.‘ﬁ
4, 1.35»
II‘

.

ll.

‘

inﬁrm. .

v.12

....

:9 A
sligl.‘
.53....1...

I.

 

I4 ‘ .
.imswﬁuua

a?

\r x, .

. ‘Dbsf . A".

.u. Emuuiiﬁuﬁ...

t ,t.‘ . .
qua-ant}; a

tall: «2 .
"nub“.hnfmﬁ \
.,..;audﬂl- .conuﬁﬁ. «EN
{xi-Kai? .
Ail-RN.

     

 

 

 

lllllllllllllllllllllllllllllllllllllllllll

3 1293 02060

This is to certify that the
thesis entitled

Contingent- Negative Variation (CNV)
in Autistic Children ‘

presented by
Christel Beretta

has been accepted towards fulﬁllment
of the requirements for

__MA___degree in anuage
Pathology

Ma g ”(6/2 mmﬂl,

Major professor

Date 19/16199

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

PLMI IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECAU.ED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

MAP. 2 9 2005

 

331—0—9—2005

 

FEB: (L6 2067

 

 

 

 

 

 

 

 

 

 

 

 

woo cJClRCIDateDmpSS—p.“

CONTINGENT-NEGATIVE VARIATION (CNV)
IN AUTISTIC CHILDREN

By

Christel Beretta

A THESIS

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

MASTER OF ARTS
Department of Audiology and Speech Sciences

1999

ABSTRACT

CONTINGENT-NEGATIVE VARIATION (CNV)
IN AUTISTIC CHILDREN

By

Christel Beretta

To measure attention, three autistic subjects between the ages of 6 to 9 years with
moderately-severe to severe autism were tested on six experimental tasks using the CNV
(contingent-negative variation) paradigm. One methodological question and three clinical
conceptual questions were addressed. The methodological question, whether autistic subjects
were able to establish a CNV, was answered in the afﬁrmative. One of the autistic subjects
established a CNV while two autistic subjects produced large negative potential shifts.

Three conceptual questions addressed basic assumptions of the Affolter treatment
paradigm. The ﬁrst assumption is that the impaired child receives the input while being
guided to perform a task. This assumption was investigated in two experimental tasks. It
was found that one autistic subject established a CNV in both guided tasks, while two autistic
subjects established large negative potential shifts (NPS). The difference between the two
guided tasks was the novelty effect. The autistic subjects demonstrated a signiﬁcantly longer
duration and a signiﬁcantly larger area of the CNV or largest NPS in the novel task as
compared to the familiar task [T (N = S) = 0, p < .05].

Affolter and colleagues’ second assumption was that the input to the brain is the same
in the guided and unguided conditions. Four experimental tasks were compared to test
possible differences. For the novel task, no signiﬁcant differences were found between the

two conditions. However, the familiar task revealed more varied results. The Wilcoxon

matched pairs signed-ranks test did not ﬁnd any signiﬁcant differences for the duration and
average amplitude of the CNV / largest NPS. However, signiﬁcant differences were found
for the area of the CNV or largest NPS and the overall area of negativities [T (N = 5) = 0,
p < .05] suggesting that there was signiﬁcantly more cognitive activity recorded in the
unguided condition as compared to the guided condition.

Affolter and colleagues’ third assumption was that changes of brain activity will be
noticeable when the event stimuli vary in the amount of resistance. The effects of strong and
weak resistance were investigated in two experimental tasks. Although the Wilcoxon
matched pairs signed-ranks test did not reveal signiﬁcant differences between the two
conditions, some biases were observed. Two of the autistic subjects demonstrated greater
amplitude shifts relative to the hard ball when the soft ball was squeezed (weak resistance).
One autistic subject showed the same pattern in her second trial but not in her ﬁrst trial.

When the area of the CNV or largest NPS was compared to the overall area of
negativities recorded throughout the entire sampling time, signiﬁcant differences were found
for all tasks and conditions suggesting that there was more negative activity indicative of
cognitive functioning outside the conﬁnes of the CNV or largest NPS than within the
conﬁnes of the CNV or NPS. Hence, the measure of overall negativities recorded during the

entire sampling time holds some promise as an indicator of cognitive activity.

ACKNOWLEDGMENTS

Foremost, I would like to thank the chair of my committee, Dr. Ida Stockman, for her
unceasing support. I would also like to thank the committee members, Dr. Ernest J. Moore,
and Dr. Tom Carr, for their willingness to embark on a unique and untested research project.

My thanks extend to Diane Wilhite, the speech-language pathologist in charge of AI-
students in Lansing, the children in the AI room, who were willing to be my subjects, and
their parents for their support and participation in this research project. I would also like to
thank the principal of Riddle Elementary School, Mr. Weatherspoon, for his help. Last but
not least, I would like to thank my ﬁiend, Ewa Danielewicz, for her technical advice and
encouragement, and my family for their understanding.

This study was supported by the grant #61-6272 from the Faculty Outreach Research

Program for Michigan State University, Dr. Ida Stockman, Principal Investigator.

iv

TABLE OF CONTENTS

Page
LIST OF TABLES .................................................................................................................. ix
LIST OF FIGURES ................................................................................................................. x
CHAPTER ONE
Introduction
Theoretical Explanations of Pathology .................................................................................... 3
Physiological Theories ................................................................................................... 3
Autoimmune Responses Causing Autistic Symptoms ......................................... 3
Neuronal Organization in Brain Development ..................................................... 4
Psychological Theories .................................................................................................. 6
The Theory-of-Mind Model ................................................................................. 7
Abnormal Perceptual Processing of Nonverbal Interaction Experiences ............. 9
Theoretical and clinical assumptions .......................................................... 9
Behavioral and physiological evidence for the interaction hypothesis ..... 13
Evaluation of the Four Theoretical Perspectives .......................................................... 18
Statement of the Problem ....................................................................................................... 21
Description of Electroencephalography ....................................................................... 22
Ongoing EEG measurements to study attention in normals ..................... 24
CNV as an investigative tool to study attention in normals ...................... 26
EEG and Attention .............................................................................................. 27
Ongoing EEG-studies in impaired subjects .............................................. 27
CNV in impaired subjects ......................................................................... 28
Application of ERP/CNV to the Affolter Guided-Interaction Paradigm ..................... 29
Research Questions ................................................................................................................ 33
CHAPTER TWO
Experimental Methods and Procedures
Subjects ................................................................................................................... 35
Subject-Inclusion Criteria ............................................................................................ 36
Procedures for Subject Recruitment and Selection ...................................................... 37
Human Subjects Considerations ............................................................................................ 37
Data Collection ................................................................................................................... 37
Description of Tasks ..................................................................................................... 37

Task 1 — Push Button Task ................................................................................. 38
Task 2 — (Squeezing a Soft Ball—Unguided) and Task 3 (Squeezing a

Hard Ball—Unguided) ........................................................................................ 38
Task 4 — Squeezing a Soft Ball, Guided ............................................................. 39
Experimental Task 5 — Picking Up a Light Glass, Guided ................................. 39
Task 6 — Picking Up a Heavy Glass, Unguided .................................................. 40
Auditory Brainstem Response (ABR) ................................................................ 41
Administration of Tasks ............................................................................................... 41
Preparation of Subjects for Tasks ....................................................................... 42
Order of Task Administration ............................................................................. 43
Recording of Data ...................................................................................................... 43
Instrumental Recording of Subjects’ Responses to Tasks .................................. 43
Behavioral Observations ..................................................................................... 45
Data Analysis ................................................................................................................... 45
Dependent and Independent Variables ......................................................................... 46
Possible Differences Between Guided and Unguided Conditions ..................... 47
Possible Differences Between Stimuli Varying in the Amount
of Resistance ............................................................................................. 47
Research Questions .................................................................................................... 48
Elicitation of CNV in Autistic Subjects .............................................................. 48
Elicitation of CNV in the Guided Condition ...................................................... 48
Possible Differences Between the Guided and Unguided Conditions ................ 48
Possible Differences When the Even Stimuli Vary in the Amount
of Resistance ............................................................................................. 49
Statistical Analyses .................................................................................................... 49
CHAPTER THREE
Results
Research Question 1: Is it Possible to Elicit CNVs in Autistic Children? ........................... 52
Duration and Average Amplitude of the CNV/Largest NPS (Push Button) ...... 52
Area of the CNV/Largest NPS and Overall Negativities of Entire Sampling
Time (Push Button Task) .................................................................................... 57
Research Question 2: Eliciting the CNV in the Guided Condition, Is It Possible to
Elicit a CNV when the Autistic Subject is Manually Guided to Do an Event? ......... 58
Task 4 (Squeezing a Soft Ball — Guided) ................................................................... 60
Duration and Average Amplitude of the CNV/Largest NPS for
Task 4 (Squeezing a Soft Ball) .................................................................. 60
Area of the CNV/Largest NPS and Overall Negativities of the Entire
Sampling Time for Task 4 (Squeezing a Soft Ball) .................................. 60
Duration and Average Amplitude of the CNV/Largest NPS for
Task 5 (Picking Up a Light Glass) ............................................................................. 62
Duration and Average Amplitude of the CNV/Largest NPS for Task 5
(Picking Up a Light Glass) ........................................................................ 62

vi

Area of the CNV/Largest NPS and Overall Negativities of the Entire

Sampling Time for Task 5 (Picking Up a Light Glass) ............................ 65
Comparison of Task 4 (Squeezing a Soft Ball — Guided) and Task 5 (Picking
Up a Light Glass — Guided) ................................................................................ 67

Effect of Novelty on Duration and Average Amplitude in Task 4
(Squeezing a Soft Ball — Guided) and Task 5 (Picking Up a
Light Glass) ............................................................................................... 67
Effect of Novelty on Area and Overall Area of Negativities of
Task 4 (Squeezing Soft Ball — Guided) and Task 5 (Picking Up

A Light Glass — Guided) .................................................................... 68

Research Question 3: Guided Versus Unguided Condition .................................................. 68
Comparison of Task 2 (Squeezing a Soft Ball — Unguided) With Task 4

(Squeezing a Soft Ball —- Guided) ....................................................................... 69

Duration and Average Amplitude of the CNV/Largest NPS for Task 2
Squeezing a Soft Ball — Unguided) and Task 4 (Squeezing a
Soft Ball — Guided) ................................................................................... 69
Area of the CNV/Largest NPS and Overall Area of Negativities for
Task 2 (Squeezing a Soft Ball — Unguided) and Task 4

(Squeezing a Soft Ball - Guided) ............................................................. 71
Comparison of Task 5 (Picking Up a Light Glass - Guided) and Task 6
(Picking Up a Heavy Glass — Unguided) ............................................................ 72

Duration and Average Amplitude of the CNV/Largest NPS for Task 5
(Picking Up a Light Glass — Guided) and Task 6 (Picking Up a
Heavy Glass — Unguided) ......................................................................... 75
Area of the CNV/Largest NPS and Overall Area of Negativities for
Task 5 (Picking Up a Light Glass — Guided) and Task 6
(Picking Up a Heavy Glass — Unguided) ........................................... 76
Examining a Possible Novelty Effect Between Task 2 (Squeezing a
Soft Ball — Unguided) and Task 6 (Picking Up a Heavy

Glass — Unguided) ..................................................................................... 77

Research Question 4: Change of Resistance ......................................................................... 78
Comparison of Task 2 (Squeezing a Soft Ball — Unguided) and Task 3

Squeezing a Hard Ball — Unguided) ................................................................... 78

Duration and Average Amplitude of the Largest NPS for Task 2

(Squeezing a Soft Ball — Unguided) Versus Task 3 (Squeezing

a Hard Ball — Unguided) ........................................................................... 79
Area of the Largest NPS and Overall Area of Negativities for Task 2

(Squeezing a Soft Ball - Unguided) and Task 3 (Squeezing a

Hard Ball —- Unguided) .............................................................................. 80
CHAPTER FOUR
Discussion and Interpretation of Results
Applying the CNV as a Measurement Tool ........................................................................... 82
Addressing the Methological Aspect of the CNV ........................................................ 84

vii

Addressing the Clinical Conceptual Questions ............................................................ 88

Eliciting a CNV in the Guided Condition ........................................................... 88

Novelty Effect as a Factor in the Guided Condition ........................................... 90

CNV Differences between Guided and Unguided Conditions ..................................... 92

Change of Resistance ................................................................................................... 96

Probing Differences Between Autistic and Normal Subjects ................................................ 98
Research Question 1 -- Is it possible to elicite a CNV in the Normal Pilot

Subject under the task conditions of the autistic subjects of this study? ............ 99

Research Question 2 -- Is it possible to elicite a CNV in the guided condition:
Performance of normal pilot subject in the guided condition as compared

To the autistic group? ......................................................................................... 99
Research Question 3: Differences between the guided and unguided
Condition in the Normal Pilot Subject .............................................................. 102
Research Question 4: Change of resistance .............................................................. 102
Implications for Future Research ......................................................................................... 103
Conclusions ................................................................................................................. 104
Research Question 1: Is it possible to elicit CNVs in autistic children? ................... 105
Research Question 2: Is it possible to elicit a CNV when the autistic
Subject is manually guided to do an event? ...................................................... 105
Research Question 3: Is there a difference between the CNV elicited when
the autistic subject is manually guided and unguided to do an event? ............. 106
Research Question 4: Is there a difference between the CNV elicited from
autistic subjects when the event stimuli vary in the amount of resistance? ...... 107
Summary ................................................................................................................. 108
Appendix A ................................................................................................................. 111
Appendix B ................................................................................................................. 112
List of References ................................................................................................................ l 16

viii

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

LIST OF TABLES

Page
Dependent Measures of Largest NPS on Experimental Task 1
(Push Button — Unguided) ............................................................................. 53
Dependent Measures of Largest NPS on Experimental Task 4
(Squeezing A Soft Ball — Guided) Versus Experimental Task 5
(Picking Up a Light Glass — Guided) ............................................................. 59
Dependent Measures of Largest NPS on Experimental Task 2
(Squeezing a Soft Ball — Unguided) Versus Experimental Task 4
(Squeezing a Soft Ball — Guided) .................................................................. 7O
Dependent Measures of Largest NPS on Experimental Task 5
(Picking Up a Light Glass — Guided) Versus Experimental Task 6
(Picking Up a Heavy Glass - Unguided) ....................................................... 73
Dependent Measures of Largest NPS on Experimental Task 2
(Squeezing a Soft Ball — Weak Resistance) Versus Experimental
Task 3 (Squeezing a Hard Ball — Strong Resistance) .................................... 74

Pilot Data for Normal Subject ...................................................................... 100

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

LIST OF FIGURES

Page
C8051399, Event 4, Experimental Task 1, Push Button,
Unguided, Autistic Subject 1, Trial 1 ............................................................ 54
DB061199, Event 3, Experimental Task 1, Push Button,
Unguided, Autistic Subject 2, Trial 2 ............................................................ 55
CM061199, Event 3, Experimental Task 1, Push Button,
Unguided, Autistic Subject 3, Trial 2 ............................................................ 56
DB061 199, Event 9, Experimental Task 4, Squeezing Soft Ball,
Guided, Autistic Subject 2, Trial 2 ................................................................ 61
DB061199, Event 11, Experimental Task 5, Picking Up a Light
Glass, Guided, Autistic Subject 2, Trial 2 ..................................................... 63
CSO61199, Event 13, Experimental Task 5, Picking Up a Light
Glass, Guided, Autistic Subject 1, Trial 2 ..................................................... 64
CSO61199, Event 15, Experimental Task 6, Picking Up a Heavy
Glass, Unguided, Autistic Subject 1, Trial 2 ................................................. 66

CHAPTER ONE

INTRODUCTION

In clinical praxis, speech-language pathologists encounter children with pervasive
developmental delays who do not respond to traditional treatment methods. Such children
show multiple impairments crossing verbal and nonverbal domains (Powell & Bishop, 1992;
Scholl, 1981; Stark & Tallal, 1988; Swisher, 1994). These children may be nonverbal or
echolalic, their motoric behavior may be abnormal, their sensory and perceptional systems
may be strongly impaired, and their cognitive abilities may be very low. In the literature,
these children are described as having pervasive developmental delay (PDD) or being
autistic. Children along the autistic/PDD spectrum are considered to be more severely
impaired than children along the speciﬁc language impairment (SLI) and speciﬁc learning
disability (SLD) spectrum, who show more localized deﬁcits.

Classic children with SLI show deﬁcits in their oral language, whereas those
diagnosed with learning disability (LD) show a deﬁcit in learning one or more
cognitive/linguistic skills required for academic success in school. For years, it was assumed
that nonverbal performance was unimpaired in the latter two populations. The reason for this
assumption was that the IQ tests used to evaluate nonverbal performance relied heavily, for
instance, on simple visual-discrimination tasks, which can be accomplished by many
children with developmental delays (Johnston, 1988). More recent research ﬁndings

(Swisher et al., 1994) have indicated that typically developing children and those with SLI

differ in their responses to test items on nonverbal instruments, even when the overall test
result is in the average range for both groups.

