of severely
mentally impaired persons (SMI). Specifically, this
investigation sought to determine whether correctional
procedures or verbalization would facilitate the learning
of these students. The basic paradigm for the presenta-
tion methods was drawn from studies by Forehand and Yoder
(1973, 1974), who explored the effects of modeling and

verbal cues on concept acquisition. The major extension

Patricia J. Hecht
which occurred in the present study was the use of two
correctional procedures.

Match-to-sample tasks from the Leiter International
Performance Scale have been used by other researchers
(Forehand and Yoder, 1973; Filler and Bricker, 1976) with
educable retardates. In order to have this study relate
to other research, three tasks from the Leiter Scale
involving matching of identities were used.

Twenty-four students in educational programs for SMI
were the subjects of this study. Subjects were screened
on the match-to-sample tasks and randomly assigned to one
of four treatment groups. All training procedures involved
demonstrating the desired response for each trial. Treat-
ments 1 and 4 (cued and guided, respectively) included
a correctional phase if the student did not make the
correct response. The cued condition utilized the visual
cue of positioning the cube directly in front-of the
appropriate slot and pointing to the relevant response.

The guided condition involved physically manipulating the
student to make the correct response. In the verbal direc-
tion condition, the trainer verbalized the descriptive
phrase "put the block here" while demonstrating the correct
placement. In the verbal information condition, the
trainer labeled the relevant attribute (e.g., "put the red
with the red") while demonstrating the correct placement.

Each subject was trained on one task for a maximum of
ten sessions (400 trials) or until achieving criterion of

eight correct responses out of ten. The independent

Patricia J. Hecht
variables were tasks and treatments. Number of trials to
criterion was the dependent variable. The data were
analyzed by ranks using the Hodges-Lehman two-way analysis
of variance for blocked data.

The main hypothesis of this study was that correc-
tional procedures would facilitate the performance of SMI
students over the verbal conditions. It was further
hypothesized that the picture task would be learned more
easily than color or form. The results were inconclusive;
however, by inspection of the actual number of trials to
criterion, it appeared that the cued and guided conditions
were somewhat more facilitative. Picture and color tasks
appeared to be easier than form. Nine of the twenty-four
subjects met criterion and these were relatively evenly
distributed across treatment conditions, accounting for
the non-significant results.

A number of problems were identified in conducting
experimental research with SMI individuals. Discussion
focused on the following issues: subjects, position
preference, visual attending, age, residence, and media-
tion. The dearth of literature on this population may
reflect the many obstacles in doing research with these
individuals. Recommendations were made for questions to
be answered by additional research. It was suggested that
a more fruitful approach with this population might be
through case study or longitudinal study in the naturalis-

tic setting.

THE EFFECTS OF FOUR TRAINING PROCEDURES ON THE
MATCH-TO-SAMPLE PERFORMANCE OF SEVERELY AND
PROFOUNDLY MENTALLY IMPAIRED STUDENTS

BY

_ L;
O‘\(‘

Patricia J? Hecht

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Elementary and Special Education

1977

ACKNOWLEDGEMENTS

This research was carried out with the assistance of
many people in many places. Without each of them it could
not have been completed. The writer wishes to express
sincere appreciation to the many individuals who made
this study possible.

A special thank you is extended to the students who
were the subjects of this study. Grateful appreciation is
extended to the staff and administration of the Grand
Rapids, Lansing, Lapeer Intermediate, Pontiac, and Wayne
County programs for severely mentally impaired for allow-
ing access to their centers. This study could not have
been conducted without their cooperation and assistance
in the identification of students.

Sincere appreciation is extended to all the members
of my committee: to Dr. J. Edwin Keller and Dr. Eileen
Earhart, who lent their ideas and support; to Dr. Don
Hamachek, who encouraged me in the process of becoming;
to Dr. Don Burke, Chairman, who not only guided the
course through the doctoral program but also encouraged
me to dream and then gave his moral support, advice, time,
and direction to this project from its inception to

conclusion.

ii

Finally, deep and sincere appreciation is extended to
my family and friends, who offered personal encouragement

through the years of doctoral work.

iii

Chapter
I

II

III

TABLE OF CONTENTS

PROBLEM AND REVIEW OF RELATED RESEARCH.

Rationale .
Vicarious Learning.
Mentally Retarded
Verbal Mediation.

Match- to- Sample Tasks :

Summary Related to the Present Study.

METHODOLOGY .

Purpose . . .

Special Definitions

Design. . . . .
Overview .

Research Questions

Population. . . .
Procedure . . . .
Materials.

Selection of Subjects.

Treatments .

Training .

Data Collection.

Hypotheses. .

Statistical Analysis.

RESULTS AND DISCUSSION.

Hypotheses Tests.
Hypothesis
Hypothesis
Hypothesis

Findings. . . . .

Discussion. . . .
Subjects .

Position Preference.

1
2
3

Visual Attending .

Age. . . .
Residence.
Mediation.

iv

Chapter
IV SUMMARY AND RECOMMENDATIONS . .

Recommendations . .

Modification of the Present Study:

Additional Research. . . . . . .
REFERENCES .
APPENDICES . . . . . .
A - SCHEMATIC DRAWING OF MATERIALS.
B - FORMS FOR SCREENING AND TRAINING.

C - GRAPHS OF INDIVIDUAL STUDENT'S
PERFORMANCE O O O O O I O O O

Table

&

WVO‘UI

LIST OF TABLES

Distribution by Sex, Age, and Residence .

Means, Standard Deviations, and Ranges for
Ages in Months. . . . . . . . . . . . . .

Trials to Criterion by Treatment and Task .

Means, Standard Deviations, and Ranges for
Trials to Criterion by Tasks. . . . . .

Ranks of Those Who Met Criterion. .
Highest Percentage Achieved Per Session .
Distribution of Ages by Treatment . .

Percentage of Those Who Met Criterion by
Residence . . . . . . . . . . . . . . . .

vi

Page
18

18
26

28
28
31
32

33

LIST OF FIGURES
Figure Page

1 Variable Matrix of 3 X 4 Design . . . . . . . l4

vii

CHAPTER I
PROBLEM AND REVIEW OF RELATED RESEARCH

With the implementation of Michigan's P. L. 198
mandating educational services to all school age handi-
capped persons, the responsibility for programming of
severely and profoundly mentally impaired was shifted
from mental health to education. The definition of who
shall be educated has been expanded to include individuals
who are severely handicapped. A Bureau of Education for
the Handicapped (BEH) task force generated the following
as a working definition in launching a national campaign:

These children who because of the intensity of

their physical, mental, or emotional problems,

or a combination of such problems need educa-

tional, socia1,psychological, and medical

services which will make it possible for them

to maximize their full potential for meaning-

ful participation in society and for self-

fulfillment (Thompson, 1976).

The Federal Government has established as a priority the,
provision of services to low incidence populations and the
preparation of personnel. This current national interest
and emphasis is exemplified by two events which occurred
in the spring of 1975. The first was a national training
meeting, Educating the 24-Hour Retarded Child, sponsored

by the National Association for Retarded Citizens. The

second was the founding of The American Association for

2
the Education of the Severely/Profoundly Handicapped with
the goal of disseminating information and facilitating
communication. BEH estimates the number of such severely
handicapped youngsters at nearly 1.5 million, including
over 460,000 severely to profoundly retarded (Thompson,
1976).

Teachers whose responsibility it is to train these
students need to know whether a particular procedure or
method facilitates learning. What is known about teaching
educable or trainable mentally impaired may not be applicable
to the education of severely mentally impaired (SMI) indi-
viduals. This study explored several methods of teaching
a cognitive task to SMI students. Fundamentally, the
question being asked was, "Do training procedures have a
differential effect on student performance?" Mere specifi-
cally, we were interested in determining whether one of
several methods is most facilitative in the case of SMI

youngsters, a matter which has been virtually unstudied.

Rationale

 

The education and training of persons who are severely
mentally impaired has received increasing attention in the
past few years. Programmatic focus has shifted from one
of care to one of instruction (Roos, 1975). However, the
research on the efficacy of special education offers little
to inspire confidence that traditional practices have much
to offer as a basis for programming for the SMI (Tawney,

1974). Stephens (1976) points out the need to define and

3
address educational issues for this population. Not only
should attention be given to content or curricular areas,
but also to methods of instructional presentation. For
the most part, the technology has been the application of
operant techniques and of the experimental analysis of
behavior. None of the special teaching procedures developed
for young children has been systematically tried out for
mentally retarded individuals (Kessler, 1970). The SMI
population is not homogeneous insofar as learning is
concerned, but there is evidence that the early sensori-
motor development approximates the sequence seen in normal
children, albeit delayed (Woodward and Hunt, 1972; Inhelder,
1968).

The review of related literature will consider vicarious
learning in young normal children as well as relevant
research with retarded persons, primarily educable mentally
impaired, which has explored methods of instructional pre-
sentation. The present study was a spin-off of some of
these studies, although not a replication in the sense of
direct application with an SMI population. In essence,
the cognitive level of tasks used with higher functioning
individuals precludes a direct parallel from research done
with those persons. The paucity of literature with this
severely mentally impaired population will be evident to

the reader.

4

Vicarious Learning

 

The importance of modeling, also referred to as
vicarious learning, demonstration, and imitative learning,
has been demonstrated in the acquisition of concepts in
children, ranging from simple imitative responses to
complex linguistic rule-governed behavior (Flanders, 1968;
Zimmerman and Rosenthal, 1974; Cullinan, Kauffman, and
LaFleur, 1975). The substantial body of experimental
findings derives from social learning theory (Bandura,
1971). In the most comprehensive theoretical account of
modeling, Bandura (1965, 1971) describes both vicarious
effects and observational learning components. Our
interest is in the former, which refers to the observer's
performance as a function of observing a demonstration.
Vicarious acquisition of affective, motor and problem
solving behaviors does occur. Conceptual learning has
been demonstrated through the vehicle of modeling.

