DOCUMENT RESUME
ED 248 641
AUTHOR
TITLE
EC 170 419
Zucker, Stanley H.; Prehm, Herbert J.
Parameters of Cumulative Programming with
Severely/Profoundly Handicapped Pupils. Final
Report.
Arizona State Univ., Tempe, Dept. of Special
Education.
Department of Education, Washington, DC.
Jan 84
G008001872
51p.
Reports - Research/Technical (143)
MF01/PC03 Plus Postage.
Elementary Secondary Education; Models; *Severe
Disabilities; ^Teaching Methods
^Cumulative Programing
INSTITUTION
SPONS AGENCY
PUB DATE
GRANT
NOTE
PUB TYPE
EDRS PRICE
DESCRIPTORS
IDENTIFIERS
ABSTRACT
_ A project incorporating. .15 stjidies _o n
cumulative programming instructional strategies in severely
handicapped populations is presented. A model was developed to allow
for controlled formal investigation of potentially relevant variables
including those descriptive of the subject (such as handicapping
condition, race, or age), the content (areas Of skill or knowledge),
and the task (variables related to the actual instructional process).
The model allows for design of new studies based on the results of
previous ones. Fifteen studies on various aspects of subject,
content, and task variables are then reviewed and data are presented
in table form. (CL)
************************************************
* Reproductions supplied by EDRS are the best that can be made *
* from the original document. *
***********************************************************************
ERIC
0'
CO
o
FINAL REPORT
PARAMETERS OF CUMULATIVE PROGRAMMING WITH
SEVERELY/PROFOUNDLY HANDICAPPED PUPILS
U.8. DEPARTMENT OF EDUCATION
NATIONAL INSTITUTE OF EDUCATION
EDUCATIONAL RESOURCES INFORMATION
CENTER (ERIC)
This document has been reproduced as
received from the person or organization
originating it
Minor changes have been made to improve
reproduction q jalily
• Points of view or opinions stated in this docu
mem do not necessarily represent official NIE
position or policy.
STANLEY H. ZUCKER
HERBERT J.; PREHM
DEPARTMENT OF SPECIAL EDUCATION
ARIZONA STATE UNIVERSITY
■ . " ■ '/
GRANT NUMBER - G008001872
CFDA NUMBER - 13.443C
JANUARY 1984
er|c
2
2
Table of Contents
Page
Objectives and Need for this Assistance 3
Goal and Objective 12
Results or Benefits Expected 16 •
Approach . 17
Plan of Action 17
. Subjects 18
Procedure 18
Analysis 21
Projection of Accomplishments 22
Accomplishments * 26
Study 1 *••*.* 26
Study 2 26
Study 3 29
Study 4 29
Study 5 ' 29
Study 6 . . 31
Study 7 . . ' 31
Study 8 31
Study 9 34
Study 10 . 34
Study 11 37
Study 12 37
Study 13 37
Study 14 40
Study 15 40
Dissemination 43
References 45
Appendix A - Data Sheets 49
J
3
OBJECTIVES AND NEED FOR THIS ASSISTANCE
Special education is currently facing a dilemma brought about,
in part, by legislative mandates to provide services to the severely/
profoundly handicapped (SPH), while much of the information needed
to provide these services is incomplete. Historically, very few
public schools or institutions have providedmuch more than custodial
care for the severely/profoundly retarded, multihandicapped population.
In fact, the provision of publicly supported special education services '
to the mentally retarded has been marked by a concentration of effort
in behalf of those most competent, while the severely and profoundly
retarded have been systematically excluded and placed in institutions
where little or no i programming was available. In the 1960 ' s less
than 5% of the nation's retarded were in institutional settings
i
(Butterfield, 1969;; Dingman & Tarjan, I960), while about half the
nation's profoundly retarded were institutionalized.
i *
Recent court decisions and the civil rights nature of the
Education of All Handicapped Children Act of 1975 (Public Law 94-142)
already have initiated and will continue the implementation of
services f ^ all handicapped children in the least restrictive
environment possible. The litigation and legislation is helping
to remedy the situation described above by making services available
to previously unserved populations and perhaps even more important,
by making sure the services are appropriate. As Abeson and Zettel
(1970) point out:
"There simply will not be any grounds for depriving a
ERIC
4
handicapped child, who because of that handicap,
possesses unique learning needs requiring special
education. No longer will it be permissible for
a sch-ool person to exclude or postpone the education
of such handicapped children on the grounds that they
cannot learn, their handicap is too severe, programs
do not exist, or for any other reason ... it means
that no child is uneducable or stated in another way,
all children can learn." (p. 122)
We. not only have to take this statement seriously, we also
must take it literallv. Responsibility for insuring success in
these efforts fall in two major areas: (a) delivery agencies
(e.g., public schools, institutions, etc.) and (b) training
agencies (e.g., universities, colleges, etc.). Two problems emerge
when educational efforts are directed at the severely handicapped.
One problem is that of breaking new ground. The second problem
involves systematically testing existing knowledge and techniques
on this heretofore unserved population. Our previous educational
efforts with the SPH were mostly based on intuition. That intui-
tion, at best, was based on our previous experience. According to
Haring and Pious (1977):
"--it is our responsibility as educators to teach and
to demonstrate unmistakably to the public that severely
handicapped persons can learn.
We are beginning to recognize some of the instruc-
tional components necessary for the undertaking of educa-
tional responsibilities mentioned earlier ... the
procedures that educators have developed in their work
with moderately handicapped children will have to take
a quantum leap; what we now consider sophisticated skills
m&y well seem primitive as we develop the instructional
competencies necessary to teach our new educational
clientele . . ." (p. 5-6)
There is no question that special education is "tooling up" to
meet the educational needs of the severely handicapped. The questio
is in what areas or toward which objectives do we invest our resources?
In the past few years there have been some programmatic efforts at
identifying knowledge which would be of benefU to severely handi-
capped populations. In general, these program can be classified
into two broad categories: investigations of operant procedures with
severely handicapped subjects and development of curricula for
severely handicapped populations.
*
The first program was essentially designed to demonstrate the
lav.' of effect across many different behaviors in severely handicapped
populations. Different operant procedures that were investigated all
depended on the law of effect for their theoretical base, thus their
success or failure was not really a test of the law of effect but
instead a gauge of the effectiveness of the reinforcer used. In essence
then, most of these studies dealt with incentive motivation questions
(Siegel, 1968). They answered questions involving the relative efficacy
of various reinforcers and the relative preference of the subjects for
them. Some recent examples of this include the preference of vibratory
over visual stimulation by severely and profoundly retarded subjects
(Ohwaki, Brahlek, and Stayton, 1973; Ohwaki & Stayton, 1976) and the
use of contingent vibratory stimulation to train nonambulatory pro-
foundly retarded students (Murphy & Doughty, 197/; Zucker, D'Alonzo,
McMullen & Williams, in press).