Thus, not only do children along the PDD/autistic spectrum present an enigma to
therapists, but children along the SLI/LD spectrum do so as well. Still, the latter group can
respond with some success to traditional intervention approaches, which often require the
processing of auditory and visual information. In contrast, children along the PDD/autistic
spectrum exhibit verbal, nonverbal, motoric, and perceptual deﬁcits when their impairments
are severe. They represent the challenging cases in treatment because of their behaviors. One
moment they withdraw from touch in displeasure (tactual defensiveness), and the next
moment they might totally ignore being touched.

Similar phenomena have been observed when these children have been exposed to
visual and auditory stimuli in their environment (Ayres, 1972; Fisher, Murray, & Bundy,
1991; Rutter & Schopler, 1987). These children, for instance, cover up their ears when the
vacuum cleaner is switched on. However, when one speaks to them in a normal voice, they
often fail to notice that they have been spoken to. Yet, when the children’s hearing is tested,
it is within normal limits. When their eyesight is checked, it is usually keen. This inconsistent
reaction to sensory input has been puzzling to clinicians, parents, and researchers alike.
Children with autism are also known for their stereotypies, such as hand or ﬁnger ﬂapping, or
walking on tiptoes. When they are examined, the pediatrician may not ﬁnd any indication of
physical handicaps. They also look normal. Consequently, the behaviors of these youngsters
typically cannot be explained in terms of deﬁnite physical deﬁcits. What theory could

possibly account for the variety of deﬁcits exhibited by this population?

Theoretical Explanations of Pathology

Contemporary theories regarding the cause of autism fall into two basic categories.
The cause of autism is explained in terms of either physiological or psychological factors,
processes, or mechanisms. For example, Singh and co—workers (Singh, 1996) focused on the
cellular level by investigating the association between virus serology and autoimmune
responses. Waterhouse, F ein, and Modahl (1996) attributed the etiology to dysfunctioning of
neural mechanisms. The psychological theory put forward by Baron-Cohen (1995) explained
autism by using a "theory of mind" model, whereas Affolter and colleagues (Affolter, 1987;
Affolter & Bischofberger, 1993, 1996, in press; Affolter & Stricker, 1980) claimed that
abnormal perceptual processing of nonverbal interaction experiences is the root of the

problem.

Physiological Theories
Physiological explanations of autism are based on investigations of the organic
structures that appear to be responsible for aberrant behavior. In other words, the functions
of living organisms and the individual organs, tissues, and cells of which they are composed

are the area of study (Purves et al., 1995).

Autoimmune Responses Causing Autistic Symptoms

The hypothesis that a virus-induced autoimmune response may play a causal role in
the autistic syndrome has been put forth by Singh (1996), Warren et al. (1996), Warren and
Singh (1996), Singh, Lin, and Yang (1998), and Warren et al. (1995). This means that there is
reason to believe that autism is an autoimmune disease in which the body's immune system

attacks its own cells instead of intruder cells (viruses and bacteria). Singh and coworkers

reported an association between virus serology and brain autoantibodies in autism. They
conducted the ﬁrst serological study of the measles virus and herpes virus 6 (HHV-6) in
autistic and normal subjects.

The researchers found that positive measles or HHV-6 titers were related
autoantibodies -- in particular, to anti-MBP (an anti-myelin basic protein) in autistic children
but not in normal controls (Singh, 1996). The authors pointed out that the measles virus and
the I-II-IV-6 virus were linked to demyelination. Myelin plays an important role in the
transmission of neural impulses from the cell body of a neuron to its axon terminals, where a
neurotransmitter substance is released. The axon carries action potentials. Myelin forms a
sheath made of Schwann cells that is wrapped around an axon to insulate it electrically,
thereby increasing its rate of transmission of neural impulses.

Another point worth considering is that environmental factors such as exposure to the
measles virus or/and the HIV-6 virus should be considered as possible etiological agents
linked to autoimmunity in autism (Singh, 1996). It is not known how the association between

virous serology and autoantibodies might relate to the pathophysiology of the disorder.

Neuron_al Orgﬂization in Brain Development
Whereas Singh and coworkers conducted their investigations at the cellular level,
Waterhouse et al. (1996) looked at neuronal organization as a possible cause of autism. The
researchers identiﬁed four major dysfunctions of neural mechanism that could account for the
behaviors observed in autistic individuals: (a) canalesthesia, (b) impaired assignment of the
affective signiﬁcance of stimuli, (c) asociality, and ((1) extended selective attention.
Under canalesthesia, the authors referred to fragmentation of cross-modal information

processing for an ongoing event and long-term memory records for past events due to

islandization of hippocampal system neurons. The dysfunction is caused by too great a cell-
packing density that leads to component sensory records in the hippocampal system tissues to
be canalized. Consequently, sensory information is not normal because hippocampal mossy
ﬁbers synapse on "islandized" comu ammonis (CA3) pyramidal cells, and the incoming
information is not coded adequately.

Waterhouse et a1. (1996) deﬁned impaired assignment of the affective signiﬁcance of
stimuli as an inability to judge actions by other people on a scale of important to unimportant
(p. 458). This dysﬁmction is due to impaired amygdala function. The amygdala fails to
develop due to overcrowding with immature cells. As a result, there is an inability to assign
appropriate affective signiﬁcance to incoming novel and social stimuli such as human faces,
facial expressions, and nonverbal messages. The authors emphasized that the role of the
amygdala is to integrate memory-modulating effects of neural hormone systems activated by
learning experiences. Lesions of the amygdala in humans have been shown to result in
abnormal affect, impaired face recognition, and impaired memory for the emotional content
of stories.

Under asociality, Waterhouse et al. (1996) deﬁned the dysfunction of social bonding
due to aberrant ﬁmctioning of the neuropeptide system involving oxytocin and vasopressin.
The researchers stressed accumulating evidence suggesting that oxytocin and vasopressin
participate in a mammalian neurophysiological network that regulates social affiliation and
elaborated imprinting. The authors hypothesized that there might be two possible sources of
oxytocin and vasopressin system dysﬁmction: (a) Impairment of oxytocin and vasopressin
receptors in the amygdala and hippocampus following abnormal neuronal development in the
medial temporal lobe, and (b) disruption of oxytocin- and vasopressin—releasing neurons early

in development.

These assumptions have been supported by other researchers. Panksepp (1992)
suggested that some motor stereotypies in autism may result from abnormal levels of
oxytocin. Other researchers have stated that 25% of autistic individuals have elevated
serotonin (5-HT) levels, which contribute to abnormalities in mood (Cook, 1990; Gillberg &
Coleman, 1992; Waterhouse et al., 1996). According to Watherhouse et al., a certain number
of autistic individuals demonstrate elevated beta endorphin levels, which reduce the reward
value of sucking, satiety, and touch in nursing. These levels can also result in an abnormally
high pain threshold.

The fourth dysfunction refers to the increased selective attention observed in autistic
individuals. That is, they show decreased ability to shift attention (Waterhouse, et al, 1996).
Inadequate temporal and parietal polysensory association cortex cells might trigger an
extended selective attention loop in primary processing, according to the authors. Zeki
(1993) provided evidence to support the hypothesis of extended selective attention in autism.
He stated that properties of polysensory association cortex cells may be generated by local
circuits or convergence. Zeki put forward the idea that the association cortex cells were new
in evolutionary terms, fewer in number than the cells of the primary cortices, and more
vulnerable. He pointed out that abnormally organized polysensory association cortex cells
would disrupt formation of complex representations and increase reaction time to complex

environmental stimuli.

Psychological Theories
Psychological theories of interest to this study have focused on explaining how the

mind (our behaviors, experiences, personal identity) works and how it relates to the physical

structure (the brain and the nervous system) that are ultimately responsible for many human

behaviors observed in everyday life.

The Theory-of-Mind Model

Supporters of the theory-of-mind model have argued that a modular cognitive
capacity underlies the ability to engage in complex social interactions. Baron-Cohen (1995)
and Stone, Baron-Cohen, and Knight (1998) proposed that theory of mind is a complex,
multicomponent, cognitive ability that is a distributed circuit involving many regions of the
cortex rather than a unitary module localized in a small region of the cortex. The structures
presumably involved in this circuit are amygdala, orbito-frontal cortex, and superior temporal
sulcus. Baron-Cohen (1995) explained the autistic syndrome in terms of theory of mind,
meaning that autistic individuals lacked the ability to make inferences about other people's
mental states.

Baron-Cohen (1995) tried to discover why, in particular, autistic children have
extreme difﬁculties forming mental representations. He described the autistic deﬁcits in
terms of "mindblindness." Hypothesizing that mindreading is enabled by the language
faculty. The drive to inform, to exchange information, to persuade, or to ﬁnd out about other
people's thoughts, is based principally on mindreading. Unless language is linked to the
mindreading system, language may be seldom used -- at least, not for social purposes. It is
well known that autistic children are unable to make the associations necessary for normal
social and language development.

Stone et al. (1998) provided evidence for Baron-Cohen’s theory. Their analysis of
various lesion studies indicated that damage in the orbito-frontal cortex, superior temporal

sulcus, and amygdala causes symptoms very similar to the ones observed in autism. Baron-

Cohen and coworkers compared patients with bilateral orbito-frontal lesions to individuals
with Asperger's syndrome (a mild form of autism) on a series of advanced theory-of-mind
tasks, such as ﬁrst- and second-order make-believe tasks and faux pas tasks.

The researchers found that the patients with bilateral orbito—frontal lesions performed
similarly to individuals with Asperger's syndrome. They performed well on simpler tests such
as ﬁrst— and second-order false-belief tasks, but they encountered difficulties with higher
level tasks, such as recognizing a faux pas, which required more subtle social reasoning.
According to Stone et al. (1998), two things are necessary for an individual to detect a faux
pas:

1. He or she has to understand that one person has knowledge that another person is
unaware of or that an individual has a mistaken belief.

2. He or she needs to have an empathic understanding about what kinds of things
another person would ﬁnd upsetting or insulting.

Individuals with autism can understand other people’s emotions, provided those
emotions are caused by situations or desires (Baron-Cohen, 1991). However, emotions that
are caused by beliefs are incomprehensible even for high-functioning autistic individuals.
This deﬁcit may be due to problems connecting mental inferences with an understanding of
emotion. These assumptions have been supported by research conducted in patients with
orbito-frontal and ventromedial brain damage. Brothers and Ring (1992) found that the
amygdala and orbito-ﬁ‘ontal cortex are essential brain structures for theory of mind as they
are involved in interpreting the valence and signiﬁcance of other people's actions and

intentions.

Abnormal Perceptual Procesm Nonverbal Interaction Experiences

Verbal and nonverbal deﬁcits in children along the PDD/autistic spectrum and the
SL1 spectrum have been explained in terms of abnormal processing of perceptual information
at the nonverbal level. Affolter and coworkers (Affolter, 1987; Affolter & Bischofberger,
1993, 1996, in press; Affolter & Stricker, 1980) assumed that nonverbal interaction with the
physical environment in daily problem-solving events is the root of cognitive and linguistic
development.

Theoretical and clinical assumptions. In theoretical terms, Affolter and colleagues
assumed that a baby’s nonverbal interaction with the environment is the foundation on which
perceptual organization and knowledge of the world rest. Nonverbal interaction occurs when
one’s action changes some aspect of the physical environment and, in turn, causes an
adaptive response by the actor or agent in response to the change in the environment. For
example, one may have to change body posture to accommodate to the new situation. Thus,
interaction is reciprocal; it alters the person and his or her environment simultaneously. Thus,
one comes to experience that “there is a world that resists the body when it is touched”
(Affolter et al., 1996, p. 2).

When the body touches its environment and the various objects, people, and animals
in it, the nervous system is assumed to register changes in resistance. The infant’s adaptations
to new situations are also experienced in the form of changes in resistance. From interaction
experiences, the nervous system knows, for instance, how hard or soft a ball can be. This
kind of tactual information is presumed to have been coded on a scale of maximal to minimal
resistance.

Affolter and colleagues concluded from these observations that the nervous system

stored tactile information in the form of changes in resistance offered by the various objects

and people in the environment. They assumed that the nervous system created a scale of
resistance that would allow the individual to relate patterns of resistance to certain tactual
experiences. In child development, these tactual explorations are observed early, long before
the child starts to speak

The following example explains the notion of interaction as described by Affolter and
colleagues. A 5-month-old baby is sitting in his high chair, and an orange is on the support.
The baby (actor) reaches out to touch the orange, and the orange moves around in response to
the baby’s touching and exploring. The orange rolls farther away from the baby. The baby
has to adjust his own posture to the new situation in order to keep track of the orange. The
baby accommodates to the new situation. Then the infant rolls the orange over the edge of
the support. The orange falls to the ground and is out of the baby's sight. The baby interacted
with his environment and achieved a topological change (which means a change of the baby's
body in relation to the objects that form his environment). Grasping, taking off, and putting
on, for instance, are topological changes. Once the orange was gone, the child related to the
new situation.

The interaction event has all of the primitives for organizing a perceptual system, as
assumed by Affolter and colleagues. When interacting with the environment, the baby
receives tactual input (Primitive 1) that is coordinated with other sensory systems, such as
vision, audition, taste, and smell. During an interaction event, the sensory systems can be
correlated so that the tactual input is connected to other systems. Thus, multimodal
connections (Primitive 2) of perceptual information can be established. When a baby
performs an interaction, there will be a sequence in which the individual actions take place.

The event sequence will enable the child to organize the perceptual information serially

10

(serial organization — Primitive 3). In solving daily life problems, the child will be exposed to
all necessary sources of perceptual information, and that is where learning takes place.

Another important consequence or effect of interaction -- apart from its role in
perceptual organization -- is knowledge about causes and effects. To give an example, the
baby rolled the orange over the edge of the support (cause), and the fruit disappeared (effect).
By having a multitude of these interaction experiences, the baby gains knowledge about its
own body and about the physical world around it. Knowledge of causes and effects supports
conceptual development, which frames the semantic foundation of language and adaptive
problem solving in daily-life events.

In clinical terms, Affolter and colleagues applied their interaction hypothesis to
diagnosis and intervention. They believed that severe developmental problems result from
abnormal interaction experiences. These abnormalities are assumed to be rooted in central
nervous system pathology that prevents the child from sensing enough input, connecting the
input with the modalities, or organizing it serially. Hence, there are three diagnostic groups:

1. Children who are unable to interact with their environment to extract the necessary
information due to Tactile-Kinesthetic (T-K) deﬁcits. This group of children typically learns
best when the environment is changed so that they can receive adequate T-K input by feeling
more resistance. Stockman (in press) suggested that these children behave like those in the
United States who receive a diagnosis of PPD or high-functioning autistic children.

2. Children who can extract the necessary information from the environment but who
are unable to process more than one modality at a time. These youngsters fail to make
intermodal connections. Stockman’s (in press) review of the literature indicated that children

with classic autistic symptoms show intermodal difﬁculties.

ll

3. Children who can perceive tactual, multistimuli and, as far as serial and temporal
aspects are concerned, can process more than one modality at a time, but who fail at
sequencing the information. This group of children is unable to serially organize the input
they perceive. As a result of this deﬁcit, these children face storage and recognition
problems. According to Stockman (in press), this subgroup behaves like children in the
United States who are diagnosed with SLI.

The subgroups stated under items 1 and 2, above, children with problems extracting
the necessary tactile information from their environment and coordinating their intermodal
information, were of particular interest in this study. These groups are the most severely
impaired, in clinical terms.

If failure to interact is the cause of the deﬁcits due to a perceptual problem, then
treatment must facilitate interaction. To help the impaired child interact with the
environment, Affolter and colleagues recommended guiding the child through a problem-
solving event that is meaningful to the child. In other words, a clinician moves the child’s
body with him or her, not for him or her. They hypothesized that the input to the brain is the
same when the child is manually led to solve a problem as when the child solves the problem
independently. The focus on facilitating nonverbal interaction experiences in problem-
solving situations is assumed to be relevant to both nonverbal and verbal development.

Affolter and colleagues’ theory is also based on functionality. They assumed that
while the child is being guided (hand over hand), the child will make some of the necessary
associations required for concept formation. They claimed that the input will get to the brain
whether the child effects the interaction on his or her own or is being guided to do so. They

stressed that one moves the child’s body with him or her and not for him or her.

12

Just touching objects is not necessarily beneﬁcial to learning. Severely impaired
autistic children can often be seen running around a room and touching everything they can
get their hands on, but the children will not interact with the touched objects or people in any
way. Affolter and colleagues pointed out that these children need to be guided to perform a
frmctional goal-directed task. A functional goal-directed task would be, for instance, opening
a milk carton. A cause-effect relationship is preserved in this task (you open the milk carton
and you can drink milk).