Studies with normal children in the cognitive domain
have primarily focused on comparisons of modeling with
and without verbal cues as the basic paradigm. However,
results have not been consistent with regard to the effects
of verbalization on vicarious learning. Rosenthal, Moore,
Dorfman, and Nelson (1971) found that verbal rule cues with
modeling did not lead to better acquisition of a novel
concept by three, four, and five year olds than did modeling
alone. Of interest is the fact that all twelve trials
were demonstrated prior to the observer's performance. In

another study, Rosenthal, Alford, and Rasp (1972) found

S
that those second graders who were in a verbal labeling
condition exceeded all other modeling groups; however,
of note is the fact that those in a group which received
a description of the model's actions were outperformed by
those who observed a silent demonstration. It may be that
the differences lie in the complexity of the tasks or in
the conceptual level of the subjects.

In a series of investigations, Rosenthal and Zimmerman
(1972) studied the acquisition of conservation through
modeling by first graders and four year olds. In Experiment
I conceptual rule verbalization by the model did not facili-
tate the subjects' judgments of equivalence, but did affect
their explanations. Experiment 111 compared the effective-
'ness of modeling with a verbal instructions technique for
Chicano children from Spanish speaking homes and demon-
strated the superiority of modeling over instructions for
these youngsters with limited facility in English.

Rattan (1974), in a study of acquisition, retention,
and transfer of conservation by first graders, found that
providing verbal rules in conjunction with either direct
experience or demonstration was significantly better than
verbal directions. One additional study of interest is
that of Rosenthal and Zimmerman (1973), in which they
investigated the relative effectiveness of modeling and
guided practice. The guided practice procedure, which
constrained the child to perform without error by being
physically put through the movements, was less effective

than simple observation of a model performing the task.

6
Mentally Retarded

 

Several studies (Forehand and Yoder, 1973, 1974; Yoder
and Forehand, 1974) with educable mentally impaired (EMI)
persons have explored the effects of vicarious learning
(modeling) and verbal cues on concept acquisition. Utiliz-
ing match-to-sample tasks from the Leiter International
Performance Scale, which involved categorizing based on
an underlying conceptual rule (e.g., apple goes with pear,
man goes with woman), their focus was on task completion
time and number of errors. In all of their investigations
the model demonstrated one trial and then the observer
performed. In the first study (1973) three treatment
groups were compared: demonstration with verbal concept,
demonstration only, and exposure to materials without
demonstration. Their findings indicated that modeling
with verbalization of the concept produced fewer errors
than did the model only and no model conditions. In a
second study Yoder and Forehand (1974) investigated four
treatments utilizing the same match-to-sample tasks. Two
levels of verbalization were used in conjunction with
demonstration: verbal directions and verbal rule or
concept. They found that providing a verbal description
with modeling did not improve performance over the model
only condition; however, labeling the concept enhanced
the learning of more difficult conceptual tasks. They
suggested that task complexity is important in interpreting
the effects of rule provision on imitative learning. Their

interest in the third study (Forehand and Yoder, 1974) was

7
in whether observing the model was a critical variable.
Might a student acquire the concept as readily by being
exposed to the completed task with concurrent verbal
labeling of the concept? Four treatment groups were
studied: model only, verbal concept with demonstration,
verbal concept without demonstration, and control. The
findings again demonstrated that performance of both
normals and retardates was facilitated by modeling. The
combination of demonstration and verbal cues was more
beneficial than modeling alone.

Litrownik, Franzini, and Turner (1976) conducted a
study with trainable mentally retarded (TMR) students at
a mental age level of 5.67 years. They found that TMRs
could acquire a novel rule-governed concept vicariously
by observing a model. Their results suggest that demon-
stration in which the observer performs following each
trial (imitation) facilitated response matching, while
massed modeling resulted in better transfer. Matching
on imitative trials of an easy concept was not affected
by providing a conceptual rule, a finding which is con-

sistent with Yoder and Forehand.

Verbal Mediation

 

It has been shown that young normal children are
unable to perform as well as older children on a variety
of tasks which require verbal mediation (Landau and Hagan,
1974). However, when the materials were appropriate to

the children's predominantly nonverbal representational

8

modes, younger children were able to remember just as much
as were older ones (Corsini, 1969). In the case of retarded
individuals, several investigators have posited attentional
deficits (Zeaman and House, 1963), an excessively limited
breadth of attention (Zeaman, 1973), attention to multiple,
not necessarily salient, dimensions (Fisher and Zeaman,
1973), or mediation deficiency (Luria, 1961). The studies
by Forehand and Yoder reviewed above appear to suggest that
in the case of EMI subjects, vicarious learning is facili-
tated by labeling the verbal concept. Similarly, in teach-
ing multiple classification skills to EMI students, Choe
(1974) found that labeling and identifying attributes
facilitated the performance of nine to thirteen year olds
over both specific verbal instructions and a control condi-
tibn. The role of verbal mediation with this population
appears to be clear cut in the case of school age subjects.

Whether verbalization facilitates the learning of SMI
students has not been extensively researched. Luria (1961)
argues that the retarded individual does not acquire the
second signal system (language) control over the first
signal system (motor) that is normally acquired in develop-
ment. In a similar vein, Milgram (1973) notes that if
equivalence is assumed between language and thought, the
verbal medium is likely to be relied on heavily in the
education of retarded children. He points out that an
individual may adequately solve problems in the absence of
verbal product either expressively or receptively.

Rosenthal and Kellogg (1973) hypothesized that vicarious

9
learning might have special efficacy for individuals with
limited verbal skills. In a study with adult retardates,
they compared silent demonstration with equivalent informa-
tion provided verbally and found that modeling surpassed
verbal instruction in fostering concept learning. On a
two choice discrmination task with institutionalized
retardates, Bricker and Bricker (1971) found that those
children who were given training did not differ signifi-
cantly on post-testing as a function of training under
conditions of naming or no-naming of the relevant response.
Both groups performed significantly better than the con-
trols, possibly as a result of contingent reinforcement.
The relationship between receptive language and verbal

mediation is not clear in low functioning individuals.

Match-to-Sample Tasks

In a study of teaching styles of mothers with their
pre-school EMI children, Filler and Bricker (1976) utilized
four-choice match-to-sample tasks from the Leiter Scale.
The six items which they selected required matching of
identities. Three were taken directly from the Leiter:
color, picture, and form. Three were created from Leiter
materials and involved matching according to number. In
the study information-containing and information-devoid
verbal directions accounted for 53 percent of the mothers'
preresponse behavior. Limited choices in presentation was
the only behavior which was significantly related to

children's performance. Apparently, verbalization did not

10

relate directly to the performance of these young children.
In a followup with the same subjects, Filler (1976) modi-
fied maternal teaching styles and demonstrated the impor-
tance of preresponse nonverbal manipulation of materials
as compared to postresponse verbal feedback and control
conditions. There was a marked improvement of pre-schoolers'
performance when mothers systematically maximized the
probability of a correct match by limiting choices.

It will be recalled from the earlier discussion about
EMI students that the group of studies by Forehand and
Yoder utilized items from the Leiter Scale. Their tasks

involved understanding a concept as a basis for matching.

Summary Related to the Present Study

 

Studies with normal subjects have explored concept
acquisition, retention, and transfer through vicarious
learning and verbal cue. Although the conceptual level
of these subjects was higher than the population of
interest in the present study, the various methods of
instructional presentation have applicability: silent
demonstration, modeling with verbal rule labeling, and
modeling with verbal description. Investigations with
young normal subjects, as well as those with pre-school
retardates, suggest that the ability to use verbal cues
may relate to the functional level of the individual.
Younger children appear to be less able to benefit from

verbal cues. With older EMI students, the verbalizing of

11
a conceptual rule facilitated performance on a complex
cognitive task.

Of importance in learning through observation is the
observer's ability to attend to and remember what is
modeled. The issue may not be simply one of verbal media-
tion. The role of imitative learning has been considered
in language training for autistic children (see for
example Lovass, 1967). In the training of cognitive
skills for students whose developmental disability is
primarily in the cognitive domain, persons who are SMI,
the interrelationship of vicarious learning and verbal
cues has not been studied.

In the present study, in an attempt to determine the
effects of various training procedures all of which
involved modeling, the three match-to-sample tasks were
used with SMI students which were used previously with

pre-school retardates (Filler and Bricker, 1976).

CHAPTER II
METHODOLOGY

Purpose

This study was undertaken as basic research to examine
the relative effects of four instructional presentations on
the match-to-sample performance of severely mentally
impaired individuals. Specifically, it sought to determine
whether there was a facilitative effect on number of trials
to criterion under the conditions of cued response, verbal
direction, verbal information, or guided response, all of

which involved demonstration (vicarious learning).

Special Definitions

The following terms were used in the description of
this study. Definitions are provided here to facilitate
a common basis of understanding.

Severely mentally impaired - defined as developmental
disability primarily in the cognitive domain such that the
rate of development is approximately one-third or less than
that of normal as identified by an Educational Planning and
Placement Committee.

Match-to-sample performance - defined to be the number
of trials to criterion on tasks (color match, picture match,

form match) from the Leiter International Performance Scale.

12

13

Training session - a block of time, not to exceed ten
minutes, during which forty successive trials were pre-
sented. The maximum number of training sessions was ten.

Trial - defined as the presentation of a single
stimulus block to be matched, in both screening and train-
ing of subjects. The maximum number of training trials
was 400.

Success criterion - eight correct responses out of

ten successive trials.

Design

Overview

As an experimental study, this investigation involved
screening subjects on the match-to-sample tasks and random
assignment to one of four treatment conditions. Each
subject was trained on one task for a maximum of ten
training sessions or until reaching criterion.