The second program was geared toward generating systematic and
specific curricula to meet the needs of severely handicapped popu-
lations. The focus of these programs was to use interdisciplinary
and practitioner input for generation of competencies applicable to
the severely handicapped. Evidently, this program has been quite
successful as indicated by the dissemination of many books, reports
and guides on curriculum for the severely handicapped (e.g., Donlon
& Burton, 1976; Sontag, 1977). These documents address topics
ranging .Vom general content areas such as c Mnunication and self-
help to specific objectives for eye movements and tongue positions.
The feature which is most noticable in the majority of these curriculum
specification efforts is that they are not watered down versions of
curriculums for moderately retarded children. Rather, these curricula
are directed solely toward training areas which relate to SPH popu-
lations. . ~
Notwithstanding the accomplishments of the above programs,
there seems to be a major gap in our research efforts at elucidating
information applicable to severely hand' capped populations. The
focus of research must change from what to teach to how to teach.
Answering the how to teach question by identifying additional
reinforcers or novel ways in which they may be applied is circum-
venting the question. Haywood (1977) stated:
"We do not need any more studies of single subjects
designed to demonstrate that the law of effect still
works and can be applied to yet another aspect of
behavior. The law of effect is as valid today as it
was in 1927." (p. 315)
Generating more curriculum, even appropriate curriculum, does
not answer the process question of how to achieve the competencies.
One example would be a recent special publication of the Division
on Mental Retardation of the Council for Exceptional Children
(Sontag, 1977). This volume entitled "Educational Programming
v
7
for the Severely and Profoundly Handicapped" devotes very little
actual space to the "how to" question. While many of the articles
have provocative titles, close inspection reveals they are generally
curricular descriptions. Implementation is generally assumed to
be operant in nature. For example Bricker and Iacino (1977)
relate this process:
"The next stage is systematic implementation which may
be largely dependent upon the personnel's knowledge
of and skill in using teaching strategies based on
behavioral principles." (p. 175)
The important point to be gleaned here is that operant techniques
cannot substitute for instructional strategies. These two dimensions
are not interchangeable, they are complementary. Operant technique?
are only as good as the instructional strategy which determines
their use. The rewarding and often dramatic results that are
accomplished using operants with severely handicapped populations
should not limit our investigations and development of appropriate
instructional strategies which would further enhance our success.
It is only through research that we can systematically test
existing strategies and based on the results develop new ones to
be further tested. Haywood (1977) emphasizes that:
"We need basic research on the fundamental mechanisms
by which human beings learn. We need somewhat more
complex research on the interactions among learning
strategies, personal characteristics of individuals,
types of material to be learned, settings for learn-
ing, and incentive conditions." (p. 316)
The identification and investigation of learning strategies
would seem to be of paramount importance if we wish to meet the
needs of low functioning children. In discussing the learning
9
ERIC
6'
performance of mildly handicapped students, Prehm (1976) has indicated
that handicapped children fail to use learning strategies spontaneously
and that we should correct this by teaching the strategy along with
the information to be learned. Learning strategies in this sense
are techniques used to help people remember and use information.
Providing learning strategies for severely handicapped populations
which effectively enhanced learning would fill the gap mentioned
earlier between operant procedures and curriculum implementation.
One learning strategy approach which has been used with non-
handicapped students is called cumulative programming. Becker,
Engelmann and Thomas (1975) define cumulative programming as follows:
". . . two concepts from a related set are brought to criterion. Then
new concepts are added one at a time and brought to criterion." (p. 257)
The use of this program as a method of instructio. as been explored
by a few researchers. A brief review of these studies follows.
Carnine (1976) demonstrated the effectiveness of introducing
similar sounding stimuli cumulatively in a letter-sound correspondence
task. The study was conducted with three groups of normal preschool
children. In one group the stimuli were presented simultaneously,
i.e., all the stimuli were on the same presentation sheet. For
this group, the similar sounding letters, e and i, were introduced
separately with four letters in between each presentation. The
stimuli were presented to the other two groups cumulatively, i.e.,
a new letter was added to the set of previously learned letters
only when the child had reached criteria on all the letters in
the current set. For one of the cumulative introduction groups,
9
the similar sounding stimuli were presented together, i.e., i
followed e. The similar sounding stimuli for the other cumulative
introduction group were separated, i.e., e and i were presented
with four other letters in between. The findings indicated that
the cumulative introduction group in which the similar sounding
stimuli were separated reached criterion more quickly than the
other two groups. Also, posttest scores for both cumulative intro-
duction groups were higher than the simultaneous group's posttest
scores .
Staats, Brewer & Gross (1970) studies alphabet reading in
11 preschool children. The letters were presented sequentially in
upper-case form. Pictures, used early in the program, were eliminated!
when the child could identify several letters. When a new letter
had been mastered, it was presented in sequence with several pre-
viously learned letters, i.e., for P the sequence would be L M N 0 P.
When the child learned the letter in sequence, it was randomly
presented with all the earlier letters in the alphabet, thereby
constantly reviewing those letters. The results demonstrated that
the use of this type of cumulative programming strategy resulted
in the children's acquisition of the letters at an increasingly
faster rate, i.e., they required fewer trials to master the second
half of the alphabet (N-Z) than were required to master the first
half of the alphabet (A-M) .
Ferster & Hammer (1966) used a cumulative programming strategy
to teach chimpanzees binary arithmetic. They compared successive
pairs and cumulative programming strategies on a number-symbol paired
10
10
associate learning task. The experimenters found that discrimination
of randomly presented numbers was accurate when cumulative programming
was used but when successive pairs programming was used, the animals
only responded at a chance level to randomly presented numbers.
Gruenenfelder & Borkowski (1975) tested the spontaneous transfer
of cumulative rehearsal during serial learning to a new list. They
divided 60 normal first grade children into two groups: (a) no instruc-
tion; and (b) cumulative rehearsal. From the data reported, the
authors concluded that some of the children in the second group did
not use the cumulative rehearsal strategy spontaneously after training.
The serial lists (three 9-item and one. 4-item) were slides constructed
using pictures from the Peabody Picture Vocabulary Test. The children
who used cumulative rehearsal needed fewer trials to reach criterion
than the other two groups (those with no instruction and those who »
received cumulative rehearsal instruction but did not spontaneously
transfer it after training). The cumulative rehearsal group took a
significantly longer time to reach criterion. The time measure was
contaminated because each subject in the study controlled the length
of stimulus item exposure time. Subjects in the cumulative rehearsal
group held items for significantly longer exposure periods. The
children who were instructed with cumulative rehearsal were able
to transfer their learning to a new serial list a week later where
no instructions were given.
Up to this point the studies cited have employed intellectually
normal children. There have been two studies, however, which investi-
gated cumulative programming strategies with moderately and severely
ERIC
li
handicapped children. Fink and Brice-Gray (1979) conducted a pilot
study investigating cumulative programming with 10 moderately and
severely handicapped preschoolers. Half the subjects were assigned to
a successive pairs programming group while the other half were assigned
to a cumulative programming group. The results indicated that subjects
who received the cumulative programming instructional treatment reached
terminal criteria in significantly less trials than the subjects who
received the successive pairs programming instructional treatment. Also,
subjects in the cumulative programming instructional treatment made
significantly more correct responses on an immediate recall measure
than did the successive pairs programming instructional treatment group.