The child will sit down at a table with good support (feet touching the ground, arms
lying on a ﬁrm surface). The therapist will help the child open the milk carton, ﬁnd the straw,
put it in the carton, and drink the milk. This activity is meaningful to the child. There is a
beginning, middle, and end to this particular task. The random touching seen before has been
converted into goal-directed touching and learning, such that concept formation (open, drink)
can take place.

Moral and physiological evidence for the interaction hypothesis. Support for the
guided-interaction framework comes from Affolter’s own behavioral research and other
supportive studies. As summarized in Affolter and Bischofberger (in press), this research has
involved several comparative behavioral studies of normal children and children with severe
developmental disorders, including autism. They compared sensory-deprived (blind or deaf)
subjects, language-impaired (including autistic) subjects, and normal control subjects in
cross-sectional studies of ages ranging from 4 to 21 years, on form and pattern-recognition
tasks, serial tasks, and classic problem-solving tasks. All tasks allowed them to compare
tactual, visual, and auditory input (Affolter & Bischofberger, in press; Affolter & Stricker,

1980)

13

The learning and visually impaired groups were able to develop cognitively after
showing initial developmental delays. The language-delayed group and the autistically
impaired subjects, on the other hand, failed to show the same developmental pattern as the
sensory-deprived subjects. They continued to fall behind in their cognitive development. In
particular, subjects with severe impairments did more poorly on tactual tasks than did normal
children and those with sensory deﬁcits. Based on these ﬁndings, the authors suggested that
tactual input was basic to perceptual organization and deve10pment.

In a longitudinal study (Affolter & Bischofberger, in press; Affolter & Stricker, 1980),
the researchers found that language-impaired subjects (autistic subjects included) did not
demonstrate nonverbal interaction performances in daily problem-solving activities as
comparable to the sensory-deprived and normal groups. The language-impaired as well as the
autistic subjects interacted with their environment to a lesser degree than the sensory-
deprived groups. If problem solving occurred in the language-impaired group, it was done in
a peculiar way. Usually it was not done at all.

The most striking ﬁnding was that “event representation” had to be established before
true language would emerge. Consequently, Affolter and coworkers assumed that the
common link between verbal and nonverbal performance was successful problem solving at
the nonverbal level.

The theory stated by Affolter and colleagues was strongly inﬂuenced by Affolter’s
mentor, Jean Piaget, who often is considered to be the ﬁrst contemporary scholar to focus on
physical interaction as an important construct in developing cognition. Piaget considered
assimilation and accommodation to be reciprocal mechanisms of developmental change that
resulted in a cause-effect understanding of the world. This basic understanding would form

the basis for cognition. In his writings, Piaget (1977) emphasized the importance of these

14

early interactions with the enviromnent in the “sensory-motor stage” of development. Piaget
proposed that the child had to acquire a number of sensory-motor skills before language
could develop.

Evidence in the literature supports some of the assumptions made by Affolter and
colleagues about T-K deﬁcits in children with developmental delays. Stockman (in press)
cited studies that have shown a relationship between sensory-motor deﬁcits and cognitive-
linguistic delays. Grandin (1995) described in her book her own autistic sensory problems as
a child and the remedies she created herself. Grandin would crawl into a cattle chute at her
aunt's ranch to relax by feeling more pressure against her sides and neck. Later she built her
own machine to experience more pressure. Affolter and colleagues believe that individuals
with perceptual disturbances need to feel more resistance than those without such
disturbance.

There is also evidence in the literature that children with severe developmental delays
may have intermodal-processing deﬁcits, as Affolter and colleagues have claimed. When
F rith and Baron-Cohen (1987) reviewed a variety of studies that compared autistic children to
mentally retarded and normal children, they concluded that at the sensory, lower processing
levels, no deﬁcit in any modality appeared to be speciﬁc to autism. The autistic subjects
showed insufﬁcient coordination with other senses in time and space. This observation was
supported in a 1969 study by Frith and Hermeline (as cited in Stockman, in press). The
researchers demonstrated that autistic subjects performed as well as normals when they just
visually tracked a display and even better than normals when they just tracked from feeling.
These children were less successful when they had to look and feel at the same time. Such
ﬁndings support the assumption made by Affolter and colleagues that autistic children have

problems on the intermodal level of organization.

15

There is a rich body of contemporary literature describing how the normal child
learns. Part of learning is concept formation. Medin (as cited in Booth, 1998) suggested that
concepts and categories serve as building blocks for human thought and behavior. Booth
investigated the role of functional information in the development of object concepts in
infancy. She tested 14-month-old infants by applying, among other things, the operant-
discrimination learning paradigm. Infants were given either static, dynamic, or functional
experience with exemplars. Infants in the static condition observed stationary pairs of
exemplars. In the dynamic condition, they observed the exemplars moving in category-
speciﬁc ways. In the ﬁmctional condition, the infants produced the motions themselves. The
operant-discrimination learning paradigm revealed that infants in the functional condition not
only were more likely to learn, but also they learned faster than they did in the static or
dynamic conditions.

Besides behavioral evidence for the claims of Affolter and colleagues, there is a
physiological basis for them as well. In traditional intervention paradigms, T—K input as a
possible source of learning has been widely neglected. One reason for this is that research in
exploring the senses has focused primarily on vision, the auditory system, taste and smell.
Affolter and colleagues are among the few contemporary scholars who have placed
importance on T-K input for development. The nervous system of a new-born child is still
immature in many ways; various processes and neurological mechanisms need to be
established before, among other things, normal language development can be triggered.

Because the nervous systems of mammals are very similar to that of humans, a study
by Wallace et al. (1997) is of interest. They investigated the development of multisensory
neurons and the multisensory integration in the superior colliculus of cats. They found that

although a high proportion of multisensory neurons and a high level of multisensory

l6

integration were found in the deep layers of the superior colliculus of adult cats, there were
no such neurons during the ﬁrst period of postnatal life in kittens. All sensory-responsive
neurons were rmimodal. This means that experiences and a functioning nervous system are
necessary to establish the connections that enable adult cats to be successful as a species. One
can assume that human infants explore their environments with the help of their senses. A
well-frmctioning nervous system will form intramodal and intermodal connections that are
necessary for development.

The roles of active touch and perception in one’s exploration of the environment have
not been fully explored. Bonda et al. (1996) investigated the neural systems involving
memory processing of experiences through touch. The researchers found that a ventrally
directed parietoinsular pathway, leading to the posteroventral insula and perirhinal cortex,
constitutes a system by which long-lasting representations of tactual experiences are formed.
The perirhinal cortex was found to be further involved in the integration of long-lasting
representations of somatosensory experiences. This study indicated that certain pathways and
brain areas are allocated to somatosensory experiences and representations.

Carvell and Simon (1996) investigated abnormal tactile experiences early in life in
rats by trimming their whiskers. They found that the deprived rats were severely impaired in
their ability to distinguish between smooth and rough surfaces. The rats' sensorimotor
integration underlying active touch was substantially and perhaps permanently impaired.

Klatzky (1987) looked into the salience of object attributes for haptics with and
without vision. The results of her study supported the contention that haptic and visual
systems have distinct encoding pathways, with haptics oriented toward the encoding of
substance rather than shape. This may reﬂect a direct inﬂuence of haptic exploratory

procedures.

17

Evaluation of the Four Theoretical Perspectives

Theory should account for behaviors in normal and impaired populations so that a
better understanding of developmental principles of hmnan learning can be achieved and
intervention principles formulated. To account for the behavior of children with autism, one
needs to look at how theory can account for both nonverbal and verbal cognition, inclusive of
social/pragmatic deﬁcits.

Physiological theories have been formulated to explain the cause of autism by
pointing to neurological structures that may be involved in the behaviors exhibited by autistic
persons. These theories fall short in that they are not aimed at capturing broad developmental
principles of human learning. It is not clear whether treatment strategies can result from the
ﬁndings of research based on such theories. Singh's (1966, 1997, 1998) work at the cellular
level might pave the way for the development of some future drug or vaccine. His line of
thinking is clearly promising with regard to treatment outcomes in the future, although it
leaves the clinician empty handed for the time being. Waterhouse et al. (1996) focused on
identifying structural and organizational differences that might explain the deﬁcits observed
in autism. Their ﬁndings on vasopressin and oxytocin might lead to medical intervention at
some time when neurotransmitters and hormones, such as vasopressin and oxytocin, are
better understood. Again there is some promise for the future. For the time being, though,
their theory also will leave the clinician empty handed.

Both psychological theories have shortcomings, as well. Baron-Cohen's theory-of-
nrind model pertains to a speciﬁc deﬁcit (lack of mindreading ability) that, in turn, is
responsible for language deﬁcits. He asked relevant questions and looked for answers to
them. He did not investigate perceptual mechanisms speciﬁcally, nor did he explore the issue

of verbal and nonverbal performances. His theory does not provide any direction for

18

intervention, rather it is a theoretical model looking at autism primarily from a theory-of-
mind perspective while touching only up on the mind-brain relationship. In addition, by
reducing the autistic deﬁcit to a lack of mindreading ability, researchers using this theory
address only a mild form of autism such as Asperger’s syndrome. Supporters of a theory-of-
mind model have not offered any explanations for the pervasive developmental deﬁcits that
more often characterize children with autism.

The theory stated by Affolter and colleagues, on the other hand, is the only one that
casts a broad explanatory net in ways that connect to more general principles of learning and
normal development. Diagnostic and treatment practices follow directly from their theoretical
assumptions. Still, this theory is not well known, and its basic tenets go against the grain of
most existing treatment frameworks. This is because the focus of their theory and treatment
principles on physical interaction with the environment implies a sensory-motor orientation
to therapy. Sensory-motor approaches developed by occupational and physical therapists have
emerged, but these approaches remain an undercurrent in rehabilitation (Stockman, in press).
In addition, there is no evidence that the few sensory-motor approaches that have been
developed by occupational and physical therapists actually work.

One reason why sensory-motor approaches are missing in contemporary treatment is
the waning inﬂuence of Piaget's theory on cognitive development, in which sensory-motor
performance is foundational. Although some assumptions remain that do support Piaget's
theory and reemerged in the new wave of cognitive constructionism (Langer, 1994; Nelson,
1986; Thelen, 1995), it is generally neither understood nor accepted that tactual input is
important to essential aspects of cognitive and linguistic development. A lack of knowledge
about the tactual system as compared to audition and vision is another reason why any

paradigm involving sensory-motor mechanisms is not well received.

19

Documentation and/or testing of the efficacy of the guided-interaction treatment
paradigm developed by Affolter and colleagues is needed. Until now, a major source of
support for this treatment paradigm has been clinical observations substantiated by video
archives. Children and brain-injured patients are videotaped when therapy is initiated, and
patients’ progress is documented in the course of treatment. For instance, at the School for
Perceptually Disturbed Children in St. Gallen, Switzerland, clinicians are required to write
down their observations during treatment. During a guided-interaction task, the child’s
behaviors are documented. For example, the clinician writes down when the child is
attentive, and the conditions under which attention is sustained. The clinician notes progress
in solving problems of daily life -- for instance, if the child anticipates the next step in the
task, if the child takes over and continues performing the task independently, how often the
clinician has to redirect the child’s attention to the task, and so on.

It is well established that clinical efficacy studies, the approach by Affolter and
colleagues included, are plagued by a lack of control. There are many variables that affect the
internal and external validity of such research. These include clinician bias and maturation
effects. In view of these difficulties with documenting progress and establishing valid criteria,
support for treatment outcomes from a variety of different sources is desirable.

Encephalographic studies are a valuable addition to traditional empirical
investigations in the ﬁeld of speech-language pathology. Over the past decades, the number
of electrophysiological studies conducted to investigate physiology and human behavior has
increased considerably. These studies allow researchers to correlate brain activity with
cognitive functioning, and they have the potential to be a valuable additional tool in assessing

treatment efﬁcacy for reasons stated earlier.

20

Another important factor inﬂuencing the present researcher to undertake this study is
that the premises by Affolter and colleagues are not well accepted (for reasons described
later). By measuring actual brain activity while testing Affolter and colleagues' assumptions,
this investigator might be able to shed some light on the validity of their theory with a

different kind of measurement than solely video observations.

Statement of the Problem

The present investigation was motivated by the advantages of studying mental
processing in relationship to measures of brain activity. Neuropsychological research that
involves relationships between mental processing and brain activity is potentially useful for
informing both physiological and behavioral theories of causation for two reasons. The ﬁrst
one is obvious; it allows the investigator to combine physiology with psychological behaviors
-— something that is missing in contemporary psychological and physiological theories of
autism, which concentrate on one dimension or the other. Although brain/behavior studies
cannot yield theory in and of themselves, they can provide some support for or refute critical
assumptions on which a physiological or psychological theory is based.

The second reason is that measures of brain activity appear to be less sensitive to
cultural differences as cognition is being measured in the form of electrical activity and not
with formal or rigidly standardized test instruments, which often are culturally biased. The
kinds of mental processes that can be measured, are considered.

Attention is a prerequisite for learning. That is, if a child can pay attention for a
certain period of time, he or she has a chance to learn some aspect of a task. When a child is
unable to pay attention, it is important to enable him or her to do so. Getting the child's

attention is an important goal when treating youngsters with severe deﬁcits along the

21

PDD/autistic spectrum. Hence, one criterion for evaluating the efficacy of any treatment
paradigm is its capacity to motivate attention to the task. In the guided-interaction paradigm
designed by Affolter and colleagues, attention is the consequence of information received by
the child. For impaired persons within this framework, the type of input needed to get the
attention of the nervous system is anchored by tactual input.

On the one hand, the investigator focused on ﬁnding an electrophysiological
measurement that would give some indication of whether a child is attentive or inattentive;
on the other hand, the investigator wanted to observe brain activity in different conditions
while sensory input was given involving the tactual system. Another consideration was to
ﬁnd an electrophysiological measurement that normal and autistic children could tolerate.
Electroencephalography (EEG) turned out to be the measurement of preference for the

present study.

Description of Electroencephalography

The EEG is composed of electrical rhythms and transient discharges, which may be
distinguished from one another on the basis of location (brain region), frequency, and
functional properties. The alpha band has a range of 8 to 13 cps, and the beta band's range is
14 cps. and above. The theta band ranges from 4 and 8 cps, and the delta band's range is 0.5
to 3 cps. Wavelength is employed to indicate the duration of a single event. Functional
properties of cerebral electrical events are those variations that occur in relation to a shift in
the level of consciousness (change in the type of mental activity).

The ongoing EEG records spontaneous brain activity driven by internal subject
responses. Historically, ongoing EEGs have been used to compare normal with abnormal

brain activity, as will be described later. To record brain activity in response to speciﬁc

22

stimuli, the evoked-response paradigm was developed. These stimuli are associated with
changes in the EEG that are called event-related potentials (ERP). Evoked potentials (EP) are
signals that are generated by neural populations that become active when tirnelocked to the
stimulus. This signal is summed to the ongoing EEG activity (Lopes da Silva, 1993). Time
analysis of the EP is based on two assumptions:

1. The electrical response evoked from the brain is invariably delayed relative to the

stimulus.

2. The ongoing activity is a stationary noise. The detection of the EP becomes a

question of improving the signal-to-noise ratio.

One extensively studied ERP is the contingent—negative variation (CNV), ﬁrst
described by Walter (1964). The CNV is picked up from a background of raw EEG activity.
It consists of slow surface negativity that depends on the association of two successive
stimuli. A ﬁrst stimulus (81) serves as a preparatory signal for an "imperative" stimulus (82),
to which a motor response is made. The time interval between S] and S2 is usually 1 to 2 s
(Tecce & Cattanach, 1993). Present responses are recorded and averaged. Positive and
negative components make up the evoked response, whereas the negative component is
indicative of cognitive/mental activity. McCallum (1988) and Tecce and Cattanach (1982)
suggested that the CNV is a potentially useful measure of brain-behavior functions,
particularly psychological processes.

A sawtooth waveform is typical of the CNV that rises steadily until the termination
takes place during the positive phase of the evoked potential to the imperative stimulus
(Reneau & Hnatiow, 1975). Cohen (1969) indicated that 40% of the adults he tested
produced CNVs having a ramp shape and 33% produced rectangular CNVs. The remaining

subjects produced mixed and atypical CNVs. The shape of the CNV depended on the

23

subject’s uncertainty with the prediction. If the subject was certain when the response
stimulus would occur, the CNV showed a ramp-like form. Ifthe uncertainty was total, the
CNV showed a square form. If uncertainty was reduced, the CNV showed reduced amplitude
(Reneau & Hnatiow, 1975).