The independent variables were tasks and treatments.
Number of trials to criterion was the dependent variable.
All cells contain two subjects. The data matrix is shown

in Figure l.

14

 

 

 

 

 

 

 

 

 

 

 

TASKS TREATMENTS
1 2 3 4
Color
Picture
Form
Figure 1

Variable Matrix of 3 X 4 Design

-Research Questions

This study addressed the basic question of whether or
not various training procedures have a differential effect
on the performance of severely mentally impaired students.
The main question asked was: What is the relationship
between instructional presentation and match-to-sample
performance when all methods involve vicarious learning
(demonstration)? Will there be significant differences
between the four treatments as reflected in number of
trials to criterion on the matching identities task?
Secondarily, the question was asked: Will any of the

three match-to-sample tasks be learned more easily?

Population

 

The population for this study comprised severely
mentally impaired students. Students were limited to those

enrolled in educational programs for severely mentally

15
impaired who had been identified by an Educational Planning
and Placement Committee. The delimitation of program
placement was provided to circumvent the difficulties
inherent in definitions utilizing intelligence scores
with this population.

It was considered important to select students who
were most likely to benefit from short daily training
sessions and who were at a functional level consonant with
a concept of matching. Students who met the following
criteria were identified for further screening:

1. no evidence of severe vision or hearing loss.

2. no evidence of severe behavioral problems.

3. ability to cooperate in ten minute training
sessions.

4. no evidence of gross neurological disease.
5. chronological age not less than seven years
six months.

Procedure

Materials

 

Three items taken directly from the Leiter Interna-
tional Performance Scale (1969) comprised the match-to-
sample tasks. These were three of the tasks used by
Filler and Bricker (1976) with pre-school retardates.

The tasks were: Year II - 1, colors (red, green, blue,

and yellow); Year II - 3, pictures (elephant, chair, doll,
and wagon); and Year III - 1, forms (cross, circle, square,
and star). In order that there would be a task available

at the apprOpriate level for each student when training

16
occurred, the three sets of materials varied in difficulty.
In addition, all subjects were screened on the tasks prior
to training.

In presenting the task, the four stimulus items were
presented side by side on a single strip. This strip was
mounted on a wooden frame with slots to accommodate four
cubes which are matched to identical stimuli. The frame
used in this study was modified from the original Leiter
(see Appendix A for schematic representation). To minimize
the motor component of placing the one-inch tubes, each
slot was widened by one-fourth inch. To eliminate
extraneous reSponses, the frame was shortened from eight

slots to four.

Selection of Subjects

Potential subjects had been assessed on the Reasoning
and Problem Solving OhecKUst (Hecht, 1975). Program
supervisors in eight educational programs for SMI students
identified those individuals whose assessments indicated
the ability to match identical subjects (item 23) but not
match forms (item 32). A concept of matching identities
at the object level was regarded as the minimum functional
level for the training tasks.

These students were then screened individually by the
experimenter on the three training tasks in sixteen trials
of each for a total of forty-eight trials. Approximately
125 individuals were screened. The criterion for exclusion

from a task was originally set at 25 percent or better

17

(chance). The criterion was raised to SO percent (eight
correct) because the majority of students exhibited a
position preference during screening and did not choose
the four alternatives equally; their performance was
better than chance as a function of the position preference.
In the screening procedure each stimulus cube for each
task was presented singly and the student asked to "put it
where it goes." Each block was removed before presenta-
tion of the next. The four stimuli were presented in a
pre-determined random order within each task. The order
in which the tasks were presented was determined also by
random order. Following screening, eight subjects were
randomly assigned to each of the three match—to-sample
tasks. Each of the subjects for each task was randomly
assigned to one of the four treatments.

Twenty-four severely mentally impaired students in
four educational programs in Michigan were the subjects
of this investigation. Table 1 shows the breakdown of
these subjects by sex according to chronological age and
place of residence. Chronological ages ranged from 8-3
years to 25-0 years. There were seventeen males and
seven females. Sixteen of the subjects lived either at
home or in foster care; eight were residents of an insti-
tution for mentally retarded.

The means, standard deviations, and ranges for ages
in months are shown in Table 2. Six subjects are repre-
sented in each treatment. Subjects in Treatment 2 tended

to be younger than those in other conditions. A Kruskal-

18
TABLE 1
DISTRIBUTION BY SEX, AGE, AND RESIDENCE

 

 

Sex Chronological Age
8 9 10 ll 12 13 14 15 l6 17 18 19 20 21 22 23 24 25 N

Day

M 3 3 2 l l 10
F 1 1 1 l l S
Institution

M l l 2 l l 1 7
F 1 l 2

 

TABLE 2

MEANS, STANDARD DEVIATIONS, AND RANGES
FOR AGES IN MONTHS

 

 

TREATMENT X SD Range
1 193.0 87.8 104-300
2 143.1 37.1 100-213
3 186.3 62.5 99-278
4 219.7 53.1 125-290

 

19
Wallace (Siegal, 1956) computed on ages across treatments

was not statistically significant (H=2.97, df 3, p<.30).

Treatments

In all conditions the experimenter was the trainer.
The student was seated at a table with the trainer to the
left and slightly forward. This position was selected
both because it facilitated the student's observation of
the match to sample in the apprOpriate left to right
orientation and it permitted the trainer to tally the
subject's visual attending to the stimuli during demon-
stration, albeit in an inexact way.

The common component in the four training procedures
in this study was a demonstration of the relevant match.
Two of the treatments involved correctional procedures,

two did not. The four conditions utilized were:

Treatment 1. Demonstration with cued response. A

 

treatment condition in which the trainer elicited the
student's attention, modeled the required response by
matching a single block, removed the block and gave the
command, "You do it."

If the student did not respond, the trainer allowed
a time lapse of 5-10 seconds prior to beginning

the next trial. If the student did not respond

on two consecutive trials, the trainer requested

an unrelated behavior known to be in the student's
repertoire (e.g., "Look at me”, "Hands on table",
"Give me the ...") to elicit involvement before

the next trial.

If the student responded incorrectly, the trainer
positioned the block directly in front of the
appropriate slot and pointed to the relevant
response, saying, "Here."

20

If the student responded correctly, the trainer
gave verbal praise.

Treatment 2. Demonstration with concurrent verbal

 

directions. A treatment condition in which the trainer
elicited the student's attention, modeled the required
response by matching a single block, gave the verbal
command, "Put the block here", removed the block, and
gave the command, "You do it."

If the student did not respond, the trainer
allowed a time lapse of 5-10 seconds prior to
beginning the next trial. If the student did
not respond on two consecutive trials, the
trainer requested an unrelated behavior known
to be in the student's repertoire (e.g., "Look
at me", "Hands on table", "Give me the ...")
to elicit involvement before the next trial.

If the student responded incorrectly, the
trainer gave a neutral comment, "OK", prior
to beginning the next trial.

If the student responded correctly, the trainer

gave verbal praise.

Treatment 3. Demonstration with concurrent verbal
information. A treatment condition in which the trainer
elicited the student's attention, modeled the required
response by matching a single block, verbally labeled the
relevant attribute (e.g., "Put the chair with the chair"),
removed the block, and gave the command, "You do it."

If the student did not respond, the trainer

allowed a time lapse of 5-10 seconds prior to

beginning the next trial. If the student did

not respond on two consecutive trials, the

trainer requested an unrelated behavior known

to be in the student's repertoire (e.g., "Look

at me", "Hands on table", "Give me the ...") to
elicit involvement before the next trial.

21

If the student responded incorrectly, the trainer

gave a neutral comment, "OK", prior to beginning

the next trial.

If the student responded correctly, the trainer

gave verbal praise.

Treatment 4. Demonstration with guided response. A
treatment condition in which the trainer elicited the
student's attention, modeled the required response by
matching a single block, removed the block, and gave the
command, "You do it."

If the student did not respond, the trainer

allowed a time lapse of 5-10 seconds prior to

beginning the next trial. If the student did

not respond on two consecutive trials, the

trainer requested an unrelated behavior known

to be in the student's repertoire (e.g., "Look

at me", "Hands on table", "Give me the ...") to

elicit involvement before the next trial.

If the student responded incorrectly, the

trainer made physical contact and guided him

in making the relevant response, saying, "Here."

If the student responded correctly, the trainer
gave verbal praise.

Training

Each training session was comprised of forty trials.
Subjects were trained to a criterion of eight correct
responses out of ten trials, or a maximum of ten sessions,
totaling 400 trials. Training sessions were scheduled
twice daily for five consecutive days. The order in which
the students were trained was randomized across sessions
within each educational program. Due to missed sessions,

five subjects had their ten sessions over four days rather

22
than five. In no case did a student have more than three
training sessions in a single school day.

The stimulus cubes were presented in a predetermined
random order within each training session. Blocks were
presented singly on a standard referent point. Each block
was removed before the demonstration and presentation of
the next. A single trial was defined as the demonstration
of the required response and the subsequent presentation
of the block for the student to match. The stimulus strip
was reversed following each training session to determine
whether consistent choice was matching or a position
reference.

Training of the first seven subjects was conducted
as a pilot study. Treatment 1 for this pilot was a condi-
tion of demonstration only. It was felt during this
initial training phase that students who exhibited a posi-
tion preference would likely improve their performance only
under a correctional procedure. At that time the one
treatment condition which utilized a correctional pro-
cedure also involved physically touching the student,
which might be reinforcing to some subjects. Consequently,
Treatment 1 was rewritten to provide visual cueing as a
correctional procedure equivalent to guidance without
physical manipulation. In both of these conditions the
initial presentation could be considered as demonstration
only, since the correctional component was utilized only

when the student did not reSpond correctly.