P'/ehm, Zucker and Roth (1979) also tested the cumulative programming
paradigm, this time with severely and moderately retarded school age
children. They found subjects in a cumulative programming group required
less trials to reach criteria than did a successive pairs group. Unlike
Fink and Brice-Gray (1979), however, immediate recall did not differ be-
tween groups. On seven day recall the cumulative group performed sig-
nificantly better than the successive group.
The studies reviewed above using intellectually normal and primate
populations certainly reveal the potential utility of cumulative pro-
gramming strategies for teaching. Especially promising are the salutory
effects reported by Fink and Brice-Gray (1979) and Prehm, Zucker and Roth
(1979) in handicapped populations. These positive results and the
alarming dearth of other evicence in this area clearly call for more
research, on a larger scale, which will systematically identify, investi-
gate and validate cumulative programming strategies and the associated
parameters which interact with cnem.
Goal and Objective
The major goal of this project is to investigate the parameters of
cumulative programming instructional strategies in severely handicapped
populations. Realization of this goal will result in pragmatic techniques
which will have direct and immediate application to training in these
populations.
The major objectives which will accomplish this goal are:
1) investigation of variables descriptive of the subject;
2) investigation of variables descriptive of the content;
3) investigation of variables descriptive of the task.
In order to facilitate systematic exploration of these variables,
the following Model has been developed (Figure 1 on page 17). This
Model (adapted from Altman, 1973; Alt ' ' '-.Iking ton, 1971) will allow
for controlled formal investigation v U'^e .tntially relevant vari-
ables including: (a) those variable:, • of the. subject; (b)
those variables descriptive of the conte, (c) those variables
descriptive of the task. The list of variables comprising the dimensions
of this Model is not meant to be exhaustive. The list will serve as the
starting points for the beginning studies. Variables may be deleted or
added to the Model as deemed necessary depending on the results of our
explorations.
The Subject Variables of this model will allow us to ask questions
related to the effects of cumulative programming techniques on different
handicapped populations. For example, there may be differences between
severely and profoundly retarded, or between retarded and autistic, or
between ambulatory and non-ambulatory, etc. Chronological age of the
child may be a factor as could be sex or race. In addition, whether or
13
$9/
Task Variables
x
Subject Variables
Ability level
Chronological Age
Sex
Race
Institutionalization vs
Non-ins ti tutionalization
£Z ! Z Z
— ^ . '
E
0)
-o
a
i
e
CL
U r-
E -r-
oj a
-a o
a
<
o
u
>>
CO
a.
i i
0» r-
cr to
<o c
3 O
cn t-
-j a
o
Content Variables
Figure 1 - Dimensions identifying cumulative programming parameters
14
14
not the child resides at home or in an institution may be related to
success at certain tasks. Subject variables then, are those character-
istics of children which may effect the outcome of the instructional
technique.
Content variables refer to the different areas of skill or knowledge :
we would be training, tfhile it is obvious that cumulative programming
V'
techniques may enjoy differential success across these areas it is also
possible that dimensions within these may also effect outcomes. For
example, in the pre-academic area do cumulative instructional strategies
work equally well for recognition of shapes as opposed to beginning
number concepts? For academics the same questions could be asked about
reading and math skills. Socia] areas might involve differences between
one-to-one skills as opposed to group skills. Psychomotor (locomotion
vs. fine motor), language (receptive vs. expressive) and vocational
(assembly vs. sorting) areas also lend themsel'.es to these kind of
questions.
Task variables reflect areas related to the actual instructional
process. The interaction of this dimension with the other two may differ-
entially effect cumulative programming outcomes. For example, consideration
of the reinforcement variable would involve the efficacy of cumulative
programming under reinforcement vs. no reinforcement. Also subsumed
under this category would be the relative effectiveness of alternate
reinforcers (e.g. social vs. edible). Under design features such questions
such as group or individual instruction could be considered, Thecategory
of antecedent variables would include items such as previous experience
with the content and current skill level. Learning criteria refers to
15
a "barometer" of success which may differentially vary from a certain
"number' of corfect-respoTises to generalization to other settings or
different tasks. Finally, stimulus conditions refers to the number, rate
of presentation, order of presentation, etc. of stimuli in the cumulative
programming instructional sequence. >
Thus the subject, content, and task variables outlining the dimensions/
of this Model comprise the possible parameters which interact with cumu-
lative programming. At first inspection it. may appear that each cell
of the Model represents a study in itself. Closer examination, however,
reveals that each cell represents many separate studies within those
relevant dimensions. We have the option of manipulating the variable
dimensions while testing cumulative programming or holding the dimensions
constant while comparing cumulative programming to another type of pro-
gramming, or doing both simultaneously. Many times these decisions will
be based on the number of subjects that are available from each category.
One example of this would be the cell defined by the intersection of
reinforcement under Task Variables, ability level under Subject Variables,
and vocational under Content Variables. These dimensions suggest a variety
of studies one of which might be the effectiveness of cumulative programming
with three types of reinforcers (e.g., edible, social, and vibratory), on
two types of vocational task (e.g., sorting and assembly) on the learning/
performance of severely handicapped children (e.g., profoundly retarded
and autistic). Or, we could hold the dimensions constant and vary the
type of instructional strategy (e.g., cumulative programming and succes-
sive programming), or, we could simply add this fourth factor of instruc-
tional programming to the previous three dimensions. What would determine
the specific elements of a given study would be the number of subjects
available and the specific questions under investigation.
It is obvious that the model depicted could be used to generate
a seemingly endless number of studies. In order to limit the scope of
our studies we have a number of high priority variables which would
receive. early attention. Specifically, along the content dimension we
would highlight pre-academic and academic variables, along the subject
dimension we would highlight ability level and chronological age variables
and along the task dimension we would highlight stimulus conditions and
learning criteria variables. These variables seem most relevant based
on previous research and pragmatic value in the classroom. Thus, our
initial series of studies will concentrate on these factors and the re-
sults of these studies will aid in the choice of subsequent investigation
variables from our model.
RESULTS OR BENEFITS EXPECTED
As indicated in the preceding section, there is very little available
knowledge related to the effectiveness of various instructional strategies
with SPH children. Educational efforts are dependent upon the teachers'
knowledge of reinforcement techniques and rather standardized programming
techniques. The effectiveness of alternative programming techniques,
particularly as these techniques effect the learning of SPH children,
is not known. If teachers are to make sound instructional judgments,
evaluations of the parameters of potentially powerful educational pro-
gramming techniques are needed.
This three year project is iesigned to fill a void in our knowledge
of effective instructional programming for SPH individuals. Completion
of the project will provide teachers of the SPH and program planners
with extensive information regarding the cumulative programming instructional
17
17
strategy. The knowledge gained through the proposed project can be used
i
to plan curricula, to design specific instructional activities, and to
implement actual instruction. Through this project more effective and
efficient instruction for the SPH will be developed. The primary benefits
of the proposed project will be the increased skill levels of (a) the
pupils involved in the project and (b) the pupils taught by persons using
the results of the project. Additional benefits will be increased ability
of educators to design effective instructional programming strategies for
SPH children.