Ongoing EEG measurements to study attention in normals, In whole-scalp EEG
studies, brain activities in the various bandwidths (such as alpha, beta, theta, and delta) have
been related to cognitive processes such as attention. The studies discussed in the following
paragraphs have demonstrated a relationship between electrical activity in the various
bandwidths and cognitive functioning and, in particular, attention.

Researchers in Austria (Klimesch et al., 1996) used EEG and evoked potentials to
study attention and memory in a normal population. Klimesch et al. demonstrated that the
upper 9.1M band is sensitive to semantic memory, whereas the lower alpha band seems to
reﬂect attentional processes. The authors assumed that EEG frequencies from the alpha
bands originate from the thalamus and that activity of thalamocortical networks reﬂects
processes that are related to searching, accessing, and retrieving information from long-term
memory.

In general, the bits! bands in the range of 20 to 40 Hz are linked to diffuse arousal and
focused attention (Steriade, 1993). The fast-frequency bands have been observed, for
instance, in the occipital cortex of a dog while the animal paid intense attention to a visual
stimulus (Lopes da Silva, 1970). Sheer (1984) observed the same waves in humans during
focused arousal before they performed a complex task. These particular beta bands are
linked to the arousal system of the reticular formation that projects to the thalamus, where
axons are found that directly innervate the cerebral cortex (Saper, 1987). The beta bands were

also the focus of research by Pﬁrrtscheller, Stancak, and Edlinger (1997). They recorded from

24

sensorimotor areas during turilateral self-paced brisk and slow ﬁnger movements. The
researchers found two different beta-components, whose function they believed to be a
correlate of active inhibition of idling of the primary motor area following movement
execution. The fact that the primary motor cortex is inhibited after execution of a motor task
suggests that the resources allocated to one task can be transferred to another task. As will be
described later, cognitive processes, such as paying attention, are possible only when certain
areas of the brain are inactivated and the freed capacity is allocated to the task. If certain
sensory and perceptual inputs cannot be blocked out, the system will overload and cease to
function adequately.

Harmony et al. (1996) investigated whether an increase in delta EEG activity during
mental tasks was related to an increase in the subjects’ attention to internal processing. The
researchers showed an increase in delta power from 1.56 to 3.90 Hz only during the difﬁcult
mental-calculation task, not during the control task. On the basis of these ﬁndings, they found
evidence for their hypothesis.

Klimesch, Doppelmayr, Russegger, and Pachinger (1996) investigated the role of
t_h_e_t_a bands during an encoding task. The researchers found that theta power increased during
successful encoding of new information, whereas alpha power decreased during encoding.
Signiﬁcantly higher theta power was noted during the encoding of words that could be
remembered in the later recall task, compared to those words that the subjects failed to
remember in the later recall task (Klimesch, Doppelmayr, Schimke, & Ripper, 1997). The
researchers believed that theta power was induced in the cortex through hippocampo-cortical
feedback loops. Theta power traditionally has been viewed as “noise” in the system.
However, the ﬁnding that alpha power decreased while theta power increased at a time when

words were learned might suggest a reciprocal relationship between the two band widths.

25

This reciprocal relationship might be indicative of cognitive processing and of attention. No
learning can take place without attention.

Burgess and Gruzelier (1997) looked at changes in human M rhythm during a
word-recognition memory task. Their results conﬁrmed that short-duration changes in human
theta rhythm were associated with recognition memory. Verbal versus visuo-spatial
recognition tasks were examined by Dujardin et al. (1995) and Rugg and Dickens (1982). The
researchers found that alpha power was signiﬁcantly lower during performance of both tasks
as compared to the rest condition. Theta power, however, was signiﬁcantly higher in epochs
recorded during the visuo-spatial as Opposed to the verbal task.

Weiss et al. (1995) examined EEG parameters that can be attributed to movements
performed with maximal subjective effort. They found that theta decrease possibly reﬂected a
down-regulation of a posterior attention system in order to minimize the inﬂuences of
external stimuli during the preparation for voluntary isometric contractions.

CNV as an investigative tool to study attention in normals. CNV has been used to
study cognitive functioning in normal adult subjects. It is generally accepted that larger CNV
amplitudes are correlated with attention to a task procedure. Differences in CNV amplitude,
on the other hand, may reﬂect changes in attention and sensitivity. McAdam (1969)
investigated the relationship between learning how to estimate a short interval of time and the
CNV. He found that an increase in CNV amplitude paralleled the actual learning of the
experimental task.

Tecce and Scheff (as cited in Reneau & Hnatiow, 1975) found greater amplitude of
the CNV when reaction times of subjects were faster. They concluded that the CNV and
reaction times were manifestations of attention. On the basis of his own ﬁndings, Donald

(1970) hypothesized that the CNV is a phenomenon that occurs during intense concentration.

26

EEG and Attention

Ongoing EEG-studies in impaired subjects. The ongoing EEG has been used as an
investigative tool since the beginning of the century. The studies discussed in this section
used ongoing EEG recordings of spontaneous brain activity in impaired subjects and
compared them to those of normal controls. The impaired group demonstrated reduced
activity in the alpha bands that are indicative of cognitive functioning, as described earlier.
Cantor, Thatcher, Hrybyk, and Kaye (1986) measured brain activity in low-functioning
autistic children with age-matched mentally retarded and normal subjects. The researchers
found that the autistic subjects exhibited signiﬁcantly more slow wave activity and less alpha
activity than the normal or mentally retarded children did.

Dawson, Klinger, Panagiotides, Lewy, and Castelloe (1995) investigated subgroups of
autistic children based on social behavior display and patterns of brain activity. Compared to
normally developing youngsters, the autistic children showed reduced EEG power in the
frontal and temporal regions, but not in the parietal region. Differences in the left hemisphere
were more prominent than in the right hemisphere. The researchers were able to establish
distinct patterns of brain activity of the children in the subgroups of autism. The subgroup of
"passive" children exhibited reduced alpha EEG power in the frontal region.

Gorbachevskaya et a1. (1992) studied, among other syndromes, infantile autism by
using EEG mapping. The Russian researchers included in this group infantile schizophrenia,
Rett's syndrome and fragile x syndrome. Their reason for including these three syndromes in
the infantile autism group was that they showed similar clinical pictures. The investigators
found three distinct patterns of brain activity among these three syndromes. The fragile x
group showed marked reductions of the alpha rhythm and predominance of a theta rhythm

with a frequency of 6 to 9 Hz and with maximal amplitude in the parieto-central region.

27

In the schizophrenic group, a reduction in amplitude of the EEG waves was noted,
with a very small increase in diffuse slow-wave activity and with inadequate representation of
the alpha rhythm. In the early stage of Rett's syndrome, a decrease in amplitude of the waves
and of a fragmentary alpha rhythm with a frequency of 8 to 11 Hz was noted. Through the
course of the disease, the abundance of the alpha rhythm decreased, foci of epileptoid activity
appeared, and the theta rhythm was considerably intensiﬁed.

Schneble (1984) compared the EEGs of autistic children and adolescents with those
of age-matched mentally retarded patients. The most striking ﬁnding of Schneble's study was
that the majority of the autistic population (67.5%) showed an abnormally accelerated
background activity. A mixture of fast alpha frequencies (11 to 12 s.) was found to be
dominant in the parieto-occipital regions bilaterally. Almost the same pattern was found in
the medial and frontal regions, with the only difference that the frequencies were more
widespread. On average, the autistic group showed less synchronization than the control
group.

As these studies have indicated, the impaired population shows reduced alpha
amplitude and/or inadequate representation of the alpha rhythm compared to the alpha
representation in normals. Of the autistic population tested by Schneble (1984), 67.5%
showed abnormally accelerated background activity.

CNV in impaired subjects. Only a few researchers have looked at the CNV in clinical
populations. Gloning, Burian, Gestring, and Haider (1970) studied the CNV with aphasic
adults. Some severe aphasics had test results similar to those of the control group, and other
severe aphasics had results that differed from those of the control group.

Walter (1966) found that it was difﬁcult for patients suffering from chronic anxiety to

develop a CNV. The same results were found when psychotic individuals were tested. They

28

had great difﬁculty establishing a CNV (Reneau & Hnatiow, 1975). Rugg et al., (1989)
recorded event-related potentials in a Go/No-Go reaction-time task from patients with closed
head injuries and normal controls. The researchers found abnormalities in those of the
clinical group but not those of the control group.

Cohen and Ofﬁrer (1967) observed CNVs in children as young as 6 years. Prevec and
Ribaric (1986) recorded a reliable CNV even in children as young as 3 years. Low, Coats,
Rettig, and McSherry (1967) found that CNVs in children do not return to the baseline
abruptly after the child makes a response. Their duration and amplitude are increased as
compared to those of adults. Cohen and Offner used the CNV paradigm to test children with
learning disabilities. The children in their study could not form CNVs.

Based on these studies, one can expect a more pronounced CNV in normal children as
compared to normal adults. There is also the possibility that the impaired population might

fail to establish a CNV.

Application of ERP/CNV to the Affolter Guided-Interaction Paradigm

The questions for the present investigation were motivated by methodological and
theoretical considerations. With regard to methodology, there is some doubt that CNVs can
be established in impaired children (as described earlier). In addition, even the literature on
CNVs in normal children is scarce, so there is a need to explore methodological
consequences of using the CNV in normal and clinical groups to gather further evidence.

With regard to theory, there appear to be four fundamental assumptions of the
framework put forward by Affolter and colleagues that can be tested using simple tasks in the

CNV paradigm. However, these assumptions are contentious on several fronts, and

29

answering questions pertaining to them would validate or invalidate the theoretical
framework that guides treatment.

The ﬁrst assumption is that the brain gets information during guiding. Although the
existence of the sensory system cannot be denied, there is a strong bias that guiding will not
stimulate the nervous system because it is allegedly passive.

Detecting a CNV, or negative potential shift (NPS) in the autistic subjects when they
are manually guided to perform a task would provide additional evidence for Affolter and
colleagues’ hypothesis that the input under guiding is received by the brain and actively
processed, although the subject is not performing the task independently. If increased brain
activity in the form of negative shifts or CNV is documented during guiding, Affolter and
colleagues’ hypothesis of working at the nonverbal level would also receive some support,
and clinicians who use the guided-interaction paradigm to enable severely impaired children
to interact with their environment and to solve functional tasks would have another piece of
evidence to demonstrate the efficacy of this type of intervention.

The second assumption is that an individual will receive the same input in both the
guided and unguided conditions. This is contended for the same reasons as the ones stated
above. Independent volitional movements are thought to provide the best source of input for
learning. This traditional view received support from the study by Carr et al., (1996).

Not many researchers have investigated guided versus unguided performance and the
resulting brain activity. Carr et al. (1996) examined somatosensory activity associated
speciﬁcally with motor control using functional magnetic resonance imaging (ﬂVIRI). The
researchers compared an active performance condition in which the subject moved his hand
independently with a passive sensory baseline condition in which the same movement was

produced by the experimenter manipulating the subject’s hand. The active performance

30

condition and the passive sensory baseline condition were compared to ﬁnd some evidence
for processes associated with endogenous motor control and motor execution.

Differences in the extent of magnitude of activity in the somatosensory cortex due to
somatosensory feedback were of particular interest to the investigators. They measured
activity in the somatosensory cortex, proximate parietal association areas, and motor cortex
during a task requiring cyclic ﬂexion and extension of the four ﬁngers of the right hand. The
subject himself carried out this movement, and then the experimenter moved the subject’s
ﬁngers in the same pattern and at the same pace.

The results showed that in the contralateral primary and supplementary motor areas,
larger numbers of activated pixels were found during the active performance task than during
the passive sensory baseline. A larger signal change occurred during active performance (9%
versus 5%) than during the passive sensory baseline condition. Increased somatosensory
activation was observed when endogenous control and active movement were required.
Increased sensory feedback and increased attention to sensory feedback were noted during
active performance. These ﬁndings suggest that independent task performance requires more
mental activation than the passive baseline condition. That is, there was a difference between
the guided and unguided conditions in the normal subject. The activation patterns were
stronger in the unguided condition.

However, what the activation patterns of an autistically impaired subject would look
like is still a question that needs to be answered. If an impaired child has to learn how to
organize the perceptual information during guiding, it is not unreasonable to assume that the
NPS would be even stronger under the guided condition as compared to the unguided
condition. If more CNVs or negative shifts are recorded in this study when the impaired

subject is manually guided to perform a novel task as compared to the condition where the

31

impaired subject performs the task independently, additional evidence will be provided in
favor of guiding.

A third assumption has to do with the role of resistance in tactual representation in the
nervous system. Affolter and colleagues asserted that the most elementary kind of
information needed to interact is the change of resistance one feels when touching objects in
the environment. In her view, then the child searches for resistance to organize the T-K input.

To cite Affolter's Oyer Lecture (1996),

“When I peel an orange, I know that the orange changes. I get information

about these changes. I assimilate them. At the same time I also change or I

have to accommodate my own movements in order to hold the orange and

continue to peel it. While touching the orange and moving my own body

through space, I also perceive the resistance offered by the environment.

Judging from the degrees of resistance experienced, for instance, I know how
much effort I have to use to grasp an object, or to push something out of my

99

way.

Hence, guided-interaction treatment places emphasis on resistance. If
resistance is a critical source of input, one should be able to detect changes in NPS on
the EEG recordings. These changes should be observable when changes in topological
relationships are correlated with changes in resistance.

Affolter and colleagues assumed that the various degrees of resistance play a key role
in coding T-K information. This hypothesis is also addressed in this study. Affolter and
colleagues believed that the nervous system has to exert more effort when an object is
grasped that offers weak resistance than when an object is grasped that offers strong
resistance. Ifa CNV of longer duration and higher amplitude is recorded in a condition where
a soft ball is squeezed (rather than a hard ball), this is evidence that more mental energy is
exerted to complete this experimental task. The individual has to search harder to perceive

the T-K feedback resulting from touching the soft ball as compared to the hard ball.

32

The implications for therapy are that children with perceptual disturbances would be
seated in such a way that they know where they are in space. This can be achieved by having
them sit on a solid surface, feet placed on the ﬂoor and hands on a table offering a solid
surface as well. Wherever and whenever possible, the objects used for intervention should be
solid, too. Instead of using a plastic spoon, where the impaired child cannot feel much when
he or she touches it, a metal spoon would be preferable. These accommodations would make
it easier for the child to orient himself or herself in space and to the functional task he or she
is about to tackle.

Positive research ﬁndings would also have implications for children with speciﬁc
language impairment (SL1). Completion of problem-solving functional tasks would help them
develop the sequencing skills very much needed in daily life and in a school setting.
Metalinguistic performance (retelling stories, summarizing events, and so on) will be
impaired unless the child with SLI acquires sequencing skills.

As the studies on CNVs in autistic and normal children are scarce or nonexistent, the
results of this study will increase understanding of CNV measurement and the kind of
information we might get from it.

The present study is pioneering work. To date, no studies have been conducted testing
fairly complex segments of human behavior or, in Affolter and coworkers’ terminology,

“topological changes.” The results will shed light on the feasibility of this kind of endeavor.

Research Questions
The following research questions were posed to guide the collection of data for this
study:

1. Is it possible to elicit CNVs in autistic children?

33

2. Is it possible to elicit a CNV when the autistic subject is manually guided to do an
event?

3. Is there a difference between the CNV elicited when the autistic subject is
manually guided and unguided to do an event?

4. Is there a difference between the CNV elicited from autistic subjects when the

event stimuli vary in the amount of resistance?

34

CHAPTER TWO

EXPERIMENTAL METHODS AND PROCEDURES

In this study, brain electrical potential shifts occurring between paired auditory stimuli
were measured. The formation of the contingent negative variation (CNV) was of particular
interest as it is indicative of attention/cognitive functioning. Two independent variables were
manipulated: (a) guided versus unguided condition, and (b) changes in the resistance of
stimulus material. The four dependent measures of brain wave activity included:

1. The duration in ms of the CNV or the largest negative potential shift (NPS).
2. The average amplitude in volts of the CNV or the largest NPS.
3. The extent of the area as a function of both duration and average amplitude of the

CNV or the largest NPS.

4. The overall extent of the area as a function of both duration and average amplitude of

the negativities recorded during the entire sampling time.

Subjects
Three autistic subjects were selected. They included two boys and one girl, all ranging
in age from 6 to 10 years. They had been diagnosed with autism in the range from moderately

severe to severe in accordance with the criteria stated in DSM-IV.

35

Subject-Inclusion Critep'a

The autistic subjects were diagnosed with autism in accordance with the criteria stated
in DSM-IV (fourth edition) of the American Psychiatric Association by a qualiﬁed
professional (psychiatrist, psychologist, or physician) (see Appendix A). In addition to
meeting DSM-IV criteria, subjects also met the following criteria for inclusion in the study:

1. They had no known cause of brain damage, such as infantile stroke.

2. They had no further medical conditions such as diabetes, or seizures.

3. They had not been born prematurely, and therefore the clinical proﬁle was not

assumed to be the result of slow maturation.