23

It was evident in the pilot that the screening pro-
cedure did not identify accurately those individuals
whose visual attending to the task varied within a training
session. It was decided to tally occasions during training
sessions when the student was not visually attending to the
materials during demonstration. No record was kept of
inattention during the student's placement of the cube
nor during the correctional phase of the cued and guided

conditions (1 and 4, respectively).

Data Collection

The experimenter recorded correct responses and the
positions of incorrect responses during both screening
and training of each subject. The forms are shown in

Appendix B.

Hypotheses

 

As was previously mentioned, the unit of analysis in
this study was the number of trials to criterion on the
matching task. Although the research does not clearly
support any direction for a prediction, it was the writer's
hunch that verbalization would not facilitate the per-
formance of SMI individuals. Therefore, it was hypothe-
sized that:

1. There will be a differential effect for training
procedures on student performance.

In the case of normal young children, the attainment

of the match-to-sample tasks is developmental with color

24
easiest and form hardest. Therefore, a further prediction
was:
2. There will be a differential effect for tasks
on student performance.
3. There will be no interaction between tasks and

treatments .

Statistical Analysis

 

A Hodges-Lehman (1962) two-way analysis of variance
for blocked data was used to test the hypotheses. The

alpha level for significance was set at p<.05.

CHAPTER III
RESULTS AND DISCUSSION

The primary hypothesis of this study was that there
would be a differential effect for treatments. Although
the literature did not support firm predictions with an
SMI population, it was the writer's hunch that students
in the cued and guided conditions (Treatments 1 and 4,
respectively) would achieve criterion in fewer trials
than would students in the verbal directions and verbal
information groups (Treatments 2 and 3). As was previously
explained, training under all conditions was continued
until the student met the criterion of eight correct
responses out of ten trials, or a maximum of 400 trials.

Twenty-four subjects were trained under four condi-
tions. One would have expected that the majority would
reach criterion or that a clear trend would be evident.
As will soon be observed by the reader, the training
sessions resulted in fewer than half of the subjects
achieving criterion and these were distributed relatively
evenly across treatment conditions. In all, nine of the
twenty-four subjects met criterion; five did so within the
first training session, or fewer than forty trials. The
remaining four students required from 140 to 305 trials.
The mean number of trials to criterion was 108.7 with a

25

26
standard deviation of 110.6 and a range of 10 to 305 for
those who achieved criterion.
Table 3 shows the number of trials by treatment and
task for those subjects who met criterion. As can be seen,

at least two subjects learned under each treatment and

three subjects learned each of the tasks.

TABLE 3
TRIALS TO CRITERION BY TREATMENT AND TASK

 

 

 

 

 

TREATMENTS
Cued—G ’V”Direct V Info *Guidedfi_
TASKS 2 3 4
Color X X X 150 X X 36 18
Picture 29 10 X X X 10 X X
Form X X 269 305 151 X X‘ X

 

 

 

 

 

 

X designates subjects who did not meet criterion.

By inspection of the actual number of trials, it appears
that the cued (Treatment 1) and guided (Treatment 4) con-
ditions were somewhat more facilitative and that the picture

and color tasks were easier than form.

Hypotheses Tests
The Hodges-Lehman is a non-parametric two-way analysis
of variance for blocked data which tests the effects in

both directions and the interaction, providing a single

27
computed value. If significance were found, simultaneous
confidence intervals would be used to locate the source of
differences among the distributions by pair-comparisons in
the mean ranks. Since so few subjects achieved criterion,
the chances of finding support for the hypotheses were

decreased.

Hypothesis 1

 

There will be a Treatment main effect favoring either
the cued or guided treatment over the verbal direction or
verbal information treatment.

The difference in the mean ranks between the four

groups was not statistically significant (HL=.030, df 3).

Hypothesis 2

There will be a Task main effect with picture and
form learned more easily than color.

The statistic of .030 was not significant for the
difference in mean ranks between the three tasks. Thus

the data did not support the hypothesis.

Hypothesis 3

 

There will not be a Treatment by Task interaction.
The data supported the hypothesis; however, the
failure to find main effects weakens the support of this

hypothesis.

Findings
Means, standard deviations, and ranges for the three

tasks are displayed in Table 4. These block means were

28
TABLE 4

MEANS, STANDARD DEVIATIONS, AND RANGES
FOR TRIALS TO CRITERION BY TASKS

 

 

TASK X SD Range

Color 275.5 165.8 18-400
Picture 256.1 185.8 10-400
Form 340.6 86.6 151-400

 

used to align the data prior to assigning ranks for compu-
tation of the Hodges-Lehman test. Eight subjects are rep-
resented in each block. Three subjects achieved criterion
in each block, so the mean includes five students at the
arbitrary ceiling of 400 trials.

In Table 5 the ranks of those who met criterion are

 

 

 

 

 

displayed. Subjects who did not meet criterion in 400
TABLE 5
RANKS OF THOSE WHO MET CRITERION
TREATMENTS
TASKS l 2 3 4
Color ___a "' "- 5
--- 6 --- 3
Picture 4 '-' -‘- ---
1.5 --- 1.5 ---
--- 3 7 ---
Form ___ . 9 -__ ___
X 12.25 12.33 12.75 12.67
aThose who did not meet criterion have a tied rank

of 17.

29
trials had a tied rank of 17. Because the absolute value
of number of trials required to achieve eight correct
responses out of ten trials had the arbitrary ceiling, the
data were analyzed by ranks.

It will be recalled that inattention was defined as
the failure of a subject to visually attend to the
materials during the demonstration of correct placement
in each trial. The tally of occurrences of inattention
apparently was too inexact to record the behavior it was
intended to measure. Totals ranged from zero to 122 out
of a maximum of 400, if the subject were inattentive for
every trial. Passive inattention, such as glassy-eyed
staring, was not recordable on this measure. In addition,
it was the observation of the writer that there were other
points during the entire sequence of a trial in which
visual attending was crucial for successful task comple-
tion. Not only was it important that the subject observe
the demonstration, but also that he look at the materials
to make his choice and place the cube. Some of the sub-
jects were particularly prone to maintain eye contact with
the trainer rather than to attend to the materials during
this choice phase. For students in the cued and guided
conditions, inattention during the correctional phase
would, in effect, negate the treatment condition. There-
fore, these data were not meaningful and were not included.

Fewer than half of the subjects achieved criterion
and these were relatively evenly distributed across treat-

ments, accounting for the non-significant results. In an

30
effort to better understand why the findings were incon-
clusive and to identify trends, the data were scrutinized
for commonalities among subjects.

In Table 6 it can be observed that the majority of
subjects who did not meet criterion did not improve their
performance across the ten training sessions. Only Subject
15 showed such a trend. In fact, many declined in their
percentage of correct responses from the first to the
tenth session, although only one subject did not perform
better than screening at any time. These data are more
fully presented in Appendix C.

Since there was a lot of variation in a subject's
performance both within and across training sessions, the
highest percentage in a given session was selected as the
best effort and is displayed in Table 6. The procedure
of utilizing the highest percentage of correct responses
was consistent with the identification of those who
achieved criterion. What is evident is the lack of con-
sistency in performance across sessions. For many subjects,
marked variation occurred within training sessions as well.
In effect, achieving criterion was not necessarily synony-
mous with mastery of the task. If the performance of the
15 subjects who did not meet criterion is indicative, then
the arbitrary nature of the criterion is apparent.

There may be a point where maximum benefit occurs.
Fourteen of the twenty-four subjects doubled the percentage
of correct responses during the first training session over

the percentage attained in screening. Another four subjects

31

TABLE 6
HIGHEST PERCENTAGE ACHIEVED PER SESSION

 

 

 

 

Session
Prer 1 2 3’ 4 5 6 7T7 8 9 10
Subject
Day Center
1 30 60 50 60 60 30 20 30 30 40 40
2 25 70 40 30 30 40 40 40 40 30 30
3 25 80 -- -- -- -- -- -- -- -- --
S 20 20 40 10 20 40 40 30 20 20 20
6 30 30 20 10 10 10 30 30 20 0 10
7 15 50 50 60 50 60 40 40 70 40 40
9 25 60 40 40 30 30 30 40 30 30 30
10 25 70 50 50 30 20 30 30 20 20 30
12 20 40 40 10 20 40 10 20 80 -- --
13 25 30 40 40 50 40 50 60 50 40 40
14 25 40 50 30 30 30 30 30 40 20 40
15 20 30 30 30 30 20 30 30 40 60 60
17 25 50 50 50 80 -- -- -- -- -- --
19 30 80 -- -- -- -- -- -- -- -- --
21 25 40 50 20 30 30 30 40 40 40 50
23 25 40 50 20 30 30 30 40 40 40 50
Institution
4 25 90 -- -- -- -- -- -- -- -- --
8 25 40 30 60 80 -- -- -- -- -- --
ll 30 60 70 70 60 70 70 90 -- -- --
16 25 90 -- -- -- -- -- -- -- -- --
18 25 40 30 50 40 40 40 40 40 30 30
20 25 90 -- -- -- -- -- -- -- -- --
22 20 30 40 40 30 10 20 20 30 30 20
24 20 30 10 30 30 40 50 40 20 20 40
a
Percentage correct on the training task during

screening.

32
did so in the second training session. For thirteen of
these eighteen subjects, the highest percentage achieved
across trials occurred in the first or second training
session. There did not appear to be any pattern relating
to treatment or to task.
The distribution of the subjects' ages by treatment

is shown in Table 7. Essentially the distribution was

TABLE 7
DISTRIBUTION OF AGES BY TREATMENT

 

Treat- Chronolo ical A e
ment 8 9 I0 II I2 I3 I4 I5 I6 I7 I8 I9 20 2I 22 23 24 25

 

 

 

 

1 2 1 1 1 1*
t a
2 1. 1 1 1 1* 1
a a
3 1 1 1 1 1 1
a a
4 1 I 1 1 1 1

 

*
Those who ach1eved cr1ter1on.

bimodal with ten subjects younger than twelve years and
thirteen subjects older than sixteen years. Those achiev-
ing criterion were dispersed across the age range (8-4 to
25). Age did not appear to be a relevant variable.