APPROACH
Plan of Action
The specific and general procedures to be detailed are to be con-
sidered generic to all studies to be conducted in the research program.
The model presented allows us to pick variables and construct investi-
gations based on the actual results of our ongoing studies. This method
is preferred to one which specifies in advance the exact dimensions of
each individual study. It may be that certain dimensions of' the model
will give us more useful information, thus, a flexible approach can capit-
alize on this. The purpose of the following description is to outline
the procedures which will be used independent of the variables under
investigation. The variables used here are examples taken from the sub-
ject, task, and content model.
The pjrpose of chis sample study is to investigate the relative
efficacy of two instructional strategies (e.g. cumulative and successive
pairs programming) under two types of reinforcement (e.g. social and
edible) on the learning of two-syllable functional sight words by severely
18
retarded and autistic children. Referring back to our model, this study
would be varying two of the dimensions (task and subject) while holding
the third constant (content) and adding and varying a fourth (instructional
strategy). This demonstrates the flexibility available in structuring
and investigating research questions.
Subjects
Severely retarded and autistic subjects will be randomly selected
from their respective populations. Severely retarded subjects will be
randomly assigned to one of four treatment conditions as will autistic
subjects, thus, constituting eight total conditions. After assignment,
subject characteristics such as age, sex etc. will be statistically
compared to check on differences between conditions. It is anticipated
that 10 subjects per condition will be required to satisfy general sub-
ject/variable ratio requirements. Numbers of subjects may be a factor
but, in general, the greater Phoenix area has many public and private
facilities which serve severely handicapped children. Many of these
have cooperated in tne past with either funded or individual projects
and a sample of these have responded favorably to our current inquiries.
ERIC
Procedure
Each subject will be run individually in a predetermined random
order each day. The experimental session will take place in an available
room inthe student's school. Essentially, this room will be minimally
distracting and contain a table and chairs for the experimenter and
subject. The subject will sit across the table from the experimenter.
The stimuli will be five two-syllable functional sighii words (e.g.,
19
entrance, women, danger, exit, and poison) presented individually on
13 x 20.5 em cards, randomly ordered from a through e. All stimuli
used will be novel to the subjects.
Instruction will begin with stimulus a. The experimenter will
present stimuli a and b to the subject and touch stimulus a and
say, "This is . Touch ." If the subject responds correctly,
i.e., touches the correct stimulus, the experimenter will immediately
reinforce the subject with praise or edibles, depending on which
reinforcement treatment condition the subject is in. If the subject
responds incorrectly, the experimenter will immediately say, "No,
this is (and touch the correct stimulus). Touch . " The
experimenter will then switch the left-right position of the stimuli and
instruct the subject to touch the stimulus again. The order of the
arrangement of the cards will be randomly determined prior to the
beginning of the study. The subjects will be given twenty trials per
day or the number required to reach criterion if less than twenty.
Training will continue until the subject reaches a criterion of eight
consecutive correct responses. When the subject reaches criterion
on stimulus a, the experimenter will follow the same procedures to
teach stimulus b. This would be step 2.
In step 3, the subject will discriminate stimulus a from stimulus
b. The experimenter will present both stimuli to the subject and
instruct the subject to identify one stimulus at a time. For each
trial, the experimenter will follow the order of switching stimuli
referred to in step 1 and instruct the subject to choose stimuli in
the order given on the data sheets in Appendix A. From this point on,
ERIC
20
20
the training will differ depending on instructional condition.
For the next twelve steps the subjects in the successive pairs
instructional programming group will be instructed as follows: the
next single stimulus will be introduced with stimulus a and brought
to criterion. Then, that stimulus will be discriminated from stimu-
lus a. This procedure will continue until all single stimuli had
been introduced and all possible pairwise discrimination are made
by the subject.
For the next six steps the subjects in the cumulative instruc-
tional programming group will be instructed as follows: the next
single stimulus will be introduced and brought to criterion with
stimuli a and b, i.e., the subject will only have to identify the
new stimulus. Then the subject will be required to identify the
new stimulus as well as a and b. This procedure will continue until
all single stimuli have been introduced and all possible stepwise
discriminations have been made (see Appendix A). Subjects will be
required to respond with at least 75 percent correct accuracy on
previously learned stimuli after moving beyond stimulus c (step four)
i.e., in step five, when the subject is required to identify stimuli
a, b, and c, the subject could get 75 percent correct response or six-
out-of-eight on a and b until the subject correctly identifies stimulus
c eight consecutive times.
When a subject has reached criterion on the last step of his
respective program, a test will be administered. The experimenter will
place all five stimuli on the table in front of the subject. Then,
following the random order given on the data sheets in Appendix A,
2i
the experimenter will instruct the subject to touch each stimulus.
/The subject's responses will be recorded. In addition to this data
the experimenter will also record responses to each step of the instruc-
tional program on the data sheets. This will allow tabulations of the
• total number of trials each subject takes to reach criterion.
Analysis
The data will consist of scores on two dependent measures (trials
to criterion and number correct on test) which reflect the dependent
variables of rate and recall. The statistical technique that will be
used to analyze the data and test appropriate null hypotheses will be a
three-way (reinforcement x subject x strategy) multivariate analysis of
variance (MANOVA) (Anderson, 1958; Morrison, 1967; Tatsuoka, 1971).
Using the MANOVA reduces the possibility of a spurious rejection of null
hypotheses, which increases when spearate analysis of variance (ANOVA)
tests are performed for each of the dependent measures. The MANOVA can
yield a number of criteria to test effects and interactions. These
include Hotel ling-Lawley's Trace, Pillai's Trace, Wilks" Likelihood Ratio
Criterion, and Roy's Maximum Root Criterion (Heck, 1969; Pillai, 1960;
Schatzoff, 1966). The first two can yield approximations of the F
distribution while the latter two are based on their own distributions.
The evidence available at present does not indicate any superiority of
one criterion over any other (excluding computational factors) for the
purposes of this study (Ghosh, 1964; Kshirsager, 1972; Mikhail, 1965;
Pearson, 1971, Pillai & Dotson, 1969; Pillai & Jayachandran, 1967; 1968).
If the MANOVA indicates rejection of any of the null hypotheses,
then separate univariate ANOVAs will be performed on each dependent
/
22
measure. In addition, where indicated, post-hoc Newman-Keuls comparisons
will be used to indicate differences between means across conditions and
simple or simple simple effects will be ascertained for significant
interactions (Steel & Torrie, 1960). The probability level for rejection
of all null hypotheses will be .05.
It should be emphasized that following these procedures and based
on the results, new studies will be designed and investigations of the
parameters of cumulative programming will continue. It is anticipated
that from three to five such studies can be conducted each year for the
three year grant period. By utilizing the overall model and approach
described above, the researchers can capitalize on the results of each
of the experiments.
Projection of Accomplishments
It is anticipated that from 70 to 120 SPH persons (preschool,
school age, and adolescents) will participate in project activities
during' each of the three years of the research program. The Phoenix
metropolitan area has sufficient numbers of SPH persons to support the
research program proposed. Criteria consistent with the AAESPH defini-
tion of SPH persons will be used as the basis for selecting the subjects
to participate in the project.