A

. They had no additional syndrome such as Fragile X.

LII

. They had normal hearing and vision, as confnmed by teacher report.

6. They were native speakers of American English.

7. They had no physical impairments that might have decreased their agility and

independent mobility.

8. They were unmedicated: One subject used Ritalin irregularly but was free of

medication on the days of testing.

The investigator knew the autistic subjects from working with them regularly in the
autistically impaired resource room in the Lansing School District. Permission was granted to
visit the autistically impaired resource room to interact with the children. The investigator
relied on psychological and teacher reports to ascertain that the subjects were autistic. The
investigator relied also on her own observations made in the autistically impaired classroom

to determine handedness of the subjects while working with them. All autistic subjects were

right-handed.

36

Procedures for Subject RecruitmenL and Selection
Subjects were recruited from the Lansing Public Schools, using guidelines speciﬁed
by the MSU Committee on Research Involving Human Subjects (U CRIHS). They were
recruited from the special education autistically impaired program. Only three of the ﬁve
students in the Autistically-Impaired Resource room met the inclusion criteria. Parental

consent was obtained for each subject who was tested (see consent form in Appendix B).

Human Subjects Considerations
UCRIHS reviewed and approved this project and also the consent form to be used
(see Appendix B). Each subject was paid an honorarium of $25 after each testing session at

the Neuro-Audiologic Laboratory, Department of Audiology and Speech Sciences, MSU.

Data Collection
Description of Tasks

The experimental tasks were developed and pilot tested. More than 30 hours of actual
observation time were invested in getting the most reliable results possible within the
constraints of the laboratory equipment available to the investigator. Approximately 10
graduate and undergraduate students served as pilot subjects on a voluntary basis to assist
with developing the experimental protocols, including the tasks and instrumental settings.

The subjects were presented with six experimental tasks. In addition to the tasks, the

subjects’ auditory brain-stem responses (ABR) were recorded.

37

Task 1 -- Push Button Tag

The subject sat at a table with the left hand resting in the lap and the right hand resting
on the table next to the response box. The subject heard white noise and a click delivered
over headphones. The noise was stimulus one (81) and the click was stimulus two (S2). The
2 stimuli were presented at a preset interval of 2 seconds. When the subject heard the noise
(S l), a green light was displayed on the response box, and when the click (32) was presented,
a red light was displayed on the response box. The subject was prepared to push the button
when the noise was heard and the green light was observed. After the click was presented
and the red light was displayed, the subject immediately pushed a button on the response
box. Six cycles of S1 and 82 presentations were delivered and averaged over time.

The Push Button task had been used in the Neuro-Audiologic Laboratory at MSU in
previous experiments and was known to produce a CNV in unimpaired adult subjects.
However, the light stimuli were added to the Push Button task to make it more interesting for
the autistic subjects. The Push Button task was selected to answer the question of whether
autistically impaired children could establish a CNV. Consequently, the feasibility of the

CNV paradigm for this population was investigated.

Task 2 (Squeezing; Soft Ball -- Unguided) and Task 3 (Squeezing a Hard Ball -- Unguided)
For each task, the subject sat at a table with the left hand resting in the lap, and the
right hand resting on the table. A ball (23 cm circumference) was placed next to the subject's
right hand. White noise and click were presented at 2-second intervals as described
previously for Task 1. When the noise (S 1) was heard, the subject prepared to perform the
task. After the click (S2) was perceived, the subject squeezed a ball. Six cycles of SI and 82

were presented and averaged over time.

38

In Task 2, the subject squeezed a soft ball. It was made of foam and had a
circumference of 23 cm. In Task 3, the subject squeezed a hard ball. It was made of
Styrofoam and had a circumference of 23 cm. These tasks were designed to investigate

whether negative potential shifts were altered by changes in resistance.

Task 4 -- Squeezing a Soft Ballﬁuided

The subject sat at a table with the left hand resting in the lap and the right hand resting
on the table. The same soft ball (made of foam, circumference of 23 cm) that was used in
Task 2 was placed next to the subject's hand. The investigator sat next to the subject and
listened for the same stimuli presented to the subject over a second set of headphones. When
the investigator and subject perceived the noise (SI), they waited for the click (82). As soon
as the click was delivered, the investigator took the subject's hand and placed it around the
ball. Then the ball was squeezed. When the investigator guided (hand-over-hand) the subject,
the investigator’s ﬁngers were on top of the subject’s ﬁngers, and the investigator’s palm
rested on the back of the subject’s hand so that both hands could move together
simultaneously. Six cycles of SI and 82 were presented and averaged over time.

This task was designed to determine possible differences in brain activity as related to
the CNV or large NPS between Task 2 -- where the subject squeezed the soft ball
independently -- and Task 4 -- where the subject was led manually (hand-over-hand) to

perform the same task.

Experimental Tﬁk 5 - Picking up Light Glass. Guided
The subject sat at a table with the left hand resting in the lap and the right hand resting

on the table. A glass (made of plastic, height = 14 cm, diameter of the glass at the bottom =

39

5.5 cm, diameter of the glass at the top = 8.5 cm, weight = 75 g) was placed next to the
subject's hand. The investigator sat next to the subject. As with other tasks, signals for the
subject and investigator were delivered through headphones. The white noise and the click
were delivered at preset 2-second intervals. When noise (31) was presented, the investigator
and subject waited for the click (S2). As soon as the click was presented, the investigator
guided the subject to grasp the glass, and to pick it up and put it down manually. Six cycles of
SI and 82 were presented and averaged over time.

Unlike Task 4, in which subjects were guided to perform a task they had done
previously, Task 5 involved guiding the subjects in an unfamiliar event. Task 5 was designed
to control for the order effect as compared to Task 4 (Squeezing a Soft Ball — Guided). In

addition, this task was used to investigate the guided condition in a different task.

Task 6 - Picking Up a Heavy Glass, Unguided

The subject sat at a table with the left hand resting in the lap, and the right hand
resting on the table. A glass, identical to the one used in Task 5, was used. The only
difference was that the glass in Task 6 was ﬁlled with sand. The weight difference made the
task more interesting for the subjects because it added novelty, and it also added resistance.
The glass (made of plastic, height = 14 cm, diameter at the bottom of the glass = 5.5 cm,
diameter at the top of the glass = 8.5 cm, weight = 575 g) was placed next to the subject's
hand. The white noise (S 1) and a click (S2) were delivered in sets of 2-second intervals. S1
was the warning signal. As soon as 82 was heard, the subject grasped and picked up the
glass, and put it back down. Six cycles of SI and 82 were presented and averaged over time.

Tasks 5 and 6 were designed to compare task performances in the guided and

unguided conditions and to reveal possible differences in CNV or NPS. In addition, Tasks 5

40

and 6 were designed to be identical except for the aspect of weight. The glass used for Task 6
was 500 g heavier than the one used for Task 5. This aspect was changed to add novelty to

Task 6 so that the subjects would not lose interest.

Auditory Brainstem Response (ABR)

An ABR recording was obtained for each subject. As the subjects did not undergo
formal hearing evaluations, an ABR recording was obtained from each subject to determine
that the auditory system was functioning normally. This non-volitional procedure required the
subject to lie back in the chair, close his or her eyes, and relax. The subject was presented
with 1,024 clicks 200 micros in duration at a repetition rate of 11.1/s.

An ABR recording provides information about the functioning of the auditory system,
and the auditory nerve in particular. The classic ABR response consists of ﬁve positive
principal peaks. Wave I is generated in the eighth nerve, and Waves III to V are generated in
the brainstem. All other aspects of the waves (their sources and what they reﬂect) are still
under discussion. The diagnostic power of brain stem auditory evoked potentials relies on the
interpeak latency of Waves I to V, which represents brainstem transmission time and

therefore brainstem auditory processing (Celesia & Brigell, 1993).

Administration of Tasks
Data collection took place in the Neuro-Audiologic Laboratory in the Department of
Audiology and Speech Sciences at MSU. The investigator picked up the autistic subjects
from school and drove them to the Neuro-Audiologic Laboratory during school hours. They

were accompanied by their school aide.

41

Each subject was subjected to a standard protocol of tasks in the S 1-82 CNV
response paradigm as described above. The investigator administered the tasks. Additional
personnel provided support, as needed, to operate the computer or to monitor subjects’
responses during tasks.

Each subject came to the laboratory twice. The same tasks were administered in the
same sequence a second time to ensure reliability of results. The second testing took place

within a month of the ﬁrst testing.

Preparation of Subjects for Tasks

The subjects were prepared for the experiments in one room of the laboratory. In this
room three electrodes were applied to the subject's scalp in accordance with the International
10-20 system. One electrode was placed in the F 2 position. A second electrode was clipped
on the subject's left ear to serve as a ground, and a third electrode was placed on the subject's
right ear to serve as a reference. The impedance of all electrodes was kept below 5k ohms
throughout the recordings, in accordance with the standard procedures used in the Neuro-
Audiologic Laboratory at MSU, and as used in research reports on eliciting EEG responses
(Dawson, Klinger, Panagiotides, Lewy, & Castelloe, 1995; Pfurtscheller, Stancak, & Neuper,
1996)

Once the electrodes were placed, the subject was led into a sound-proof room of the
laboratory. It contained a table, a chair, and instruments for receiving and monitoring auditory
signals, namely a preampliﬁer box, headphones, and so on. There was another chair 1.5 m

away from the subject, where an aide could sit and observe the child.

42

Before each experimental task, the investigator gave nonverbal instructions. Intention
implies that this was not always done. The investigator often took the subj ect's hand and

guided him or her to perform the task after the imperative stimulus (82) had occurred.

Order of Task Administration

The subjects were tested individually. They performed the tasks in the same order,
which was as follows:

1. Push Button Task

2. Squeezing a Soft Ball -- Unguided

3. Squeezing a Hard Ball -- Unguided

4. Squeezing a Soft Ball -- Guided

5. Picking Up a Light Glass -- Unguided

6. Picking Up a Heavy Glass -- Guided

7. ABR

Recording of Data

I_n_§trumental Recording of Subjects’ Responses to Taslg

A signal generator (MI 2xx) was triggered by a custom-written program to produce 81
and 82. The subject's electrodes were connected to the preampliﬁer box that allowed for the
stimuli to be ampliﬁed (MI 2xx) and attenuated (MI 2xx). CNV data were ﬁltered with a
high-frequency cut-off of 10 Hz and a low-frequency cutoff of 1 Hz. The preampliﬁer (Grass
sss) gain was set at 30K x 6. The signal was fed to computer memory, which was connected
to an oscilloscope and a computer monitor so that the investigator could observe the wave

conﬁguration on line. Electrophysiological activity was averaged on-line with signal

43

averaging software (MI 2xx) on an IBM-XT computer. One channel was used. Sampling time
was 2,499.71 ms, and sampling frequency was 3.02 KHz. Responses were generated across
six trials for each experimental task.

The number of cycles for each subject in every experimental task was determined on
the basis of extensive testing. The pilot subjects were asked to perform the experimental tasks
in cycles varying from 4 to 30. It was observed that the number of cycles coincided with
ﬁndings stated by Walter (1965), whose pilot testing showed also that the CNV was observed
best after only a few cycles of presentations had been given.

As far as the interference of eye movements with the CNV was concerned, Low et al.
(1966) found that horizontal eye movements were not signiﬁcant but that upward eye
movements were associated with a positive potential shift. As an evaluation of positive
potential shifts was not of interest in the present study, the effect of eye movements could be
ignored. The CNV occurs only in the form of NPS.

The ABRs were recorded with the same equipment, and the computer software
program "dacpulse" (MI 2xx). The number of trials of each ABR was 1,024. Sampling
frequency was 100.0 KHz, and sampling time was 10.0 ms.

The CNV experimental parameters were as follows:

Interstirnulus interval (181) = 2000 ms
Alternating polarity = A

Click and noise stimuli used = C and N
Click duration = 2 ms
Noise-masking duration = 50 ms
Rise/fall duration = 10 ms
Delta T = 500 ms

44

Behavioral Observations

In addition to measures of brain activity, the investigator also made behavioral
observations. She noted whether the subject seemed attentive or inattentive to the tasks.
She wrote down, for instance, whether the subject switched hands during an experimental

event.

Data Analysis

A CNV is a slow, protracted potential (Walter, 1965). Gri'rnewald (1979) applied the
term CNV to cerebral NPS occurring in the interval between a warning stimulus (SI) and an
imperative or otherwise signiﬁcant stimulus ($2). The CNV takes into account the
measurements of average amplitude and duration of the NPS.

In this study, an NPS was deﬁned as a CNV whose duration extended over a period of
400 ms and whose average amplitude reached a value of —0.500 volts/division. Values less
than 400 ms and less than -0.500 volts/division were regarded simply as NPS.

Previous researchers have reported CNV criteria for durations ranging from 200 ms to
1,000 ms. Rebert and Knott (1970) concluded from their investigations that the average time
for onset of the CNV was 467 ms. As far as the duration of the CNV is concerned, Walter
(1965) observed that the response to the ﬁrst stimulus always contained a small negative
variation, with a duration of about 200 ms equal in amplitude to the negative component of
the second response. Walter (1964) also showed data where the CNV between two stimuli
(click and ﬂashes) was recorded over a time interval of 1 second. In general, the duration of
the CNV varies and the length of the interstimulus interval varies too, depending on the test

equipment available to the investigator.

45

Walter (1964) stated that the CNV usually has an amplitude of less than 20
microvolts, but Low et al. (1966) reported it to be as high as 50 microvolts in subjects
younger than 12 years of age. As described earlier, the criteria for deﬁning a CNV took into
account the 2-second interstimulus interval, placement of the electrode in the F2 position, and
the ﬁndings of Walter (1964) and Low et al. (1966).

For the present study, all NPS with a duration of 400 ms and an average amplitude of
-0.500 volts were characterized as a CNV. Values below 400 ms and -0.500 volts were

characterized as NPS.

Dependent and Independent Variables

The four dependent variables that were used to interpret the inﬂuence of the
independent variables were:

1. The measured duration of the CNV or the largest NPS as identiﬁed by the
investigator’s visual inspection (Klimesch et al., 1996); the negative depression from the
onset to the offset of the NPS was measured in ms along the x-axes.

2. The measured average amplitude of the CNV or the largest NPS as calculated by
ﬁnding the sum of all amplitudes in the region of interest (NPS) and dividing the obtained
value by the number of data points for that particular region (NPS).

Average amplitude = 2 (v/V), where v = amplitude in volts, and V = all the data points.

3. The measured area of NPS between the two stimuli (S1 and 82); the area was
approximated as a sum of consecutive rectangles, with one side being the time interval and
the other side the amplitude. The formula was as follows:

Area = E (t - v), where t = the time interval of 0.331 ms, and where v = the

amplitude for each individual data point.

46

4. The measured overall area of negativities of the entire sampling time, as obtained
by going through the raw data and adding the individual negative areas below the baseline.
Overall area = the product of the time interval (0.331 ms) and the sum of the negative areas.

The four dependent variables were analyzed to test each of the experimental
hypotheses. Waveforms for each task and each subject were plotted from the raw data.
Afterwards, the dependent variables were identiﬁed by using the mathematical operations
described above. For each analysis, data were pooled across the subjects and trials to obtain
the best possible estimate of the parameters under study.

The two independent variables used in the study were (a) guided versus unguided
condition and (b) changes in resistance. The independent variables pertained to Research
Questions 3 and 4. The obtained values for the guided and unguided conditions were
compared in Tasks 2 versus Task 4, and Task 5 versus Task 6. Changes in resistance were

obtained by comparing the values recorded for Task 2 versus Task 3.

Possible Differences Between Guided and Unguided Conditions

To determine whether there was a difference between the CNV elicited when the
autistic subject was manually guided and unguided to do an event, Task 2 (Squeezing a Soft
Ball — Unguided) was compared to Task 4 (Squeezing a Soft Ball — Guided). Also, Task 5
(Picking Up a Light Glass — Guided) was compared to Task 6 (Picking Up a Heavy Glass -

Unguided).

Possible Differences Between Stimuli Varying in the Amount of Resistance
To determine whether there was a difference between the CNV elicited from autistic

subjects when the event stimuli varied in the amount of resistance, Task 2 (Squeezing a Soft

47

Ball — Unguided) was compared to Task 3 (Squeezing a Hard Ball — Unguided).

Consequently, the researcher determined possible changes in resistance.

Research Questions
Elicitation of CNV in Autistic Subjects
To answer Research Question 1, regarding whether autistic subjects were able to
establish a CNV, the EEG recordings of Task 1 (Push Button) were analyzed. Then the

values obtained for duration and average amplitude were compared to the criteria deﬁned for

a CNV.