A much higher percentage of institutional residents

met criterion than did subjects from day training settings,

33
as can be seen in Table 8. Subjects were not stratified

by residence prior to assignment to treatment groups.

TABLE 8
PERCENTAGE OF THOSE WHO MET CRITERION BY RESIDENCE

 

 

 

. TREATMENTS
RESIDENCE If 2 3 4 %
Day Training 1/5 1/4 1/4 1/3 .25
Institution 1/1 2/2 1/2 1/3 .62

 

Of interest is the fact that three of the five institutional
residents who met criterion did so within the first forty
trials. Overall the individuals in the institution tended
to have considerably less educational experience relative

to their ages.

Discussion

 

There is no obvious conclusion to be drawn from these
data, since so few students achieved criterion and those
who did were rather equally distributed across treatments.
These data do not permit conclusions in either direction,
pro or con, with regard to the effects of the four training
procedures. We are not in the position of saying that any
treatment facilitated learning, nor can we conclude that
it does not make a difference what method of instructional
presentation is used with an SMI population. For those

students who did achieve criterion, it appeared by simple

34
inspection of the number of trials that cued and guided
conditions took fewer trials and that color and picture
were easier than form.

While we do not have data to fully answer the questions
which this study sought to answer, there are a number of
issues which warrant discussion. If this study is any
indication, experimental research with SMI subjects is
fraught with problems. Clearly those students who are
being programmed as SMI are not homogeneous with regard
to learning. Numerous factors impinge on a placement
decision such as behavioral problems, attending difficul-
ties, lack of other apprOpriate program, etc. In an
attempt to understand the findings, several possibilities
were considered. We began by looking for commonalities
among those who achieved criterion. The discussion which
follows will touch on these topics: subjects, position
preference, visual attending, age, residence, and

mediation.

Subjects

One of the initial difficulties experienced in this
study was finding subjects within the specified parameters
and at an appropriate functional skill level (i.e., had
a concept of matching identical objects, but could not
yet match forms at a representational level). It will be
recalled that the three match-to-sample tasks from the
Leiter Scale were at the Year 11 and Year III levels. With

normal children attainment of these tasks would be expected

35
to occur over the developmental span of about twelve months.
In the case of this population of SMI individuals, who by
definition have developmental disabilities primarily in
the cognitive domain, this develOpmental span should
logically be longer; however, the segment appeared to be
much narrower for these subjects. Essentially, we encoun-
tered an all or none phenomenon. Those students who could
consistently match at the object level could perform the
three matching tasks utilized in this research and were
screened out. By and large, the subjects who were screened
in did not perform above chance on any of the tasks. It
would appear, based on this sample, that the developmental
sequence for matching may not hold for SMI individuals.
In the planning stages, consideration was given to adding
tasks such as matching of number quantities (using those
described by Filler and Bricker, 1976). In field testing
of these materials with students in an SMI setting, it was
obvious that the matching of number quantities was a giant
leap cognitively from the identical matching of color,
picture, or form. Matching of number quantities required
counting and was not simply a direct matching, although
it could conceivably be achieved by configuration. The
net effect of tapping a narrow segment of the population
seemed to be a bimodal distribution of subjects: younger
individuals who might be at a readiness stage for a concept
of matching identities and older individuals who for some
reason have not attained the concept. Those in the middle

were screened out because they were able to do the three

36
match-to-sample tasks. Possibly the older subjects were an
artifact of amount of educational experience or the fairly
recent shift in programmatic focus for these students.
This characteristic was not unique to the subjects used
in this study; the observation held for those students who
were screened out as well as for those who met the criterion
for inclusion in all of the centers. It is not clear why

finding subjects was so difficult.

Position Preference

The reader will recall that during the screening of
subjects and the pilot study, a position preference was
exhibited by the majority of students who were potential
. subjects. A student was considered to have a position
preference if he failed to choose all four locations with
equal frequency, 25 percent of the time. The choice of
any one position 40 percent or more of the time or a
combination of the two positions 65 percent of the time
during any single training session (forty trials) was
defined as a position preference in the present study.
All subjects demonstrated a position preference on the
tasks during screening. The preferred position was not
the same location across training sessions for most students,
but a position preference held within a single training
session. With the exception of a brief notation in Bricker
and Bricker (1971) that position preferences are a frequent

problem with low functioning individuals, there was no

37
indication in the literature to suggest a problem of the
magnitude encountered here.

Based on the experience in the pilot study and the
dearth of literature on this stereotypic responding, the
decision was made to proceed with the study, even though
it was unlikely that the questions could be fully answered.
It seemed reasonable to assume that a position preference
would not be altered without some type of intervention or
correctional procedure; consequently, Treatment 1 was
modified from a demonstration only condition to a visual
cueing roughly analogous to the physical guidance in
Treatment 4, but without touching the student. Interestingly,
many of the subjects could match three or even four identical
objects, a task with which they were familiar; so it would
not appear that a position preference is something that
students have been taught. Rather, when confronted with
the unfamiliar materials utilized in this study, they
failed to make any discriminations. What appeared to be
occurring was an inflexibility in approaching a new task.

It makes one wonder about the students who were screened

out because they could do the tasks. Even given that instruc-
tion in the school setting has the potential for a more task
analytic approach than was possible within the confines of
this study, one wonders how students who could do match-to-
sample tasks learned to do so, especially in view of the

fact that most of the subjects did not show improved per-

formance over time.

38

Might there have been some selective factor that
related to developmental phase, educational experience,
rigidity in thinking, learning style, or some other
variable? Obviously, some gaps exist in our ability to
assess where these students are in terms of readiness for
acquiring a new skill or concept. There are more bits of
information that are not known than are known about SMI
individuals. The major checkpoints from normal deve10pment
give a framework, but the steps in between have not all
been charted. Perhaps a position preference is a develop-
mental phase that a student goes through which is pre-
requisite to being able to respond to a four-choice array
in a manner similar to that of an infant holding objects
in both hands before being able to transfer hand to hand.
Might those who were screened out differ from the subjects
in their style of learning or their approach to problem
solving? Zigler (1973) identified two contrasting learning
styles, one characterized by reliance on concrete situa-
tional cues (outer-directedness) and the other characterized
by active attempts to deal with abstract relations among
elements (inner-directedness). He postulated that the high
incidence of failure experienced by retarded individuals
generates a sensitivity to external cues. Rosenberg (1968,
p. 22) defines learning style as "an individual's charac-
teristic pattern of behavior when confronted with a
problem." Of the four styles he identified, the "rigid-
inhibited" approach in which the individual is closed to

both external and internal sources of information may have

39
applicability to this group of subjects. He sees this
learner as not developed beyond the concrete level of
managing information in a problem solving situation.
Perhaps for the subjects who did not meet criterion, the
representational level of pictured materials was too
complex. Other factors which might have impinged are
such things as amount of educational experience, flexi-
bility in accepting changes in routines, or adaptability
in working with an unfamiliar adult. It will be recalled
that potential subjects in eight educational programs were
screened for possible inclusion. It was the writer's
impression that students in two of these programs were
less perturbed by the process and seemed to relate more
easily. These two centers are in close proximity to teacher
training programs which use them extensively for field
placement, practicum, and student teachers. One possibility
is that students in these centers tend to take such changes
in stride. The opposite extreme was evident with institu-
tional residents, many of whom exhibited resistant behavior
such as balking at leaving their classrooms or refusing to

cooperate in the screening.

Visual Attending

Fundamental to any type of learning is attending.
Klausmeier (1974, p. 29) defines attending as "orienting
one's sensory organs so that external stimuli may be
received." His model postulates that attending along with

discriminating and remembering are involved in motor as

40
well as perceptual experiencing. For an individual to be
able to discriminate between several stimuli, he must
attend to the features of an object or picture which dif-
ferentiate it from others. The distinctive features to
which one attends vary as a function of age. In the case
of low functioning individuals, the literature suggests
that inattention may be a factor. In visual discrimination
learning, Zeaman and House (1963) postulated that the
retarded subject does not focus on the relevant dimensions
of a stimulus display. Further, Routh (1973) cites
Drotar's work, which found that retardates may be deficient
in their ability to disengage or withdraw their attention
from irrelevant stimuli. We do not have data to bring
to bear on this issue. Those students who obviously did
not focus their attention on the materials during screening
were excluded from the study, since the design did not
include the shaping of visual attending. As presented in
the findings section, the inattention measure which was
kept on the subjects did not differentiate in any meaning-
ful way. The experience of those who instruct SMI indi-
viduals is that visual inattention is a significant impedence
in the learning process. This issue needs to be studied

systematically.

Age
Age did not appear to be a relevant variable for those

subjects who met criterion. Of interest is the fact that

the age distribution was bimodal (see Tables 1 and 7) and

41
those of ages twelve to sixteen were minimally represented
in this study. One possible explanation would be that
as students receive appropriate programming at younger
ages, it is conceivable that the potential for skill acqui-
sition is increased. Perhaps age, as such, is not so much
the factor as is experience, specific training, or some

other variable.

Residence

 

In Table 8 the higher percentage of institutional
residents who achieved criterion was displayed. These
individuals had been in a structured educational setting
for one year prior to the time the study was conducted.
Previously, programming was done on the ward in the insti-
tution and focused primarily on self-care skills. That
these students were screened in may have related to a lack
of experience in test taking. Once the task was demon-
strated during training, then the student understood what
was expected. In effect, two of these subjects achieved
criterion under demonstration only, since ten trials were
all that were necessary. In the case of day training, for
the most part these students have been in structured pro-
grams for a longer period of time, even though the focus
has shifted in the past few years. Another way of looking
at this phenomenon is to consider that under Zigler's
(1973) hypothesis the critical variable in retardate learn-

ing is failure experiences; one would expect the

42
non-institutionalized to have the greatest sensitivity to

external cues.