The proposed project is projected for a three year period. Pro-
jected accomplishments of the project are listed below.
Year 1
July - September 1980
Select parameters for study for first three experiments
Develop materials
ERIC
23
23
Select subjects
Hire Graduate Assistants
October - December 1980
Conduct Study 1
Begin Study 2
Submit proposals for dissemination at conferences and workshops
Prepare continuation request
January - March 1981
Conduct Study 2
Conduct Study 3
April - June 1981
Complete all Year 1 studies
Write final report of Year 1 activity
Design Year 2 Studies
Make appropriate presentations
Year 2
July - September 1981
Conduct Studies 4 and 5
Submit proposals to appropriate professional organizations
for conference presentations
October - December 1981
Conduct Studies 5 and 6
Prepare continuation request
January - March 1982
Conduct Studies 6 and 7
Conduct one local dissemination workshop
ERIC
2<1
24
April - June 1982
Conduct Studies- 7^nd 8
Make two national conference presentations
Conduct one local dissemination workshop
Complete all Year 2 activities
Year 3
July - September 1982
Write report of Year 2 activity
Design Year 3 Studies
Submit professional meeting presentation proposals to
relevant professional organizations
October 1980 - March 1983
Conduct Studies 9-12
Conduct local dissemination workshops three, four and five
Make appropriate conference presentations
April - June 1983
Conduct dissemination workshops six and seven
Complete Year 3 activity
Make appropriate conference presentations
Write final report
The benefits to be derived from this project are clear in terms
of their impact on handicapped children. The results of clarification,
validation, application and implementation of cumulative programming
techniques tested in this research program and their related parameters
can be immediately utilized in educational applications for the severely
handicapped. There will be no research to practice lag in implementing
these findings. The research program proposed here is entirely prag-
matic in nature, it involves application of teaching strategies to various
ERIC 2 0
children across various tasks and areas.
The dissemination plan involves much more than the typical re-
search presentations at national conferences and publication in
relevant journals. While these are important areas of dissemination,
the time lag between exposure and implementation with children is too
great. Therefore, in addition, project staff will conduct a series
of dissemination workshops throughout the state for public and private
agencies dealing with severely handicapped children. These workshops
would enable immediate implementation of the findings by these respective
agencies. In this way, over 1000 severely handicapped children would
directly benefit from this project while many more would benefit
indirectly.
Also, this same type of approach (dissemination workshops) is
planned for the national level. Organizations such as The American
Association on Mental Deficiency and The American Association for the
Education of the Severely and Profoundly Handicapped sponsor special
courses and workshop sessions at their national meetings which are
practitioner oriented. We would submit proposals to conduct these types
of activities, thus enhancing the probability that our findings would
be implemented at a national level.
ERIC
2d
^1
26
ACCOMPLISHMENTS
Using the timeline presented in the previous section as a reference,
the following study descriptions will indicate that we have completed
all project activities. All studies described below followed the gen-
eral experimental procedures detailed earlier in this report. Data
analysis was accomplished using a Mann-Whitney U test. The parametric
analyses previously described were abondcned due to the small size of
each study. It was decided that the non-parametric test would be much
more robust in giving us accurate information regarding group differences.
Study 1
This study was conducted at Loloma School in Scottsdale. We
compared cumulative and successive programming for teaching sight words--
to elementary severely regarded subjects, thus, it was a replication of
Prehm, Zucker and Roth (1979). Data are presented below in Table 1.
The data analysis indicated that cumulative programming was superior
to successive programmong on the number of trials to criterion (p=,05),
on the one-day post- test (p= .05 ) and on the seven-day post-test (p<.05).
Study 2
This study was conducted at the Arizona Preschool for Retarded
Children in Phoenix. We compared cumulative and successive programming
for teaching sight words to severely retarded preschool subjects, thus,
it was areplication of <fink and Brice-Gray (1979). Data are presented
below in Table 2. The data analysis indicated that cumulative pro-
gramming was not superior to successive programming on the number of
trials to criterion (p>.05) or on the one-day post-test (p?.05). However,
on the seven-day post-test, cumulative was superior to successive (p=.05).
erJc 2 "i
TABLE 2
Cumulative
Subject Trials 3 Pl b " P2^
1 10 • 8 10
2 16 8 9
3 17 8 10
4 17 9 10
5 28 10 10
6 34 10 10
Successive
X 20.33 x 8.33 X 9.83
1 16 3 4
2 16 4 v.5
3 37 5 7
4 49 9 9
5 78 10 10
X 39.20 X 6.20 I 7.00
a = blocks of twenty
b = one day posttest
c = seven day posttest
Study 3
This study was a comparison of cumulative programming across age
levels on an academic task. This was accomplished by comparing the
cumulative group from Study 1 with the cumulative group from Study 2.
In other words, we have two cumulative groups learning sight words
witn the only difference between the groups being age of the subjects.
The data are presented below in Table 3.
The data analysis indicated that cumulative programming was dif-
ferentially effective across age levels. While there were no signi-
ficant differences between pre-school and elementary age students on
one-day and seven-day post- tests, the preschool students took signi-
ficantly more trials (p*.05) to reach criterion than the elementary
age group. -
Study 4
This study was conducted at the Arizona Preschool. We used
cumulative programming to teach the names of colors to a group of
eleven preschool severely retarded subjects. The school year
ended before we could complete data collection.
Study 5
This study was conducted at Loloma School. Seven severely
retarded elementary subjects were instructed with cumulative pro-
gramming to teach them manual signs for practical words. Data from
this task will be compared to the cumulative data from Study 1 to
ascertain differences between psychomotor and non-motor academic tasks.
2B
TABLE 3
Cumulative Programming
• Elementary -
Subject Trials 3 Pl b
' 1 29 10
2 20 4
3 20 10
4 18 8
•5 12 9
6 11 7
7 10 10
8 10 10
9 9 10
10 9 10 .
11 9 10
51 14.27 X 8.91
Pre-School
1 10 8
2 16 8
3 17 8
4 17 .9
5 28 10
6 34 10
I 20.33 X 8.33
a = blocks of twenty
b = one day post-test
c = seven day posttest
30
Additional data for this study were collected from subjects in Study 14.
The data and results are presented under Study 14.
Study 6
This study was conducted at Loloma School. Ten elementary severely
retarded subjects were instructed with cumulative programming to teach
them sight words using 100% as the criterion for learning. Data from
this task was compared to the cumulative data from Study 1 (where the
learning criterion was 80%) to ascertain differences attributable to
c anges in learning criteria. Data are presented below in Table 4.
The data analysis indicated no significant differences between the two
learning criteria groups (p>.05)o
r
Study 7
This study was conducted at CIDS/ESP housed at Papago School in
Phoenix. Sixteen elementary age profoundly retarded subjects with
multiple handicaps were taught the names of common objects using the
cumulative programming technique. The study was discontinued after
80 days. No subjects were progressing toward criterion.
Study 8
This study was conducted at Loloma School. Six elementary
severely retarded subjects were instructed with cumulative programming
to teach them manual signs for practical words to 100% criteria.