Elicitation of CNV in the Guided Condition

To answer Research Question 2, concerning whether autistic subjects were able to
establish a CNV when they were manually led to do an event, the EEG recordings of Tasks 4
(Squeezing a Soft Ball — Guided) and Task 5 (Picking Up a Light Glass — Guided) were
analyzed and compared to the criteria described above. Consequently, the researcher

determined the absence or presence of a CNV.

Possible Differences Between the Guided and Unguided Conditions

To answer Research Question 3, regarding whether there was a difference between the
CNV elicited when the autistic subject was manually guided and unguided to do an event,
Task 2 (Squeezing a Soft Ball -— Unguided) was compared to Task 4 (Squeezing a Soft Ball —
Guided). Also, Task 5 (Picking Up a Light Glass — Guided) was compared to Task 6

(Picking Up a Heavy Glass -—Unguided).

48

Possible Differences When the Event Stimuli Vary in the Amount of Resistance

Research Question 4 concerned whether there was a difference between the CNV
elicited from autistic subjects when the event stimuli varied in the amount of resistance. To
answer this question, Task 2 (Squeezing a Soft Ball - Unguided) was compared to Task 3

(Squeezing a Hard Ball - Unguided).

Statistical Analyses
After recording the data, the researcher converted the data points from the DOS
program to Excel to calculate the measures for the dependent variables. The Wilcoxon
matched pairs signed-ranks test was used to test for significant differences between responses
and conditions. The data were pooled across two trials for each subject. This amounted to a

maximum of six data points to analyze each experimental variable.

49

CHAPTER THREE

RESULTS

In this study, EEG recordings were analyzed with regard to NPS and the possible
formation of CNV. A CNV is an NPS extending over a certain period of time and reaching a
certain amplitude. According to the literature (Low et al., 1966; Prevec & Beric, 1990: Ikeda,
1997; Walter, 1964), the exact values for determining a CNV vary depending on the length
of the interstimulus interval, the kind of task to be performed, and the location of the
electrodes.

The criteria for deﬁning a CNV in this study took into account the 2-second
interstimulus interval, placement of the electrode in the F 2 position, and the ﬁndings of
Walter (1964), and Low et al. (1966). An NPS was deﬁned as a CNV whose duration
extended over a period of 400 ms, and whose average amplitude reached a value of -0.500
volts/division. Values less than 400 ms, and less than -O.500 volts/division were regarded
simply as NPS.

A CNV in a 81-82 response paradigm contains two basic components (Gritnewald,
1979; Ikeda, 1997) -- an early component of the CNV that is called Bereitschaftspotential
(BP), and a late component that occurs before 82. In the present study, the components of the
CNV were not separated because there is evidence in the literature that both are indicative of
attention (Walter, 1965), although the two components might be generated in different areas

of the brain (Ikeda, 1997).

50

In the present study, three autistic subjects (two boys, one girl) varying in age from 6
to 9 years were tested twice on six experimental tasks as described earlier. In the data
analysis, the autistic boys were referred to as Subjects 1 and 2, and the autistic girl was
referred to as Subject 3.

In the ABR (Auditory Brainstem Responses) recordings the 5 expected peaks were
visible. Consequently, it was assumed that there was no gross abnormality of the auditory
system that would have excluded the subjects from participating in the experiments.

The results of the study on experimental tasks are displayed and described below for
each of the questions posed for investigation. Data analyses for each question involved an
analysis of each of the following dependent variables:

1. average amplitude of CNV or largest NPS

2. duration of CNV or largest NPS

3. area of negativity of CNV or largest NPS

4. area of negativities for the entire sampling time

The Wilcoxon matched pairs signed-ranks test was used to test for signiﬁcant
differences between responses and conditions. The results for the three subjects were pooled
across ﬁve or six trials for each of the four dependent variables.

To answer each research question, the averaged wave forms generated during each
. task were inspected to determine whether the amplitude and duration criteria for a CNV were
met for each subject in each trial, as described above. In addition, the values for the area of
the CNV or the largest NPS and the overall area of negativities for the entire sampling time

were calculated.

51

Research Question 1
Is it possible to elicit CNVs in autistic children?

The ﬁrst question addressed a methodological issue -- that is, whether an autistic
subject can establish a CNV. A Push-Button task similar to Task 1 had been used in the
Neuro-Audiologic Laboratory at MSU in the past and had produced CNVs in unimpaired
adult subjects. Hence, the investigator assumed that this kind of task was suitable to address
the question. Only Task 1 was selected to investigate whether the autistic subjects could
establish a CNV.

On Experimental Task 1 (Push Button), Subjects 1, 2, and 3 were tested twice. The
results for the three subjects pooled across six trials for each of the four dependent variables
are displayed in Table 1. Figure 1 displays a CNV for subject 1, and Figures 2 and 3 display

large NPS on Task 1, for Subjects 2 and 3, respectively

Duration and Average Amplitude of the CNV or Largest NPS (Push Button)

As is shown in Figures 1, 2, and 3, the duration of an NPS serves as a measure of the
length of time that attention can be sustained, and the average amplitude of the NPS serves as
a measure of the magnitude of the brain response as judged by the steepness of the negative
wave. The group's mean value for the duration of the CNV/largest NPS was 321.100 ms (SD
= 124.3), and the average amplitude was -0.891 volts (ﬂ = -0.421). The group's mean value
for average amplitude exceeded the CNV criterion of —0.500 volts by -0.391 volts. But the
group’s mean value for duration did not exceed the preset CNV criterion of 400 ms. On
average, the group did not meet both criteria deﬁned for duration and average amplitude.

The group’s trend did not reﬂect individual subject data. The criterion for duration

was met only by Subject 1, Trial 1. However, the two large NPS observed for Subject 1, Trial

52

.68 was» 33:95 own—98 com corpora >20 “m5 @525 a
.88 33 commune com donate >20 35 macaw a
.328 near; oéamoz aowéaoo .

 

 

 

SEN oooSN N13- 83$ 8
83? oocNNm awe- SZNN 50:
NNNSe $32 DEE- 9.38 N N
3:.NN4 838 1:2- :32 N N
”N28 232 Sec. :22 a N
N863 $33 .82- $32 m _
3.4.2: man: age..- 832 N a

8983 2.4.8» 38;- Nazi _ .L
$84333 @8333 as? 8? >28 a8 8% >28

32 €226 32 one: avenue... .9... as 8225 38.3% are

 

«taco 35.1555 at: $2 523
newness a 5;» E: can a 30.3% as 5.1 $2 Ewan

.9025ch I cozsm 53% ﬂ xmah EcoEtonm :o mmZ “mowbﬁ mo 3:582 “dungeon A 033.

53

._ E; ._ 5033 22:2 .BENS 523m :95 ._ s15 Eastman .1. 23m .32 88 ._ 25mm

002:

salon

 

54

,N RE. .N Suﬁsm 23.5.4 602395 £025 is; 4 V7,: 3:05:35 .m Eo>m m2 Eomo

A25 oE:

 

.N 239.»

‘51“)A

55

.N RE. .m Suﬁsm 26:3,. .uoEsmcD 695m £5 4 v7.8. EcoEtonm .m Eo>m o2 $220

:25 «E:

 

m 2ng

sum

56

2, and Subject 2, Trial 2, came close to the criterion. Thus, only Subject 1, Trial 1, displayed
a CNV (see Figure 1) with a duration of 540.192 ms and an average amplitude of -1 .639
volts. Its duration and average amplitude were more than 1.5 ﬂ above the mean. However,
Subject 2, Trial 2, established a large NPS (see Figure 2) that displayed a duration of
358.877 ms, falling short of the deﬁned criterion by 41.123 ms, and an average amplitude
of -0.711 volts exceeding the criterion. Altogether, Subject 2, Trial 2, met 89.7% of the
criterion for duration and 142% of the criterion for average amplitude. Subject 3's largest
NPS was very similar in shape to a CNV, which is deﬁned as a depression, displaced in the
negative area of the curve, and terminated by the points where the onset trough-like and
offset of the depression cross the baseline. However, Subject 3’s largest NPS lacked the
necessary dimensions. It gave the impression of a “depressed CNV” as compared to the CNV

displayed by Subject 1 (see Figure 1).

Area of the CNV/Largest NPS and Overall Negativities of Entire SampligLTime
(Push Button Task)

The area of the CNV/largest NPS is a combined measure of duration and average
amplitude, and so is the measure of overall area of negativities extending across the entire
sampling time of a task. Area and overall area of negativities represent measures of cognitive
ﬁmctioning as described earlier.

The restricted area of the CNV was compared to the entire area of negativities. The
group's mean for the area of the CNV/NPS was 322 (S_D_ = 281), and the group's mean for
overall negativities of the entire sampling time was 735.0 (SD = 257.0). The Wilcoxon
matched pairs signed-ranks test pooled across subjects and trials (N = 6) revealed a

signiﬁcant difference between the area of the CNV/largest NPS and the overall area of

57

negativities of the entire sampling time [I Q = 6) = 0, p < .05]. To be statistically
signiﬁcant, the T-value had to be less than or equal to l for the present sample size. This
comparison showed that there was a great amount of negativity outside the area of the CNV/
largest NPS while the subjects were performing the tasks.

Subject 1, Trial 1, not only established a CNV representing the largest area of all
measured trials in Task 1, but he also established the largest area of negativities for the
overall area. The largest negative area measured 886.413 volts*ms, which was more than 2
S_D above the mean. The area of overall negativities for the entire sampling time was
1.193.64 volts*ms (1.5 SD above the group' 5 mean).

The value for the area of the largest NPS (255.263 volts*ms) was within 1 S_D of the
group’s mean for Subject 2, Trial 2. Subject 2 showed an overall area of negativities
(422.184 volts*ms) that was more than 1 SD below the group's mean. Subject 3's values for

area of the largest NPS and overall area of negativities were within 1 S_D of the mean.

Research Question 2: Eliciting the CNV in the Guided Condition
Is it possible to elicit a CNV when the autistic subject is manually guided
to do an event?

Table 2 contains the results for Experimental Tasks 4 (Squeezing a Soft Ball ) and 5
(Picking Up a Light Glass) in the guided condition. Subject 2, Trial 2, established two
CNVs in the guided condition. One CNV was produced in Experimental Task 4 (Squeezing
a Soft Ball -- Guided), and a second CNV was produced in Experimental Task 5. As

mentioned earlier, the basic difference between these tasks was a potential novelty effect.

58

.38 33 0832888 owﬁoﬁ 8.“ morocco >20 35 83% .

.uoa mm?» Goya-"St How GOCOuT—o >20 Hag mBOJm a

.928 a335> 3882 80888 .

 

 

 

59

8.8.le 8N. H m a l .o- 83m $38 886 am

832 831. 8 _ a- 888 5.8 886 am

8. _ No 8N.:N NNwd- 8°.on 9.28 885 532

82mm 82”: mm;- 2: .NmN 9.8 Base as:

83.8 8N2 N .488- 2 a. 8 ON 9.28 cacao N N

8:88 882 .83- N88. :38 Beao N N

£38 838 .32.- .58.. $38 885 N .N

N588 88.8 .83- .E...:.. 288 885 N ...N

Ben: 3 SN .83- 838 9.28 8830 _ N

83: NEON .88...- 82 .8N 3.8 886 _ N

$8.88 $38 .28- mNNNNN 938 Base m a

3%? 838 .88- N88N 238 825 N a

82: m 08.2: .48...- 82: @228 Beao N _

o: .2 m 832 .83.- N28N 9.8 886 N _
- - - - cacao _ _

Ame-£98 98.80: 82, 83A 26 as 8% >28
80.24 =80>O 38¢. 4333 33:92 .w>< $88 8985 803880 88 Sam Btu.

 

885 Bacon-Eu 382 98 523
88085 a. a; 35 80am 5 .838 88 8 $2 Ewa-H

 

.Goeao 1 ago am: a 8:
wqﬁommv m xmm-H ﬁﬁaonaomxm msmuo> 30280 I 35 ﬂow a mENooscmv v xmmh 8828-8qu :0 mmZ “mow-8..— mo 3.8302 806530 .N 033-

Task 4 (Sgueezinga Soft Ball — Guided)
On Task 4, subjects were tested twice. The results for the three subjects pooled across
5 trials for each of the dependent variables are displayed in Table 2. Figure 4 displays a

CNV for Task 4.

Duration and Average Amplitude of the CNV/Largest NPS for Task 4 (Squeezing;
Soft Ball)

The group's mean for the duration of the CNV/ largest NPS in Experimental Task 4
(Squeezing a Soft Ball) was 252.100 ms (S_D = 93.9), and the mean for average amplitude
was -0.755 volts (S1; = 0.180). The mean value for average amplitude exceeded the
minimum -0.500 voltage requirement by 0.255 volts. All subjects met the minimum
requirement for average amplitude in every trial for this particular task, but not the duration
criterion. Consequently, when individual subject data were examined, just one subject met
both CNV criteria. The CNV established by Subject 2, Trial 2, in Task 4 (Squeezing a Soft
Ball) had a duration of 4 1 7.476 ms and an average amplitude of -0.548 volts (see Figure 4). It

was more than 1.5 & above the mean for duration and within 1 SD of the mean for average

amplitude.

Area of the CNV/Largest NPS and Overall Negativities of the Entire SamplingTime for

Task 4 (Squeezing a Soft Bam

The group's mean for the area of the CNV /largest NPS was 182.700 volts*ms (S_D =
44.6). The group's mean for the overall area of negativities was 558.300 volts*ms (SQ =
128.0). In Trial 2, Subject 2’s value for the area of his CNV was 1 $2 above the mean, and

his value for the overall area of negativities was within 1 S2 of the mean.

60

.N BE. .N Suﬁsm o:£§< 6028 5am «om N wcﬁoozcm .v was. _8:oEtoqu .o Eo>m 69 Gomo

850a 3333:.

 

.v oSwE

sum

61

The group’s mean for the overall area of negativities during the entire sampling time
was 558.300 volts*ms (S_D = 128.0). A Wilcoxon matched pairs signed-ranks test was used
to compare the negativity of the area of the largest shift to the area of overall negativities of
the entire sampling time. The result showed a signiﬁcant difference [I (E = 5) = O, p < .05]
between the two dependent variables. The value for I had to be below or equal to 1 to be
statistically signiﬁcant. This comparison showed that there was signiﬁcantly more negative

activity during the entire sampling time than during the recording of the CNV/largest NPS.

Task 5 (Picking Up a Light Glass — Guided)
On Task 5, Subjects 1, 2, and 3 were tested twice. The results for the three subjects,
pooled across ﬁve trials for each of the four dependent variables, are displayed in Table 2.
Figure 5 displays a CNV for this task. Figures 6 and 7 display large NPS meeting between

80% and 90% of criterion for duration and exceeding criterion for average amplitude.

Duration and Average Ampjitude of the CNV/Largest NPS for Task 5 (Pickinwp a Light
G_12§s_)

Table 2 also shows that the group's mean for duration of the CNV/largest NPS for this
guided task condition was 320.000 ms (S_D = 86.9). On the whole, the group met 80% of the
criterion for duration of the CNV. The group's mean average amplitude was -0.823 volts (S_D
= 0.154). The amplitude value exceeded the minimal required voltage of -0.500 volts for a
CNV by -0.323 volts. However, the group’s mean values did not meet the CNV criteria for
both measures (duration and average amplitude).

Individual subject data revealed the following: The same subject who established a

CNV in the guided condition on Task 4 (Squeezing a Soﬁ Ball) produced a CNV in the

62

.N E; .N 8.38 223. .825 .320 am: a a: 8%: .m v.5 .205:me .: 59m .02 some

.25 2::

 

.m 959m

$qu

63

.N EC. ._ 80.35 23.ng .3230 .820 Emﬁ a a3 @2on .m v7.3. EcoEtoaxm .2 Eo>m d2 Gomo

we. 2::

 

.0 233»

5110A

64

guided condition on Task 5 (Picking Up a Light Glass); see Subject 2, Trial 2. His CNV had
a duration of 430.631 ms and an average amplitude of -0.744 volts (see Figure 5). Subject
2's CNV was more than 1 S2 above the mean for duration, and it was within 1 S2 of the
mean for average amplitude.

Although subjects 2 established only one CNV in the guided condition, in trial 2,
Subjects 1 and 3 established large NPS, meeting between 80% and 90% of the criterion for
duration while exceeding the criteron for average amplitude. Subject 1, Trial 2, showed a
large NPS (meeting 91.8% of the criterion) on Experimental Task 5 (Picking Up a Light
Glass), with a duration of 367.079 ms and an average amplitude of -0.588 volts (see Figure
6). Subject 3, Trial 2, established a large NPS, meeting 80% of the criterion for duration
(see Figure 7) and achieving the highest value for average amplitude of all subjects in the

guided conditions.