Mediation

 

Operationally both attending and short term memory
require similar behaviors on the part of the learner. In
a further attempt to understand these results, in terms of
possible strategies which could be used for successful
performance, the ways in which a student might complete
the tasks were considered. It will be recalled that in
all of the instructional conditions each trial was demon-
strated for the subjects. Three possible ways that a
student might complete the task were identified. First,
he could scan the choices and make the apprOpriate visual
discrimination. Second, it is possible that he could form
an image of the spatial array as a whole and remember the
location of the placement which was demonstrated. Third,
he could represent the object internally, either verbally
or nOn-verbally, and recognize its match.

The viewpoint of verbal mediation theorists suggests
that labeling responses are necessary; therefore, the
young child and the retarded individual are limited to
non-mediated stimulus control. It is not clear to what
extent a student's language develOpment may have been
related to successful task completion in this study. No
assessment was made of the language development of the
subjects, since it was not deemed necessary for task

performance; however, it may have been a facilitative

43
factor. As an after the fact observation, five of the
subjects who met criterion had some appropriate speech.
Of particular interest is the fact that three of the five
were in the verbal direction condition. As was shown in
Table 3, those subjects achieved criterion in 150 to 305
trials. The issue of language develOpment may be particu-
larly relevant as it relates to verbal mediation and
memory strategies.

In the case of preverbal children, the mediation may
not be verbal but perceptual. Piaget has demonstrated
that cognitive activity occurs during the sensorimotor
period prior to the advent of speech. A strong case
against the equivalence of thinking and verbalization has
been advanced by Furth (1966) using the example of the
deaf. Probably the developmentally earliest means by
which information is represented is through imagery. It
seems clear that images are not verbal, but do have func-
tional characteristics. In a study with children ages
three and one-half to seven, Corsini (1969) found that
memory of younger children does not differ from that of
older children when the response is nonverbal and nonverbal
cues are used in retention. He postulated that successful
performance of a direction following task was probably
achieved through forming a more or less external represen-
tation of the spatial array through the use of sensory-
motor cues. This interpretation would seem to lend support
to the second possible strategy, remembering the location

of placement. The conditions used in this experiment do

44
not provide the basis for resolving this issue; however,

future studies might pursue the mediation aspect.

CHAPTER IV
SUMMARY AND RECOMMENDATIONS

Because of the expanded definition of who is to be
educated, including the severely handicapped, and the
changing emphasis in programming for the severely mentally
retarded from care to education, issues of the "what" of
curriculum and the "how" of instruction need to be
addressed. The major responsibility for educating these
individuals rests with classroom teachers, many of whom
have been trained to work with higher functioning
retardates.

One mode of instructional presentation which has
demonstrated results with young normal children and
moderately retarded individuals is vicarious learning,
derived from Bandura's (1971) social learning theory. The
effects of adjunctive verbal coding are equivocal, appar-
ently as a function of the conceptual level of the subjects
and the complexity of the tasks.

The purpose of this study was to compare the effects
of four training procedures, all involving vicarious
learning, on the match-to-sample performance of severely
mentally impaired persons. Specifically, this investigation
sought to determine whether correctional procedures or
verbalization would facilitate the learning of these

45

46
students. The basic paradigm for the presentation methods
was drawn from studies by Forehand and Yoder (1973, 1974),
who explored the effects of modeling and verbal cues on
concept acquisition. Research has clearly demonstrated
that modeling is superior to verbal description for sub-
jects who lack verbal facility (Rosenthal and Zimmerman,
1972; Rosenthal and Kellogg, 1973). Yoder and Forehand
(1974) utilized two levels of verbalization concurrently
with modeling: verbal description and verbal concept or
rule. The major extension which occurred in the present
research was the use of two correctional procedures.

Match-to-sample tasks from the Leiter Internatidnal
Performance Scale have been used by other researchers
(Forehand and Yoder, 1973; Filler and Bricker, 1976) with
educable retardates. In order to have this study relate
to other research, three tasks from the Leiter Scale
involving matching of identities were used.

This investigation was designed as an experimental
study. Twenty-four students who were in educational pro-
grams for SMI were the subjects of this study. Subjects
were screened on the three match-to-sample tasks and
randomly assigned to one of four treatment groups. All
training procedures involved demonstrating the desired
response for each trial. Treatments 1 and 4 (cued and
guided, respectively) included a correctional phase if the
student did not make a correct response. The cued condi-
tion utilized the visual cue of positioning the cube

directly in front of the appropriate slot and pointing to

47
the relevant response. The guided condition involved
physically manipulating the student to make the correct
response. In the verbal direction condition, the trainer
verbalized the descriptive phrase, "Put the block here",
concurrent with demonstrating the correct placement. In
the verbal information condition, the trainer labeled the
relevant attribute (e.g., "Put the red with the red")
concurrent with demonstrating the correct placement.

Each subject was trained on one match-to-sample task
for a maximum of ten sessions (400 trials) or until
achieving criterion of eight correct responses out of ten
trials. The ten minute training sessions were administered
twice a day on five consecutive days. The independent
variables were tasks and treatments. Number of trials to
criterion was the dependent variable. The data were
analyzed by ranks using the Hodges-Lehman two—way analysis
of variance for blocked data.

It was the main hypothesis of this study that the
correctional procedures would facilitate the performance
of SMI students over the verbal conditions. It was further
hypothesized that the picture task would be learned more
easily than color or form. The results were inconclusive;
however, by inspection of the actual number of trials to
criterion, it appeared that the cued and guided conditions
were somewhat more facilitative. Picture and color tasks
appeared to be easier than form. Nine of the twenty-four
subjects met criterion, and these were relatively evenly

distributed across treatment conditions, accounting for

48
the non-significant results. These data do not permit con-
clusions in either direction with regard to the effects of

the four instructional methods.

Recommendations

 

Modification of the Present Study

 

One possible approach to overcoming some of the
problems encountered, specifically the relatively few
students who met criterion and the stereotypic responding
of position preference, might be to use a training sequence
under all conditions which would move the subject through
various phases. The idea would be to systematically maxi-
mize the probability of a correct match in some way such
as by limiting the placement choices available in a manner
similar to that used by Filler (1976) as one modification
of maternal teaching style. A format of easy-to-hard
training sequences like that used by Irwin (1976) in
training a discrimination task may be applicable.

Consideration could be given to match-to-sample tasks
which might tap a broader range of subjects. It was not
clear with subjects who did not meet criterion whether the
representational level of the tasks was a relevant factor.
A possible approach would be to use concrete objects such
as toys, miniatures, or geometric blocks for identical
matching at a more concrete level.

In view of the number of problems encountered in this
experimental study and the variables which are difficult to

control with an SMI population, perhaps a more fruitful

49
method of study would be a case study approach or longi-

tudinal study in the naturalistic setting.

Additional Research

A number of issues were raised in conducting this
study which have implications for future research. Addi-
tional research is necessary to explore the following
questions and topics presented in the discussion. They
are listed here in the order in which the background and
issues were discussed.

1. Does verbalization by the instructor facilitate
the learning of SMI individuals? Might a silent demon-
stration be as effective as modeling with concurrent verbal
cues? Does verbalization serve to elicit a student's
attention for learning?

2. Are there meaningful ways by which SMI students
might be grouped which would create more homogeneity in
terms of learning?

3. Does the developmental sequence for SMI students
approximate that for "normal" children on matching tasks?

4. How did those students who can perform match-to-
sample tasks learn to do so?

5. Are there ways to approach a position preference
which might encourage an SMI student to use his past
experience in approaching new problem-solving situations?
To what extent is a position preference related to inability
to generalize or transfer prior learning? To what extent

is a position preference a develOpmental phase on the part

50
of the learner? Might a position preference be a reflec-
tion of learning style for some SMI individuals? To what
extent does a position preference relate to the represen-
tational level of the materials or the functional level of
the subjects?

6. Might visual attending be a critical variable in
imitative learning? To what extent might auditory proces-
sing as an adjunct to visual processing facilitate attend-
ing? What is the relationship between attending and
various methods of instruction?

7. What is the relationship between a student's
language development and methods of instructional
presentation?

8. If mediation can be either imagery or verbal,
what are the implications for teaching these students?

It would appear that what is known about the learning
of trainable and educable retardates does not apply directly
to the severely and profoundly mentally retarded. Only
further research efforts will allow a clearer delineation
of whether various methods of instruction facilitate or

impede task performance with this population.

REFERENCES

REFERENCES

Bandura, A. Social Learning Theory. N.J.: General
Learning Press, 1971.

Bandura, A. Vicarious processes. In L. Berkowitz (Ed.)
Advances in Experimental Social Psychology, Vol.
2. N.Y.: Academic Press, 1965, pp. 1-55.

Bricker, W. and Bricker, D. Receptive vocabulary as a
factor in the discrimination performance of low
functioning children. American J. of Mental
Deficiency, 1971, 75, 599-605.

Choe, K. Effect of instruction on multiple classifica-
tion skills of young mentally retarded children.
Unpublished doctoral dissertation, Columbia
University, 1974.

Corsini, D. Developmental changes in the effect of non-
verbal cues on retention. Developmental Psychology,
1969, 1, 425-435.

Cullinan, D., Kauffman, J., and LaFleur, N. Modeling:
Research with implications for special education.
J. of Special Education, 1975, 9, 209-221.

Filler, J. Mbdifying maternal teaching style: Effects
of task arrangement on the match-to-sample per-
formance of retarded preschoOl-age children.
American J. of Mental Deficiency, 1976, 80, 602-612.