These data were compared to the data from Study five where the criterion
was 80%. Data are presented below in Table 5. The data analysis
31
\
32
■. TABLE 4
Cumulative Programming
80%
Subject Trials 9 Pl b P2 C
1 29 10 10
2 20 4 6
3 20 10 9
4 . 18 8 10
5 12 9 9
6 11 • 7 8
7 10 10 10
8 10 10 7
9 9 10 10
10 9 10 10
11 9 _10 JO
X 14.27 X 8.91 X 9.00
100%
1
2
3
4
5
6
7
8
9
10
10
10
13
10
10
9
9
10
11
18
11.00
10
9
10
10
10
10
8
10
10
8
X 9.50
10
10
7
8
9
10
7
10
10
_ 7
X 8.80
a = blocks of twenty
b ■ one-day post-test
c s seven-day post-test
33
TABLE 5
Cumulative Programming
100%
Subject T rials 9 PI 6 ^
1 13 10 8
2 10 . 10 10
3 9 9 9
4 9 10 9
.5 9 10 9
6 _ 9 io 10
X 9.83 X 9.83 X 9.17
80%
1 31-9 10
2 38 10 10
3 33 9 9
4 12 ,8 10
5 17 9 10
6 15 10 9
7 15 8 8
8 16 10 10
9 11 9 9
10 11 10 10
11 20 10 10
12 j 16 10 ' 10
13 I 32 6 .6
14 _ 11 8 8
X 18.43 X 9.00 X 9.21
a = blocks of twenty
b = one-day post-test
c = spven-day post-test
9
ERIC
33
34
indicated it took significantly longer to learn the signs at 80%
criterion (p<.05). There were no difference.-., however, between the
groups on the post- tests (p>.05).
Study 9
This study wa? conducted at Loloma School . Ten elementary
severely retarded subje^s were instructed with cumulative progamming
to teach them the pairing of words and numbers from 1-5 to 80% cri-
terion. Data from this study was compared to the cumulative data from
Study 10 to ascertain differences between academic and motor (pre-
vocational )task. Data are presented in Table 6.
The data analysis indicated a significant difference in trials
to criterion in favor of the academic task (p<.05). There were no
differences between groups on the first and second post-test (p>.05).
Study 10
This study was conducted at Tonalea School. Elementary age
severely retarded subjects were taught a pre-vocational sorting task
using either cumulative or successive programming. The task involved
size descrimination along one dimension and a motor response. Data
are presented below in Table 7.
The data analysis indicated no differences between groups on
trials to criterion and seven-day post-test (p>.05). One-day post-
test scores were significantly different in favor of the cumulative
group (p<. 05) .
34
35
TABLE 6
Cumulative Programming
Academic
Subject
1
2
3
4
5
6
7
8
9
10
Trial s d
11
9
9
9
9
9
9
9
9
_ 9
X 9.20
7
10
9
8
9
10
' 10
10
7
10
X 9.09
P2
6
10
9
10
7
10
10
10
6
_10
X 8.80
Motor
1
2
3
4
5
65
22
9
9
11
X 23.20
9
8
9
9
10
X 9.00
4
9
8
8
JO
X 7.80
a = blocks of twenty
b = one-day post-test
c = seven-day post-test
3o
36
TABLE 7
Cumulative
Subject
1
2
3
4
5
Trials 0
65
22
9
9
_ 11
X 23.20
EI
9
8
9
9
10
X 9.00
£2
4
9
8
8
10
f 7.80
Successive
1
2
3
4
5
6
7
8
9
21
21
15
28
15
15
15
15
15
17.78
4
5
3
4
7
6
7
5
_ 8
X 5.44
2
0
5
10
9
8
8
6
_ 5
X 5.89
a = blocks of twenty
b = one-day post-test
c = seven-day post-test
9
ERIC
Jo
Study 11
This study was conducted at Tonalea School. Adolescent age
severely retarded subjects were taught a pre- vocational sorting task
using either cumulative or successive programming. The task involved
size discrimination along one dimension and a motor response. Data
are presented below in Table 8. The data analysis indicated a sig-
nificant difference in trials to criterion in favor of the cumulative
group (p<.05). There were no significant differences between groups
on the post-tests (p>.05).
Study 12
This study was a comparison of cumulative programming across age
levels on a pre-vocational motor task. This was accomplished by com-
paring the cumulative group from Study 10 with the cumulative group
from Study 11. Data are presented below in Table 9. The data analysis
indicated a significant difference between groups only on the one-day
post-test (p<.05). There were no significant differences in trials to
criterion and seven-day post-test (p?.05).
Study 13
This study was conducted at the CIDS/ESP housed at Papago School
in Phoenix. Ten elementary profoundly retarded students were taught
a pre-vocational assembly task requiring a motor response, using the
cumulative programming technique. This study was discontinued after
a year and a semester. Many of the subjects had over 1700 trials and
were not progressing toward criterion. The data are presented below
in Table 10.
38
TABLE 8
Cumulative
Subject
1
2
3
4
Trials*
10
9
9
_ 9
X 9.25
PI
8
8
8
5
X 7.25
P2
6
9
8
5
X 7.00
Successive
1
2
3
4
5
15
15
15
15
_ 15
X 15.00
8
8
0
9
6
X 6.20
9
7
6
8
_ 1
X 6.20
TABLE 9
Cumulative Programming
Elementary
Subject
1
2
3
4
5
Trials 9
65
22
9
9
J 23.20
Pl°
9
8
9
9
10
I 9.00
P2
4
9
8
8
JO
X 7.80
ERIC
Adolescent
1
2
3
4
a = blocks of twenty
b = one-day post-test
c - seven-day post-test
10
9
9
_ 9
X 9.25
33
8
8
8
5
7.25
6
9
8
_ 5
X 7.00
TABLE 10
Cumulative Programming
Subject
Trials 3
1
90
2
91
3
98
4
89
5
88
6
90
7
82
8
85
9
89
10
91
a = blocks of twenty
b = one-day post-test
c = seven-day post-test
Study 14
This study was conducted at Tonalea School. Fourteen elementary
severely retarded subjects were instructed with cumulative programming,
to teacti tham manual signs for practical words. These data were com-
pared to the cumulative data from Study 1 to ascertain differences
between psychomotor and non-motor academic tasks. The data is pre-
sented below in fable 11. The data analysis indicated no significant
differences between groups (pr.05).
Study 15
This study was conducted at Tonalea School. Fourteen adolescent
severely retarded subjects were instructed with cumulative programming
to teach them manual signs for practical words. Half the subjects
learned to QQ% criterion while the other half learned to 50% criterion.
The data is presented below in Table 12.
The data analysis indicated a significant difference in trials to
criterion in favor of the 50% group (p<.05). There were no significant
differences in post-test performance (p>.05).
In summary, it can be seen that our studies were directly related
to the stated project objective cf investigating variables related to
the subject, the content, and the task.