Area of the CNV/Largest NPS and Overall Negativities of the Entire Sampling Time for
Task 5 (Picking Up a Light Glass)

The group's mean for the area of the CNV/largest NPS was 241.800 volts*ms (S__ =
51.200). The group's mean value for overall negativities recorded throughout the entire
sampling time was 621.800 volts*ms (S2 = 125.00). The Wilcoxon matched pairs signed-
ranks test was used to compare the area of the CNV/largest NPS with the overall area of
negativities. The test revealed a signiﬁcant difference [1‘ (ﬂ = 5) = 0, p < .05] between the
two dependent variables. To be signiﬁcantly different, I had to be below or less than 1. This
ﬁnding demonstrated signiﬁcantly more negative activity outside the area of the CNV or

largest NPS than within the area deﬁned as CNV or largest NPS.

65

_N .2; ._ 8.38 233 .8258: .320 53$ a a: 8%: .0 gap @585me .2 :55 .02 $80

as; 2::

 

.5 23mm

53104

66

Comparison of Tasks 4 (Squeez'mg a Soft Ball — Guided) and Task 5 (Picking Up a
Light Glass — Guided)

As pointed out earlier, one main difference between the two guided tasks was
novelty. Task 5 (Picking Up a Light Glass) was a novel guided task. The subjects did not
know what the task would be about when the investigator guided them, whereas Task 4
(Squeezing a Soft Ball) was familiar to the subjects as they had performed the identical task
independently in Task 2. As shown in Table 2, the group’s means for all four dependent
variables (duration, average amplitude and area of the CNV or largest NPS, and overall area
of negativities) were higher for the novel task (Task 5, Picking Up a Light Glass - Guided) as
compared to the familiar task (Task 4, Squeezing a Soﬁ Ball — Guided). Although individual
data did not show this trend in every trial and for each of the dependent variables, the overall
tendency was signiﬁcant. To test whether the novelty effect was signiﬁcant, 3 Wilcoxon

matched pairs signed-ranks test was performed.

Effect of Novelty on Duration and Average Amplitude in Task 4 (Squeezing Soft Ball —
Guided) and Task 5 LPickingUp a Light Glass)

When the duration of the CNV/largest NPS in Task Four was compared to the same
dependent variable in Task 5, a signiﬁcant difference was found [I (IS = 5) = 0, p_ < .05]. The
T-value had to be below or equal to 1 to be statistically signiﬁcant. This means that the
duration of the CNV / largest NPS was signiﬁcantly longer in the novel task than in the
familiar task. For the parameter of average amplitude, no signiﬁcant difference was found
(T = 7). The magnitudes of brain activity in this dimension were not signiﬁcantly different

between the two.

67

Effect of Novelty on Area grid Overall Area of Negativities of Task 4 (Squeezing a Soft Ball
— Guided) and Task 5 (Picking Up a SigmGlass — Guided)

When the area of the CNV/largest NPS was compared in the two tasks, a signiﬁcant
difference [I (111 = 5) = 0, p < .05] was found. The area of the CNV/largest NPS was
signiﬁcantly larger in the novel task than in the familiar task. However, a comparison
between the areas of overall negativities during the entire sampling time did not reveal

signiﬁcant differences (T = 3).

Research Question 3: Guided Versus Unguided Condition
Is there a difference between the CNV elicited when the autistic subject
is manually guided and unguided to do an event?

In posing this question, the goal was to determine whether there were differences in
brain activity between the two conditions. Two sets of tasks were designed to address this
question. Tasks 2 (Squeezing a Soft Ball — Unguided) and Task 4 (Squeezing a Soft Ball -
Guided), and Task 5 (Picking Up a Light Glass — Guided) and Task 6 (Picking Up a Heavy
Glass — Unguided) were compared. The novelty aspect was the major difference between
Tasks 2/4 and Tasks 5/6, as described earlier.

The three subjects were tested twice and the test results were pooled across ﬁve trials
for each of the dependent variables. The results are displayed in Tables 3 (Tasks 2 and 4) and
4 (Tasks 5 and 6). Figure 4 displays a CNV in Task 4 (Squeezing a Soft Ball — Guided), and

Figure 5 displays a CNV in Task 6 (Picking Up a Heavy Glass —- Unguided).

68

Comparison of Task 2 (Squeezing a Soft Ball — Unggided) with Task 4 (Squeezing_a
Soft Ball — Guided)

 

 

The subjects performed a task in the unguided condition ﬁrst, followed by the
identical task in the guided condition. The goup as a whole did not meet the criteria for
duration and average amplitude in the guided and unguided conditions, as shown in Table 3.
However, the goup met the criterion for average amplitude in both conditions. The goup
also met 63% of the 400 ms criterion for duration in the guided condition, and 70.3% of the
criterion in the unguided condition. Table 3 displays the data for Task 2 (Squeezing a Soft
Ball -— Unguided) and Task 4 (Squeezing a Soft Ball — Guided). The difference between the

two conditions was statistically insigiiﬁcant, as described later.

Duration and Average Amplitude of the CNV/Largest NPS for Task 2 (Squeezigg a Soft
Ball — Unguided) and Task 4 (Squeezing a Soft Ball - Guided)

The goup's mean for the duration and average amplitude of Task 2 (Squeezing a Soft
Ball - Unguided) was 281.300 ms (S_D = 52.0), and the goup's mean for average amplitude
was -0.815 volts (S_D = 0.323). For Task 4 (Squeezing a Soft Ball - Guided), the goup's
mean for duration of the CNV/largest NPS was 252.100 ms (S_D = 93.900). The goup's mean
for average amplitude was -0.755 volts (S_D = 0.180). In the guided condition, all individual
subjects met the requirement for minimum average amplitude, in both the guided and the
unguided conditions. The duration requirement was not met by every subject on either task.
Consequently, the CNV criteria of amplitude and duration were not met by any subject in the
unguided condition on any trial. But one of the three subjects met both CNV criteria in the

guided condition.

69

.88 .83 258388 amp—9.6 How canon-8 >20 85 @595 3
.a08 ma; dogmas-av HOW GOMHOHT—O >20 “ma maoﬁm a

$28 888> 8882 88:88 -

 

 

 

8N.NN2 80.8 NNN.o- o8.NN 82:82 mm
88.2: 8N.NNN N28- 8N._NN 8285 80:
8°.NN2 88.3. 83- 28.8 8830 am
8N.NNN 828 83. 8_.NNN 82:0 =82
NNQNN v8.5 .38...- NgNoN 82:8: N N
8588 8482 p.23- NSNM: 825 N N
cNENN 5.3 .38;- NNN.: N 8285 N N
N888 8N.NNN £82.. .838 825 N ...N
NNN.N8 8N.:N £82- ENSN 8838: 2 N
83: NNN8N .88...- Na .8N 886 2 N
N838 8:88 28- :ENN 8285 N _
NNvNNN 828 88...- N88NN 886 N _
88% GNNON 92:...- N8NNN 8385 N _
83.: N 88.3 583.. N28N 885 N 2
- - - - 82:8: _ 2
- - - - 886 _ _
Ame-£38 Ame-£05 20> 80? >20 as 88 >28
«02 =95>O «0.14. 3:03 oﬁazmg .w>< Amav Gown-SQ QOUEGOU ﬂoofzm ETC.

 

70

835 858920 982 82 823
88885 8 as, 8: 8m a 838 8m 5 N2 883

 

.825 I 8m «cm a 88088 a. 689
3582895 38on Aondwcb .- =mm com m @2883me N awn-- ﬁﬁcoatomxm co mmZ “mow-84 we 3.8802 62.339 .m 033-

In Task 4 only one CNV was established by Subject 2, Trial 2 (see Figure 4). With
regard to duration, his value was 1.5 ﬂ above the mean. When the value for average
amplitude was considered, he was within 1 SD of the mean. Subjects 1 and 3 produced large
NPS that did not meet the criterion for duration. However, the criterion for average amplitude
was met on each trial. In Task 2, no CNV was produced by any of the subjects. A large NPS
was established by Subject 3, Trial 2, meeting 90% of the criterion for duration. The criterion
for average amplitude was met.

To test whether signiﬁcant differences existed between the two conditions, a
Wilcoxon matched pairs signed-ranks test was done by pooling data across subjects and
trials. When Task 2 (Squeezing a Soft Ball - Unguided) was compared to Task 4 (Squeezing
a Soft Ball — Guided), no significant differences were found with regard to duration (T = 5)

and average amplitude (T = S) of the CNV/largest NPS.

Area of the CNV/Largest NPS and Overall Area of Negativities for Task 2 (SqueezingaSoﬁ
Ball - Unguided) and Task 4 (Sgieezinga Soft Ball — Guided)

The group's mean for the area of the CNV/largest NPS in Task 2 (Squeezing a Soft
Ball - Unguided) was 239.300 volts*ms (S_D = 36.6). The group's mean for the area of overall
negativities was 709.600 volts*ms (SD = 152.2). These values were longer than those
obtained for the guided conditon. The group's mean for Task 4 (Squeezing a Soft Ball -
Guided) was 182.700 volts*ms (S_D = 44.6) for the area of the CNV/largest NPS, and
558.300 volts*ms (ﬂ = 128.0) for the overall area of negativities.

The Wilcoxon matched pairs signed-ranks test was performed to test for signiﬁcant
differences with regard to the area of the CNV/largest NPS in both conditions. It was found

that the area of the CNV/largest NPS was signiﬁcantly greater in the unguided condition than

71

in the guided condition. In addition, the Wilcoxon matched pairs signed-ranks test was used
to test for signiﬁcant differences between the overall areas of negativities, recorded through
the entire sampling time, in both conditions. It was found that there was a signiﬁcantly larger

overall area of negativities in the unguided condition than in the guided condition [1‘ (ﬂ = 5)

= 0, p < .05].

Comparison of Task 5 (PickingﬁUp a Light Glass — Guided) and Task 6 (Picking Up
a Heavy Glass — Unguided)

Another guided and unguided task comparison focused on Task 5 (Picking Up a Light
Glass — Guided) and Task 6 (Picking Up a Heavy Glass — Unguided). Table 4 displays the
data for Task 5 (Picking Up a Light Glass — Guided) and Task 6 (Picking Up a Heavy Glass
— Unguided). Figure 6 displays a CNV by Subject 1, Trial 2, in the unguided condition. The
weight of the glass was the only difference between the two conditions, and it was not
assumed to change the basic nature of the tasks. The glass-weight difference was expected to
equalize task conditions. It was expected to maintain a novelty effect in the unguided
condition that occurred after the novel guided condition. This aspect was changed in order to
sustain the subjects’ attention during task performance.

The group as a whole did not meet the preset criteria for duration and average
amplitude of a CNV. The group consistently met the criterion for average amplitude in the
guided and unguided conditions. The duration requirement was not met by every subject on
either task. Consequently, the CNV criteria of amplitude and duration were not met by any
subject in the unguided and guided conditions on any trial. The 400 ms criterion for duration
was met by 80% of the group in the guided condition and by 96.5% of the group in the

unguided condition.

72

.208 223 0222:3822 $898 88 mom-282.8 >20 85 maxim :
.208 33 82886 .28 com-282.20 >20 2225 $526 .2

.9qu 8882/ 3882 2:822:00 ..

 

 

 

228.22: 88.82 N: .22- 8322 82:22:: mm
88.3: 822.NNN 822.22. 82 .oNN 82:8: :82
228.82 8N.2N 82.22- 2288 82:0 am
88.28 822.28 N88- 8.22NN 82:0 :82
2.228 8N.NNN 2822.22- 8NNNN 82:83 N N
8328 8N8: 22222.2- 222828 82:0 N N
2.8.28 882 .2832- NSNNN 82:82 N N
2.8.228 8228 23.22- .2882. 82:0 N _..N
88.2.8.2 N882. 22222.22. 88.8.. 82:82 2 ...N
28.82. 2.2 2.2 N 2.28.... 838 82:0 2 N
822.2: @2288 2822.2- 8$NN 82:83 N 2
N888 N8N22N 2:22.22- NNNNNN 8226 N 2
828 83:. 2:22.22- .3238 82:85 N _..2
822:. 228.82 22.8.22- 8NNNN 82:0 N 2
- - - - 82:8: 2 2
- - - - 82:0 2 2
282-3298 Ame-22:3 22:: 822A >76 as 82x >28
32 22830 32.42 $229; 08822228282 .w>< $82 802885 2235800 38.35 3.2.2-

 

73

.2825 8:8N2->ZU 9822 N22 823
222282262 8 22B 22.: 22on 2 28.38 22:82 .22 9272 2:882

 

282:8: - 22o N>822 2... 22: 822.22 o 2.5
Ecoagmxm mama-g 303:0 I 820 Ems a 83 82.22:: m van-H 88082-209212 no mmZ mow-:3 mo 3.5302 Naovaoaon .v 2an

.208 223 02522228“ ow80>u 20.2 802282.20 212022222 @525 .2
.208 mm? 22022826 .82 802282.20 >20 2822 952% .2
.928 8882/ 2,882 28222280 ..

 

 

 

74

oooNN2 8de c 2 N6. 002 Nm 2 waobm Gm
cow. 2 N2- 82 .33 NNnd- 25on m waobm 8832
ooN.Nm2 ooodm mde- 25on 233 Om
ooodcn oomdmN m2w.o- cem.2wN V23? 8332
$22.03 32 .aoM gag-c- :52 m mac-am m N
2.2 .wNN. VNONMN Acme-c- mvaom 23>? m N
mmnNmo 2%.an oovd- .” $.63 mac-22m N N
0N52Nm www.mVN :vwcé- MNwN2N 2.33 N N
236.3 mm 2 .ch 22.3.22- 269w:- wcobm 2 N
wNN.mow weNEN .2ch2- 222.de 233 2 N
N2m.mvo ommNN-N 2225.:- vame waobm m 2
Nohbmw m2w60N @2222- 2Nm2NN 233 m 2
2 O2 .222 N SYNON 2.23.22- 82 .vcm wnobm N 2
28.on 2oN.NON 235.c- Sc.mwN 2.83 N 2
- - .. - mac-um 2 2
.. - - - 233 2 2
28282222 2852232 32:: 80? >20 $2: 8? >28
32 22830 32 28232 0222222925. .w>< $82 22028.25 802222800 88.325 28.22.

 

228226 82:88-26 982 N22 22223
82282222 8 223 22.22 228.2 2 28.38 82.2 22 N22 28222222

 

20288232 wcobm .- 22Nm2 285 a waﬁooswmv m 232- Enogomxm
3283/ 2888332 233 I 22mm 220m 22 maﬁooscmv N 232- 282882.892m no mmZ 233282 .20 3.8.0.832 80282209 .m 022mb

Duration and Averﬁage Amplitude of the CNV/Largest NPS for Task 5 (Picking Up a Light
_G_lass - Guided) and Task 6 (Picking Up a Heavy Glass — Unguided)

For the guided condition, Task 5, the group's mean for duration of the CNV/largest
NPS was 320.0 ms (S_D = 86.9). The mean for average amplitude was -0.823 volts (S_D =
0.154). The group's mean for average amplitude was above the criterion, which means that
every subject met the criterion on Task 5. The group met 80% of the 400 ms criterion for
duration.

However, one of the three subjects, Subject 2, Trial 2, met both CNV criteria in the
guided condition. The duration of the CNV produced by Subject 2, Trial 2, was 430.631 ms,
which means that it was more than 1 SD above the mean. The average amplitude of -O.588
volts was within 1 S_D of the mean.

For the unguided condition, Task 6 (Picking Up a Heavy Glass), the group's mean for
duration was 386.100 ms (SD = 138.6). The mean for average amplitude was -0.849 (S_D =
0.173). The group’s mean was above the requirement for a CNV. Every subject met this
criterion in each trial of this task.

Two CNVs were established in the unguided condition. Subject 1, Trial 2, for
instance, established a CNV in the unguided condition that was more than 1 S_D above the
mean in comparison with the group's mean for duration of the CNV. The value for average
amplitude was within 1 SD of the mean (see Figure 6).

For Tasks 5 and 6, the Wilcoxon matched pairs signed-ranks test was used to test for
signiﬁcant differences between the guided and unguided conditions. Test results revealed no
signiﬁcant differences between the conditions. The value for T had to be less than or equal to
1 to be signiﬁcant. On the four dependent variables the values for T were as follows:

duration (T = 4), average amplitude (T = 6), area (T = 4), and overall area of negativities

75

(T = 5) suggesting that there were no signiﬁcant differences is task performance between the

guided and unguided conditions.

Area of the CNV/Largest NPS and the Overall Area of Negativities for Task 5 (Picking Up a
Light Glass - Guided) and Task 6 (PickingUp a Heavy Glass — Unguided)

For the guided condition, Task 5 (Picking Up a Light Glass -- Guided), the group's
mean for the area of the CNV/largest NPS was 241.800 ms (S_D = 51.2), and the mean for the
overall area of negativities was 621.800 volts*ms (S_D = 125.0). These values were lower
than the comparable ones obtained in the unguided condition (Task 6 — Picking Up a Heavy
Glass -- Unguided). The area of the CNV for Subject 2, Trial 2, was 1.5 S_D above the mean,
whereas the value for the overall area of negativities was within 1 S_D of the mean.