Filler, J. and Bricker, W. Teaching styles of mothers and
the match-to-sample performance of their retarded
preschool-age children. American J. of Mental
Deficiency, 1976, 80, 504-511.

Fisher, M. and Zeaman, D. An attention-retention theory
of retardate discrimination learning. In N. Ellis
(Ed.) International Review of Research in Mental
Retardation, Vol. 6. N.Y.: Academic Press, 1973.

Flanders, J. P. A review of research on imitative behavior.
Psychological Bulletin, 1968, 69, 316-337.

51

52

Forehand, R. and Yoder, P. Acquisition and transfer of
conceptual learning by normals and retardates: The
effects of modeling, verbal cues, and reinforce-
ment. Behaviour Research and Therapy, 1974, 12,
199-204.

Forehand, R. and Yoder, P. The effects of modeling and
verbal cues on learning of retardates. J. of
Clinical Psychology, 1973, 29, 184-186.

Furth, H. Thinking Without Language. N.Y.: Free Press,
1966.

Hecht, P. J. Developmental checklist and training guide:
Reasoning and problem solving. Unpublished manu-
script, Michigan State University, 1975.

Hodges, J. L., Jr., and Lehman, E. Rank methods for com-
bination of independent experiments in analysis
of variance. The Annals of Mathematical Statistics,
1962, 33, 482-497.

Inhelder, B. The Diagnosis of Reasoning in the Mentally
Retarded. N.Y.: John Day Co., 1968.

Irwin, L. K. General utility of easy-to-hard discrimina-
tion training procedures with the severely retarded.
Education and Training of the Mentally Retarded,
1976, 11, 247-250.

Kessler, J. W. Contributions of the mentally retarded
toward a theory of cognitive development. In J.
Hellmuth (Ed.) Cognitive Studies, Vol. 1. N.Y.:
Brunner-Mazel, 1970, 111-209.

Klausmeier, H., Ghalata, E., and Frayer, D. Conceptual
Learning and Development: A Cognitive View. N.Y.:
Academic Press, 1974.

Landau, B. and Hagen, J. The effect of verbal cues on
concept acquisition and retention in normal and
educable mentally retarded children. Child
Development, 1974, 45, 643-650.

Leiter, R. G. Examiner's Manual for the Leiter Inter-

national Performance Scale. Chicago: Stoelting
Co., 1969.

Litrownik, A., Franzini, L., and Turner, G. Acquisition
of concepts by TMR children as a function of type
of modeling, rule generalization, and observer
gender. American J. of Mental Deficiency, 1976,
80, 620-628.

53

Luria, A. The Role of Speech in the Regulation of Normal

and Abnormal Behavior. N.Y.: Liveright Publ.,
1961.

Milgram, N. A. Cognition and language in mental retarda-
tion: Distinctions and implications. In D. Routh
(Ed.) The Experimental Psychol0gy of Mental Retarda-
tion. Chicago: Aldine, 1973, pp. 157-230.

Rattan, M. The role of language, manipulation, and demon-
stration in the acquisition, retention, and transfer
of conservation. Alberta J. of Educational Research,
1974, 20, 217-225.

Roos, P. Severely and profoundly retarded students - past
and future. In Educating the 24-Hour Retarded
Child. National training meeting on education of
the severely and profoundly mentally retarded,
March 31-April 2, 1975. National Association for
Retarded Citizens.

Rosenberg, M. Diagnostic Teaching. Seattle: Special
Child Publ., 1968.

Rosenthal, T., Alford, G., and Rasp, L. Concept attain-
ment, generalization, and retention through observa-
tion and verbal coding. J. of Experimental Child
Psychology, 1972, 13, 183-194.

Rosenthal, T. and Kellogg, J. Demonstration versus instruc-
tion in concept attainment by mental retardates.
Behaviour Research and Therapy, 1973, 11, 299-302.

Rosenthal, T., Moore, W., Dorfman, H., and Nelson, B.
Vicarious acquisition of a simple concept with
experimenter as model. Behaviour Research and
Therapy, 1971, 9, 219-227.

Rosenthal, T. and Zimmerman, B. Modeling by exemplifica-
tion and instruction in training conservation.
Developmental Psychology, 1972, 6, 392-401.

Rosenthal, T. and Zimmerman, B. Organization, observation,
and guided_practice in concept attainment and
generalization. Child Development, 1973, 44,
606-613.

Routh, D. The experimental study of the retarded child.
In D. Routh (Ed.) The Experimental Psychology of
Mental Retardation. Chicago: Aldine, 1973, pp.
l-l7.

Siegel, S. Nonparametric statistics for the behavioral
sciences. N.Y.: McGraw Hill, 1956.

54

Stephens, W. B. The severely and profoundly retarded:
Current issues and needs. Education and Training
of the Mentally Retarded, 1976, 11, 3-4.

Tawney, J., Allen, M., O'Reilly, C., Cobb, P., and
Aeschleman, S. Developing curricula for errorless
learning: A search for order in an unorderly
world. AAESPR Review, 1975, l (2).

Thompson, R. The severely handicapped: A new horizon.
Exceptional Children, 1976, 43, 141-142.

Woodward, M. and Hunt, M. Exploratory studies of early
cognitive development. British J. of Educational
Psychology, 1972, 42, 248-259.

Yoder, P. and Forehand, R. Effects of modeling and verbal
cues upon concept acquisition of nonretarded and
retarded children. American J. of Mental Deficiency,
1974, 78, 566-570.

Zeaman, D. One programmatic approach to retardation. In
D. Routh (Ed.) The Experimental Psychology of Mental
Retardation. Chicago: Aldine Publ., 1973, pp.
78-132.

Zeaman, D. and House, B. The role of attention in retardate
discrimination learning. In N. Ellis (Ed.) Handbook
of Mental Deficiency. N.Y.: McGraw Hill, 1963,
pp. 159-223.

Zigler, E. The retarded child as a whole person. In D.
Routh (Ed.) The Experimental Psychology of Mental
Retardation. Chicago: Aldine Publ., 1973, pp.
231-322.

Zimmerman, B. and Rosenthal, T. Observational learning of
rule-governed behavior by children. Psychological
Bulletin, 1974, 81, 29-42.

APPENDICES

AAAAAAAAA

Student

APPENDIX B

FORMS FOR SCREENING AND TRAINING

Preliminary Trials

 

Birthdate

 

Center

 

R 6 PS Goals:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Location-no match'
1 2 3 4 Color Task #
R G B Y C Y R B R B-Y G Y G R B 16
Picture Task
E D C W C W D E C D E WIW E D C 16
Form Task
w+uo+otawuo+o+xtn 16'

 

56

 

57

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Student Center
Task Treatment
Training Session # Order to
Location: no matEH
1 2 3 4 Location demonstrated
2 3 4 l 2 1 3 4 l 2 4 3 4 2 3 1
4 3 1 2 l 4 3 2 2 l 4 3 3 4 l 2
40
2 3 4 l 2 l 4 3 Trials
Student Center
Task Treatment
Training Session # Order to
Locatibn: no match
1 2 3 4 Location demonstrated
2 3 l 4 l 4 3 2 4 l 2 3 2 3 1 4
2 l 4 3 l 2 3 4 3 2 l 4 2 4 l 3
32144213 4°

 

 

 

 

 

 

 

 

Trials

 

 

 

APPENDIX C
GRAPHS OF INDIVIDUAL STUDENT'S PERFORMANCE

S8

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In.“ t
u VI am
8 tC9 .3
C— u V
III e r8
...JO '0
7 we.“ a
SCC . ~II .
5 q\In.m
i
s m ”m.
.0
6
.4 IIHHHHHHHHHHHIIIIIh 1.
A .w
3 ”1
..o
2 f ”a
- "
Eon .
aflao -
mm hoawm (\\u

 

 

m

4

. . . q
w. m m. m

season-om aoosuoo uo unsouom

30

20-1

1.0"l

TRIALS '

59

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

o. Pilllla mm
1 v
Ill 7
9 V “m
2e
I u"
tm8 A am.
8 C . . I
...JO 3w
101 so . 2
7 u oA M .
Qununu .
5 P. 1..
v
4 f
U A
.1.
3 y
lo
'8
2 A‘ v
1 c “H's ”
EOk 7
$0.23on W .
P Loaoo

 

 

mew m m m m m

. . u q . . q q

20
10"

concedesx aoscaoo uo unoohom

TRIALS

60

Sessions

 

6|7|8|9|101

Subject #3

Picture-Cued

CA: 9-2

V' V'
400

 

 

 

T l
320 360

vfr fi V

v

 

280

f-

 

240

 

200

v V V

 

160

 

I
120

 

"'l
80

 

”'46

 

will

 

 

Equa-
m m m. m m

concedesm aoosaoo uo unschsm

u

0
2

10'-

TRIALS

61

Sessions

 

"'I
400

 

 

 

 

 

 

 

 

 

 

 

 

O
1
.m
i
Au .
9 any r
I 4
sung um“
.0 f3
3 t e . .
Pvmsrb
o—Jt ”-8
7 b c so .2
uolA V
[In SPC ..l
6 “2
1
s “m
.m
4 v
r
.m
.11
3 I
T
..m
2 fl
f
EOE -
”DEG-03d“ v
P hoaoo
i 1 .1 d a u d A W 0
o
m m m. a w m s m. ..

concedeem aooaaoo no acooneh

TRIALS

100d

90"

a:
O
1

Percent of Correct Responses
8 ‘8 8‘ 3
l 1 l 1

U
0
l

B

IO'R

 

'5.