40
41
Sight Words
Signs
a = blocks of twenty
b = one-day post-test
c = seven-day post-test
TABLE 11
Cumulative Programming
Subject
Trials 3
1
29
2
20
3
20
4
18
5
12
6
11
7
10
8
10"
9
9
10
9
11
9
X 14.27
1
31
2
18
3
33
4
12
5
17
6
15
7
15
8
16
9
11
10
11
11
20
12
16
13
32
14
11
J 18.43
o
ERIC
41
Pl b ' * P2 C
10 10
4 6
10 9
8 10
9 9
7 8
10 10
10 7
10 10
10 10
_10 JO
X 8.91 X 9.00
9 10
10 10
9 9
8 10
9 10
10 9
8 8
10 10
9 9
10 10
10 10
10 10
6 6
_ 8 8
X 9.00 X 9.21
TABLE 12
Cumulative Programming
42
80%
Subject
' 1
2
3
4
5
b
Trials 0
15
16
11
11
20
16
_ 11
X 14.29
£1
8
10
9
10
10
10
8
X 9.29
P2 L
8
10
9
10
*10
10
_ 8
X 9.29
50%
1
2
3
4
5
6
7
9
9
9
. 9
9
9
9
19.00
10
10
10
10
10
10
10
10.00
10
10
10
10
10
10
10
10.00
a = blocks of twenty
b = one-day post-test
c = seven-day post-test
4
Dissemination
The project dissemination activities served as the major forum
for discussion of project study results and their utility for class-
room/traing setting implementation and the direction of future research
As indicated in the project timeline, dissemination took place over
three-years and was quite extensive. A summary of dissemination
activities is presented below:
Local dissemination workshops/staff training were conducted
at the following sites:
Loloma School
- Scottsdale
Getz School
- Tempe
Arizona Preschool
- Phoenix
Montebello School
- Phoenix
Papago School
- Phoenix
Rich School
- Phoenix
Tonal ea School
- Sco + tsdale
Tolleson Elementary
- Tolleson
National dissemination presentations/workshops were as fol^ws:
Presented: Zucker, S. H. & Prehm, H. J. Cumulative versus
successive programming with severely retarded students. The
Gatlinburg Conference on Research in Mental Retardation/
Developmental Disabilities, Gatlinburg, Tennessee, March, 1981.
Presented: Zucker, S. H. & Prehm, H. J. Alternative teaching
strategies for severely handicapped students. Annual Meeting of
the American Association on Mental Deficiency, Detroit, May, 1981.
Session Moderator: Instruction of the Severely/Profoundly Retarded.
Annual Meeting of the American Association on Mental Deficiency, 1981.
43
44
Panelist: National Workshop Conference on Vocational and Employment
Opportunities for 'he Mentally Retarded. President's Committee on-
Mental Retardation, Madison, Wisconson, March, 1982.
Presented: Zucker, S. H. & Prehm, H. J. Research on ii tructional
strategies for severely/profoundly retarded students. A,.inual
Meeting of the Council for Exceptional Children, Houston, Texas,
April, 1982.
Chair: Session on Mental Retardation. Annual Meeting of the Council
for Exceptional Children, Houston, Texas, April, 1982.
Presented: Zucker, S. H. & Prehm, H. J. Durability of training
among severely retarded children as a function of teaching
'str'atdgy; " The '6ift1t1lb'uf^'X9hTSn9fRSe 'on Research in Mental
Retardation/Developmental Disabilities, Gatlinburg, TN, April, 1982.
Session Moderator: Research Symposium on Educational Programming
for the Moderately and Severely Retarded. Annual Meeting of the
American Association on Mental Deficiency, Boston, June, 1982.
Presented: Zucker, S. H. & Prehm, H. J. Cumulative teaching
strategies for increasing the retention of severely retarded
handicapped students. Annual Meeting of the American Association
on Mental Deficiency, Boston, June, 1982.
Presented: Zucker, S. H. Stability of response choice of severely
retarded children. The Gatlinburg Conference on Research in Mental
Retardation/Devel jpmental Disabilities, Gatlinburg, TN, March, 1983.
Chair: Session on Mental Retardation. Annual Meeting of the Council
for Exceptional Children, Detroit, April, 1983.
Presented: Zucker, S. H. & Prehm, H. J. Effects of two teaching
strategies on acquisition and retention among severely retarded
students. Annual Meeting of the Council for Exceptional Children,
Detroit, April, 1983. .
Presented: Zucker, S. H. & Prehm, H. J. Instructional research
on retention of information among severely retarded students.
Annual Meeting of the American Association on Mental Deficiency,
Dallas, June, 1983.
Presented: Zucker, S. H. Cumulative teaching strategies for
severely retarded students. Southeast Regional Resource Center,
Juneau, Alaska, September, 1983.
44
45
REFERENCES
4o
0
ERIC
46
Abeson, A., & Zettel , J. The end of the quiet revolution: The Education
for all Handicapped Children Act of 1975. Exceptional Children ,
1977, 44, .114-128.
Altman, R. Implications of modeling for t.ie mentally retarded. Paper
presented at the Annual Meeting of the American Association on
Mental Deficiency, Atlanta, 1973.
Altman, R. , & Talkington, W. W. Modeling: An alternative be. : ,or modi-
fication approach for retardates. Mental Retardation , '.s/1, 9, 20-23.
Anderson, T. W. An introduction to the multivariate statistical
analysis . New York: John Wiley & Sons, 1958.
Becker, W. C. , Engelmann, S., & Thomas", D. R. Teaching 2: Cognitive
learning and instruction . Chicago: Science Research Associates,
1971:
Bricker, D. D. & Iacino, R. Early intervention with severely/profoundly
handicapped children. In E. Sontag (Ed.), Educational programming
for the severely and profoundly handicapped . Reston, Virginia:
Council for Exceptional Children, 1977.
Butterfield, E. C. .Basic facts about public residential facilities for
the mentally retarded. In Kugel , R. B. and Wolfensberger, W. (Eds.),
Changing Patterns in Residential Services for the Mentally Retarded .
Washington, D. C: President's Committee on Mental Retardation7T969.
Carnine, D. W. Similar sound separation and cumulative introduction in
learning letter-sound correspondences. Journal of Educational
Research , 1976, 69, 368-372.
Dingman, H. F. & Targan, G. Mental retardation and the normal distri-
bution curve. American Journal of Mental Deficiency , 1960, 64,
991-994.
Donlon, E. T. & Burton, L. F. The severely and profoundly handicapped :
A practical approach to teaching . New York: Grune & Stratton, 1976.
Ferster, E. B. & Hammer, C. E. Synthesizing the components of arithmetic
behavior. In W. K. Honig (Ed.), Operant behavior: Areas of research
and application . New York: Appleton-Century-Crofts , 1966.
Fink, W. T. & Brice-Gray, K. J. The effect of two teaching strategies
on the acquisition and recall of an academic task by moderately
and severely retarded preschool children. Mental Retardatio n, 1979,
17, 8-12.
9
ERIC
4b
47
Ghosh, M. N. On the admissibility of some tests of MANOVA. Annals of
Mathematical Statistics , 1964, 35, 789-794.