Color

Picture
Form

 

62

Sessions

5

 

 

6|7|a|9|m
Subject #5

Form-Cued
CA: 24-0

 

 

 

 

 

 

v

--, .
120

 

 

 

 

 

ii

 

 

 

63

Sessions

 

"‘I
400

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o r
1
III :0
9 ”x
6 . Iv
III a a.
e1 1V7 3
8 t u . e
“wrunu .
III -
.1amm .mW
.T Lu r... nlnnnnu|||||||1\\\\\1.n¢
u oA .
llllnD_F.Pu II-
.M
6 “Va 2
0|.i e
s Aim
A.
a.
4 v
.3 31.
A.
'0
2 «Ana
T T
1 m
\n
EOE .
u gfléfldh / v
P LOHOU

 

 

m

1
0
9

w

. . q n -
m. .w w m w

sssconssm essence uo assesem

TRIALS

64

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m . V .
t v
.II a To
r. ennnHHHHHHHHHH .uw
9 s1 v
VInU
u" III-D I
s9 0 .3
tV— AI“! .
8 C.7 .
Illerl 0
..JO ma
blss v
7 qu .
SCC
.2
6 v w
.m
a H
II. .m
1) sh” b ..1
\I
0
.8
2 W11 ..
EOE "
Eek—.50.; /
P .330

 

 

RI q u q .
m. m w m m

nonsense: «season we aceehsm

20"

1.0"l

TRIALS

65

Sessions

 

'

 

 

 

 

 

 

 

 

 

 

 

 

o .
1 e v
a
e W
9 r .
m w
8
'8 # U
s 8 rm
8 t v . n
C P1
III. e In 0
.30 ww
7 b 1 es .
"unusn .
.Svnunb mm
,0 n
..m
5 W
.m
4 J. H
is ..m
..1
3 up f
I0
'8
2 fl .
l w ”
EOE. v
E 0.250.: A. .
P Lenou

 

 

. . . _ -
m m w m m

sescoasem aosssoo no assessm

204

10"

TRIALS

66

Percent of Correct Responses

Sessions
93212345 6‘7'8'9'10]
1001 ‘53 Subject #9
'33; Picture-V. Direct.
°““ CA: 10-7
90-
80..
70“
60d
501
“O-
30- /\ A /\
20-
1o-
" "' "" "U " 'vr' "'—V '1 "'1 '—"I
o 65 120 160 200 amt 28d 320 :60 too

 

 

 

 

 

 

 

 

TRIALS

 

 

 

 

 

 

~a c:
8 ‘6 8‘ o o
l 1 1 l 1

Percent of Correct Responses

U
0
L

10"

ii

Picture
Form

Color

 

 

 

 

 

 

 

Sessions

5|7|a|9l1o

Subject #10

Picture-V. Direct.

CA: 9-7

 

 

 

 

 

 

 

 

 

'f'

200
TRIALS

 

 

 

 

IQB|':mo 'Sza

 

 

5

68

Sessions

2345 6|7|8|9lioi

Subject #11

Percent of Correct Responses

Oi
O
l

5 ‘3
1‘ I

U
0
L

10"

Picture

Color
Form

 

 

 

 

 

Form-V. Direct.

CA:

13-8

I

 

 

 

fivv '

 

 

 

 

 

 

TRIALS

 

. ém'.

éeo

 

 

 

v '1

400

Percent of Correct Responses

8

.g
0
l

Os
0
l

‘8

8
1

V
O
l

 

'F‘é

Color

Picture
Form

 

 

69

Sessions

5

 

‘ l

7|a|9 ‘10

Subject #12

Form-V. Direct.
CA: 8-3

 

 

 

 

 

 

 

 

 

 

 

20-
" F I F
o " "66 7150 1'6'0 W266 'a'el"2'e'o"53o|"e6

 

70

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o \ V.“
1 .I‘I‘I‘V .
II o. I
9 3..“ A ”m
1.1 .
III-I'M" II .-
tV3 - .3
8 c. - qI||I.‘|“.‘IVI w
I 3 r8
.1Jnv unu
101 so IV V”
7 qu I .
QuPunb III .
,o H mm
5 V a
v a.
‘ v
_ v
I0
x4 HMMWWHH ".6.
v
W
20% r
Goad—fined“ v

 

 

90"

q

d u a - u d
m. .w m. m m m.

neoconsem aoeasoo uo aceossk

10"l

TRIALS

Percent of Correct Responses

71

 

 

 

 

Sessions

PRE 1 2 s 4 5 6 I 7 l a 9 I10
moq 3% Subject #14

:35 Color-V. Info.

0mm CA3 18-10
90"
8°-
70"
60--
501 1
“°‘ I I II
304 A
2.. IL - I.
10-

" fi' ‘*"I T"! 'r'T 1"! r"s "'r vva r
o 80 120 160 200 240 280 320 360

 

 

 

 

 

 

 

 

 

TRIALS

 

 

 

 

 

Percent of Correct Responses

72

Sessions
Par: 1 2 3 4 5 6 l 7 8 i 9 I10
1001 33:3: Subject #15
.33?) Picture-V. Info.
09-- CA: 9-5
901
80-

4
O
I

8‘

‘8

 

5
J
K1
1

 

 

 

 

 

 

 

 

 

 

304
20- V.
10-

" "'I T" 'r' -" .-r' ---1 "‘I ---; ---y w.
o 40 86 12's 16'0 200 240i 280 320 360 400

 

 

 

 

 

TRIALS

73

Percent of Correct Responses

 

 

 

 

 

 

Sessions
FRI-S 1 2 3 4 5 6 l 7 8 I9 I10
0
100.. 38 Subject #16
33‘: Picture-V. Info.
”H“ CA: 17-2
90-1
80..
70'I
60-
501
40-
304
20-1
10.
" "'I f" "'l "*l *"I fi'l "'I ---| v-va-
0 4086120160200240I280320360A00

 

 

 

 

 

 

 

TRIALS

74

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

m I
o T
o m
o, nip: .
1 n a
luv-I.
as was
t O o v
8 CV6 .
.Jm um
b r : .
or nununn .
Quflahb .
5 h?
f
:4 ”mm
IO
.M
t. [I] v
J .m
3 .
...w
2 .
'
EOE v
Egaoum .
P LOHOU

 

 

1004

u q 1 a H J

wmmwmm

neoconsex aesanoo we aeseusm

TRIALS

7S

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m V 1
9 0 V ”m.
oopI .

Illlln
unvl 1
2 v
8 t .. ~./\e‘e .3
CV3
Illluncmmea u
.14 .
7 Sum V .2
Run.nu fin
,0 HII.III.HH.....H.UV_e “as
e
5 V Hm
MI a
.4 .1.
..w
3 0/“ ”1.
l0
2 "8
L r
1
atom .
fleeces: “\s .
p. Loaoc
q . q a . q q. . . m. _0
m m m m. m m m m m ..

seseoassm essence no aceoesm

TRIALS

 

 

 

76

 

 

Sessions
PBS 2 3 4 5 6 7 8 9 '10 I
o n
100. 3%; Subject #19
.331: Color-Gu1ded
”A“ CA: 10-5
90"
80-
O
8
:70-
O
D.
O
0
“‘60-
d
O
O
K.
3..
850‘
H
O
240-
O
O
‘2
O
0430.
20.4
10-
T' "'r "'I "U "fr 'r "‘I ---I T-vr -..1
o 80 120160200 240' 280 320360 400

 

 

 

 

 

 

 

 

 

 

a)
O
I

Percent of Correct Responses
4
8 S 8 o
l I 1 J

U
0
L

10"

00
Picture 33

Color
Form

 

77

Sessions

7|8l9
Subject #20

Color-Guided
CA: 16-3

 

'10

 

 

 

 

 

'v1'

160

 

 

V'V' f'v" fi“

240' zeq 320 360

 

 

 

 

78

Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11 r
I .w r
a .m
9 1.1. / u
If...
u" 1
. 1 ..MW
.6 tea». .14
In m7 .
1
..Jt um
7 b C so .
uslA v
RuDinh
6 A\5 “2
v
F
\_ em
5 f
.4. A r
3 //va
fl
III '0
2 HUIIHS
A; .
HHHIlf
ED“ .
abhflafldm * v
P LOHOU
1 1 q q q _ q q A u 0
m m m m m ..o. m m m. m

sescodsex «season no essence

TRIALS

Sessions 1
rat 2 3 4 5 6 7 a 9 I10
100- a; Subject #22
§§§ Picture- Guided
”M" CA: 22-4
90-
so-
I
8
r570d
O
D.
O
O
“‘60-
e3
O
O
t.
L
850‘
H
O
240- r
0
O
h
0
8‘ 30- If
201 I [-
10" D e—o-e—e
' "' *"I "*I*"I *r's "'vrvu—---' v~~.
o' 80120160200240'28032qa60I400

 

 

 

79

 

 

 

 

 

 

 

TRIALS

 

 

 

 

 

Percent of Correct Responses

80

Sessions
m: 1 2 3 A s 6 l 7 I a l 9 I10 1
100- 3}? Subject #23
'53.; Form- Guided
M" CA: 24-2
90-1
80..

O\ '4
O O
l l

8
44

8
j

 

u
0
l

 

 

 

 

 

 

 

 

 

 

 

 

' ' ' " " *"I*"I * 'r "'—r 1" 'fiv'
o to so 120 160' zoo no. 28d 3 360 400
rams

 

 

 

 

 

 

 

81

Sessions

 

10

#00

"if

 

‘T

 

 

Mv

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m

9 4d H
fiL_e
+.Munv um”

_

.o Mwnuno fi
mm - .m
SFC v

5 I” .m

JIfi

s hm

fllllllHHHflllqmm

“ 0/ r

.m

:2 1.

[I v

*0

2 "8
r

Nam

1 v
1.

Eu... .

£92535 4‘. .

P honoo
. . . q . . q \d a . O
m w m m m ..o. m m, m m

concoooom aooouoo uo unoohom

TRIALS

"TIWITIEIMLEMIETHfllflujflfiflflmfiflflfiI‘I‘ITI'“