Haring, N. 6. & Pious, C. Future directions in work with severely and
profoundly handicapped persons: An Overview. In N. 6. Haring &
L. J. Brown (Eds.), Teaching the severely handicapped , New York:
Grune & Stratton, Inc., 1977.
Haywood, H. C. The ethics of doing research ... and of not doing it.
American Journal of Mental Deficiency , 1977, 81, 311-317.
Heck, D. L. Charts of some upper percentage points of the distribution
of the largest characteristic root. Annals of Mathematical
S tatistics , 1960, 31, 625-642.
Kshirsagar, A. M. Multivariate Analysis . New York: Marcel Dekker, 1972.
Mikhail, M. N. A comparison of tests of the Wilks-Lawley Hypothesis in
multivariate analysis. Biometrika , 1965, 52, 149-156.
Morrison, D. F. Multivariate statistical methods . New York: McGraw-
Hill, 1967. :
Murphy, R. J., & Doughty, N. P. Establishment of controlled are movements
in profoundly retarded students using response contingent vibratory
stimulation. American Journal of Mental Deficiency , 1977, 82,
212-216.
Ohwaki, S. & Stayton, S. E. Preference by the retarded for vibratory
and visual stimulation as a function of mental age and psychotic
reaction. Journal of Abnormal Psychology , 1976, 85, 516-522.
Ohwaki, 5., Brahlek, J. A., & Stayton, S. E. Preference for vibratory
and visual stimulation in mentally retarded children. American
Journal of Mental Deficiency , 1973, 77, 733-736.
Pearson, E. S. Tables for multivariate analysis . London: Biometrika, 1971.
Pi Hal, K. C. S. Statistical tables for tests of multivariate hypotheses .
Manila: The Statistical Center, 1960.
Pillai, K. C. S. & Dotson, C. D. Power comparisons of tests of two
multivariate hypothesis based on individual characteristice roots.
Annals of the Institute of Statistical Mathematics , 1969, 21, 49-55.
Pillai, K. C. S. & Jayachandran, K. Power comparisons of tests of two
multivariate hypotheses based on four criteria. Biometrika , 1967,
54, 195-210.
Prehm, H. J. Learning performance of handicapped students. The High
School Journal , 1976, 59, 275-281.
9
ERIC
4 V
48
Prehm, H. J., Zucker, S. H. & Roth, K. Cumulative programming as an
instructional strategy with moderately and severely retarded
children. Working Paper No. 13, Department of Special Education,
Arizona State University, 1979
Schatzoff, M. Exact distributions of Wilks's likelihood ratio criterion.
Biometrika , 1966, 53, 347-358.
Siegel, P. L. Incentive motivation in the mental retardate. IniN. R.
Ellis, (Ed.), International review of research in mental retardation
(Vol. 3). New York: Academic Press, 1968.
Sontag, E. (Ed.). Educational programming for the severely and profoundly
handicapped . Reston, Virginia: Council for Exceptional Children,
Staats, A. W. , Brewer, B. A., & Gross, M. C. Learning and cognitive
development: Representative samples, cumulative-hierarchy learning
and experimental methods. Monographs of the Society for Research
in Child Development, 35, Serial No. 141, 1970.
Steel, R. G. D. & Torrie, J. H. Principles and procedures of statistics .
New York: McGraw-Hill, 1960.
Tatsuoka, M. N. Multi variate analysis: Techniques for educational and
psy chological research . New York: John Wiley & Sons, 1971.
Zucker, S. H., D'Alonzo, B. J., McMullen, M. R. & Williams, R. L. Training
eye-pointing behavior in a non-ambulatory profoundly mentally retarded
child using contingent vibratory stimulation. Education and Training
of the Mentally Retarded , in press.
o
ERIC
4d
49
APPENDIX A
er|c 4i)
DATA SHEET - SUCCESSIVE PAIRS PROGRAMMING
Subject: a= d=
Date:- b= e =
Experimenter: c-
Directions: circle each correct answer, cross-out each wrong answer.
STEP 1
STEP 2
STEP 3
STEP 4
STEP 5
STEP 6
STEP 7
a a
b
b
ab
ab
c
c
ac
ac
d
d
ad ad
a a
b
b
ba
ba
c
c
ca
ca
d
d
da da
a a
b
b
ab
ab
c
c
ac
ac
d
d
ad ad
a a
b
b
ba
ba
c
c
ca
ca
d
d
da Ha
a a
b
b
ab
ba
c
c
ac
ca
d
d
ad da
a a
b
b
ab
ab
c
c
ac
ac
d
d
ad ad
a a
b
b
ba
ab
c
c
ca
ac
d
d
da ad
a a
b
h
ba
ab
c
c
ca
ac
d
d
da ad
a a
b
b
ab
ba
c
c
ac
ca
d
d
ad da
a a
b
b
ab
ab
c
c
ac
ac
d
d
ad ad
STEP 8
STEP 9
STEP 10
STEP 11
STEP 12
STEP 13
STEP 14
e e
ae
ae
be
be
bd
bd
be
be
cd
cd
ce ce
e e
ea
ea
cb
cb
db
db
eb
eb
dc
dc
ec ec
e e
ae
ae
be
be
.bd
bd
be
be
cd
cd
ce ce
e e
ea
ea
be
cb
bd
db
be
eb
cd
dc
ce ec
e e
ae
ea
cb
cb
db
db
eb
eb
dc
dc
ec ec
e e
ae
ae
be
be
bd
bd
be
be
cd
cd
ce ce
e e
ea
ae
cb
be
db
bd
eb
eb
dc
cd
ec ce
e e
ea
ae
cb
be
db
bd
be
be
dc
cd
ec ce
e e
ae
ea
be
cb
bd
db
be
eb
cd
dc
ce ec
e e
ae
ae
be
be
bd
bd
eb
be
cd
cd
ce ce
STEP 15
de de
ed ed
de de
de ed
ed ed POSTTEST: a c d a b
de de
ed de e d b e c
ed de
de ed
de de
bo
DATA SHEET - CUMULATIVE PROGRAMMING
Subject: a= d=
Date: b= e=
Experimenter: c=
Directions: circle each correct answer, cross-out each wrong answer.
STEP 1 STEP 2 STEP 3 .STEP 4 STEP 5 STEP 6 STEP 7
a a
a a
a a
a a
a a
a a
a a
a a
a a
a a
STEP 8
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
STEP 9
ab ab
ba ba
ab ab
ba ba
ab ba
ab ab
ba ab
ba ab
ab ba
ab ab
c c
c c
c c
c c
c c
c c
c c
c c
c c
c c
bca acb
cab bca
bac bac
acb cab
bca cba
abc bca
cba cab
bca acb
abc abc
cab acb
d d
d d
d d
d d
d d
d d
d d
d d
d d
d d
cadb cbda
dcab bade
bacd cdba
dcab abed
dbac abed
dabc bdac
cbda dcab
abdc bade
cdba dcab
deba abed
e e adbce debca
e e dbcea daceb
e e edeba dacbe
e e adebe abdee
e e bdeae daecb
e e bcead cadeb
e e adecb ceadb
e e cbdea abced
e e edeba baced
e e adebe bedca
POSTTEST: acdabedbec