For Task 6 (Picking Up a Heavy Glass - Unguided), the group's mean for the area of
the CNV/largest NPS was 325.0 volts*ms (S2 = 122.4). The group's mean for the overall
area of negativities was 784.0 volts*ms (S_D = 248.0). Subject 1, Trial 2, established a CNV;
his area was within 1 SD of the mean, and his overall area of negativities was within 1 SD of
the mean.

As displayed in Table 4, three CNVs were observed altogether on Tasks 5 and 6. One
CNV was produced in the guided condition (Task 5, Picking Up a Light Glass -- Guided) by
Subject 2, Trial 2 (see Figure 5). Two CNVs were established in the unguided condition,
Task 6 (Picking Up a Heavy Glass — Unguided). The CNV of Subject 1, Trial 2, is displayed
in Figure 6.

The Wilcoxon Matched pairs signed-ranks test was performed to test for signiﬁcant
differences with regard to the area of the CNV/ largest NPS and the overall area of

negativities recorded throughout the entire sampling time. No signiﬁcant differences were

76

found when the areas of the CNV/largest NPS were compared between the two conditions.
However, a comparison of the overall area of negativities in the guided and unguided
conditions resulted in a signiﬁcant difference [I (E = 5) = 0, p < .05]. The T—value = 1. There
was signiﬁcantly more negative activity in the unguided condition than in the guided
condition. The fact that the overall area of negativities recorded throughout the entire
sampling time differed signiﬁcantly between the unguided condition and the guided
condition suggested that more mental activity was exerted in the unguided condition than in
the guided condition on these two tasks.

Although the group data did not indicate signiﬁcant differences between the two
conditions, the individual subject data showed different patterns. Subject I consistently
achieved higher values in the unguided condition, whereas Subjects 2 and 3 showed more
variable outcomes. At times they achieved higher values in the guided condition and at other
times in the unguided condition. In general, the effect of task presentation (novelty effect)
resulted in higher values for Experimental Task 5, where the subjects were guided without

having previous experience with the actual task itself as compared to Experimental Task 4.

Examining a Possible Novelty Effect Between Task 2 S ueezin a Soﬁ Ball - Unguided)
and Task 6 (Picki__ngUp a Heavy Glass -- (Urpguided)

Task 2 (Squeezing a Soft Ball — Unguided) was compared to Task 6 (Picking Up a
Heavy Glass -- Unguided), to test the novelty effect in the unguided condition. The Wilcoxon
matched pairs signed-ranks test did not reveal signiﬁcant differences between tasks for the
four dependent variables. The T-value for duration of CNV/largest NPS = 2, for average

amplitude T = 5, for area of the CNV/largest NPS T = 3, and for the overall area of

77

negativities T = 3. In the unguided condition, the effect of novelty versus familiarity was not

statistically signiﬁcant.

Research Question Four: Change of Resistance
Is there a difference between the CNV elicited from autistic subjects
when the event stimuli vary in the amount of resistance?

The goal was to determine whether there were differences in brain activity when the
subject independently touched two balls varying in the amount of resistance. Two tasks, one
representing weak resistance (squeezing a soft ball) and one representing strong resistance
(squeezing a hard ball), were designed to address this question. The subjects were tested
twice. The results for the three subjects pooled across ﬁve trials for each of the dependent

variables are displayed in Table 5.

Comparison of Tag 2 (Squeezing a Soft Ball — Unguided) and
Task 3 (Squeezinga Hard Ball - Unguided)

Task 2 (Squeezing a Soft Ball -— Unguided) offered weak resistance, whereas Task 3
(Squeezing a Hard Ball — Unguided) offered strong resistance. According to Affolter’s theory
of guided interaction, the nervous system codes tactual input in terms of resistance. Hence,
changes in brain activity should be observable when the form of resistance experienced by

the subject changes from weak to strong.

78

Duration and Average Amplitude of the Largest NPS for Tag 2 (SqueezinggSoft Ball;
Ugguided) Versus Task 3 (Squeezing a Hard Ball — Urggided)

For Task 2 (Squeezing a Soft Ball - Unguided, Weak Resistance), the group's mean
for the duration of the largest NPS was 281.300 ms (Si) = 52.0), and the average amplitude
was -O.815 volts (S_D = 0.323). The group as a whole did not meet the criteria for a CNV.
The group met 70.3% of the criterion for duration, as well as the criterion for average
amplitude. For Task 3 (Squeezing a Hard Ball - Unguided, Strong Resistance), the group's
mean for duration was 387.000 ms (ﬂ = 137.1). The group as a whole did not meet the
criteria for a CNV. However, the group met 96.75% of the criterion for duration. The group's
mean for average amplitude also was met.

The Wilcoxon matched pairs signed-ranks test was used to test whether any
signiﬁcant differences occurred between the two conditions (weak versus strong resistance).
The test results revealed no signiﬁcant differences between the two conditions on the
dependent variables of duration and average amplitude of the CNV/largest NPS.

Judging from Table 5, none of the three subjects met the required CNV criteria for
duration and average amplitude for either the weak or the strong resistance. Although the
group results showed no signiﬁcant differences between task conditions, data analyses for
each subject revealed some subject biases. Subjects 1 and 2 showed a greater amplitude
shift relative to the hard ball when the soft ball was squeezed (weak resistance). Conversely,
Subject 3 showed more amplitude when the hard ball was squeezed (strong resistance).
Subjects 1 and 2 showed increased duration of the largest NPS when the hard ball was
squeezed, whereas Subject 3 showed the same tendency in her second trial, but not in her

ﬁrst trial.

79

Area of the Lagest NPS and Overall Area of Negativities for Task 2 (Squeezggg Soft Ball
— Uguided) and Task 3 (Squeezing a Hard Ball — Unggided)

For Task 2 (Squeezing a Soft Ball — Unguided, Weak Resistance), the group’s mean
for the area of the largest NPS was 239.3 volts*ms (E 36.6). For Task 3 (Squeezing a Hard
Ball — Unguided, Strong Resistance), the group’s mean for the area of the largest NPS was
260.1 volts*ms (SD = 40.8). For Task 2 (Squeezing a Soft Ball — Unguided), the group's
mean for the overall area of negativities was 709.600 volts*ms (S_D = 152.2). For Task 3
(Squeezing a Hard Ball — Unguided), the group's mean for the overall area of negativities
was 721.400 volts*ms (ﬂ = 127.0).

Subjects 1 and 2 showed increased areas of the largest NPS when the hard ball was
squeezed. Subject 3 showed the same tendency in her second trial, but not in her ﬁrst trial.

With respect to the overall negativities recorded for the entire sampling time, Subjects
1 and 2 showed increased negativities when the hard ball was squeezed, whereas Subject 3
showed increased negativities when the soft ball was squeezed. Consequently, Subjects 1 and
2 showed the same biases, whereas Subject 3 almost consistently showed the opposite bias
with regard to the duration of the largest NPS, its average amplitude, its area, and the overall
area of negativities of the entire sampling time

A comparison between the areas of the largest NPS and the overall area of
negativities revealed a signiﬁcant difference [I (H = 5) = O, p < .05] in both conditions on the
Wilcoxon matched pairs signed-ranks test. This means that there was signiﬁcantly more
negative activity indicative of cognitive ﬁmctioning outside the area of the largest NPS

extending over the entire sampling time.

80

CHAPTER FOUR

DISCUSSION AND INTERPRETATION OF RESULTS

The present study was motivated by the difficulties arising from treating children with
severe impairments along the autistic/PDD spectrum because this population does not
respond to conventional treatment paradigms with the success one would expect. According
to Affolter and colleagues (1980, 1993, 1996b), one might assume that conventional
treatment paradigms require children to use their visual and auditory inputs at a supramodal
level. The prerequisite for this level is that the children have gained a multitude of
experiences following nonverbal explorations.

Within the framework, the representation of events at the nonverbal level is also
assumed to be a prerequisite for meaningful language to occur. To be functioning at this
relatively high level, a child must have experienced a fairly normal development in infancy
and early childhood. In other words, the child must have been able to explore his or her
environment by using tactile-kinesthetic input as a primary source of information to pave the
way for intermodal integration and serial organization of perceptual input.

Children with severe impairments might have problems with extracting adequate T-K
information from their environments, integrating perceptual input from various modalities
(vision and audition might not get aligned adequately with other senses), or organizing

perceptual information serially (events might not get sequenced appropriately).

8]

Assuming that the hypotheses proposed by Affolter and colleagues (1980, 1993,
1996b) are correct, it makes sense to intervene at exactly the level where the deﬁcits
occurred. For most severely impaired children, this would mean intervention at a nonverbal
level to allow the child to solve ﬁmctional problems, explore his or her environment, and
represent events of daily life mapped onto his or her experiences. Affolter and colleagues’
observations of normal infants led them to believe that guiding would help impaired children
make the connections they missed out on when they tried to make the explorations on their
own.

As Affolter and colleagues’ hypotheses address normal and impaired development of
children, the present investigation was designed to examine some of their basic assumptions
by testing exactly the clinical population that causes so many problems for clinicians -- the
children along the autistic/PDD spectrum. If tactile-kinesthetic input is primary to visual and
auditory input in development and is needed to anchor the other senses, then children with
perceptual disturbances should beneﬁt from the guided interaction paradigm and one should

be able to measure its impact.

Applying the CNV as a Measurement Tool
Over the past decades, electrophysiological studies have been designed to learn more
about human cognitive functions by measuring the electrical activity of the brain. In
particular, activities in response to one or two identiﬁable stimuli such as event-related
potentials or CNV have been subjected to many investigations. Surface negativity that
depends on the association of two successive stimuli is called CNV (Donchin et al., 1972;
Tecce & Cattanach, 1993). Its magnitude is increased when a motor response is made to the

second stimulus (Peters et al., 1970); this is due partly to enhanced attentiveness resulting

82

from greater task involvement and partly to the addition of a readiness potential
(Bereitschaftspotential). Peters et al. (1977) reported that as a paired-associates learning task
became easier, the CNV decreased. The researchers interpreted this ﬁnding as the result of
decreased interest and alertness. Nakamura et al. (1979) observed that when careful $2
discrimination required greater attentiveness. The CNV was enhanced. F enelon (1984) used
attention to explain the increased CNV with greater equivocation between the match of SI
and 82.

1f the CNV is a correlate of attention, as indicated in the literature cited above, it
might be a useful tool to measure some theoretical assumptions about cognitive processes as
claimed by the guided-interaction paradigm. The CNV has been used to test normal adults on
a S] and 82 standard-reaction-time paradigm (Donchin et al., 1972; Irnashioya et al., 1987;
Lai, 1997; Tecce et al., l978d; Washer et al., 1996). It has also been used to compare normal
adult performance of neurotics and psychotics (Timsit-Berthier et al., 1973) and traumatic
brain-injured individuals (Nativ, 1991; Rugg et al., 1987). However, the number of studies
conducted with normal and/or impaired children is relatively small. McPherson (1998)
conducted a study to investigate phonological processing in reading-disabled adolescents,
and Kemner (1994) carried out an investigation to compare visual and somatosensory event-
related brain potentials in autistic, ADD, dyslectic, and normal children. To the investigator's
knowledge, no CNV study, has been conducted with autistic children using the kinds of
stimuli used in the present study with the possible exception of the Push Button Task.
Consequently, this researcher attempted to discover (a) whether the CNV paradigm was
feasible with an impaired population -- in particular, autistic children; (b) whether brain
functioning indicative of attention was detectable in the guided condition; (c) whether there

were differences in brain functioning between the guided and unguided conditions; and

83

((1) whether there were observable differences in brain functioning when the event stimuli

varied in the amount of resistance.

Addressing the Methodological Aspect of the CNV

Experimental Task 1 (Push Button) was carried out to investigate whether the CNV
was a possible measurement for the autistically impaired population. As the CNV had not
been commonly used with children, the suitability of this paradigm for autistic subjects had
to be established ﬁrst. In particular, one researcher found that learning disabled children
failed to establish a CNV, as described earlier.

With the technology and equipment available in the Neurophysiology Laboratory at
MSU, CNVs were observed in unimpaired individuals. Consequently, the investigator
assumed that it was possible to observe a CNV with the available technology. For the stated
reasons, the ﬁrst research question concerned the methodological issue of whether a CNV
could be elicited in autistic subjects.

Altogether, ﬁve CNVs were observed in two of the autistic subjects in the present
study. On four of six tasks, including the prototypic Push Button Task, a CNV was
established. A CNV was also observed in the normal pilot subject.

With regard to Task 1 (Bush Button), Subjects 1, 2, and 3 were tested on two trials.
One CNV was observed in Subject 1 (see Figure 1), Subject 2 established a large NPS (see
Figure 2), and Subject 3 exhibited a large NPS that was similar in shape to a CNV (see
Figure 3). Although a CNV was observed on the Push Button Task (Subject 1, Trial 1), it
needs to be emphasized that a CNV was not observed every time an autistic subject was
tested. Nevertheless, all of the subjects were able to sustain attention for a certain period of

time, and the group's mean for average amplitude was above the criterion (as deﬁned by the

84

investigator as a minimum requirement for a CNV) on every task. This outcome suggests that
the autistic subjects were able to exhibit brain activity of the required magnitude on every
task. However, it was harder to sustain the negativity over the necessary period of time on
each task.

The fact that one subject was able to establish a CNV on the Push Button Task
demonstrates that on this task this particular subject was able to attend in a way expected of
an unimpaired learner, and he was able to do so on a task that had also elicited CNVs in
normal subjects.

Although Subject 2 did not establish a CNV on the Push Button Task, he established
CNVs on other tasks and conditions. The Push Button Task might not have captured his
attention as much as other tasks. Subject 3 developed a large NPS very similar in shape to a
CNV. As the conﬁguration of her largest negative potential shift on two tasks was similar to
the CNVs observed in Subjects 2 and 3, one might speculate that some autistic subjects failed
to establish a CNV at the criteria set for this study because of an inability to sustain their
attention long enough to allow the classic CNV pattern to form. Indirect support for this
speculation comes from an investigation on phonological processing in reading-disabled
adolescents and normals by McPherson et al. (1998). The researchers found a reduced CNV
in the reading-disabled group. In addition, they found ERP evidence suggesting that the
reading-disabled population had two distinct abnormal brain processes (phonological
processing and speed of processing).

Subject 3 also failed to establish a CNV on the Push Button Task and she also failed
to produce a CNV on any other task. Because this subject produced a large NPS with the
same kind of conﬁguration on another task, one might assume that her pattern of brain

activity was slightly deviant as compared to that of Subjects 1 and 2. However, whether

85

differences in attention are observable outside the laboratory is a different question
altogether. The differences between a large NPS and a CNV are expressed in ms, an amount
of time that is hardly noticeable behaviorally. Hence, the question of whether the difference
in the ability to pay attention between Subjects 1 and 3 extends beyond the laboratory setting
needs to be investigated in future research.

The dependent variable of overall negativities of the entire sampling time was not
mentioned in the literature as a measure to evaluate brain activity in CNV studies because it
does not give additional information about the actual CNV. However, it does provide a
measure of how much time and amplitude the subject invested in actual cognitive processing
during the entire task. Subtracting the CNV parameters from the entire sampling time would
also show how much cognitive functioning was taking place outside the actual area of the
CNV. The Wilcoxon matched pairs signed-ranks test revealed a signiﬁcant difference
between the area of the CNV and the overall area of negativities. This demonstrated that the
subjects engaged in more cognitive activity outside the area of the CNV than inside the CNV
area. Consequently, one might assume that cognitive processing extended over the entire
sampling time regardless of whether a CNV was elicited.

When looking at the tasks altogether, ﬁve CNVs were established by two of the three
autistic subjects. So it does appear that, in response to the ﬁrst research question, CNVs can
be elicited in children with autism. However, the interpretation of the ﬁndings as evidence of
attention needs to be tempered. On the basis of the investigator's behavioral observations,
Subject 3 appeared to be very attentive on all tasks throughout both trials. However, she
failed to establish a CNV that met the deﬁned criteria on any of the tasks.

Another mitigating factor is the severity of the symptoms displayed by the subjects.

Although the autistic subjects clearly were moderately severe to severely impaired, they did

86

not appear to have the kinds of perceptual dysfunctions that are exhibited by children who
are often referred to Affolter and colleagues’ clinic as autistic by other professionals. Affolter
and coworkers have characterized the classic autistic child as one who cannot integrate the
different modalities. A child meeting Affolter and coworkers’ deﬁnition would not be able to
perform the Push Button Task, for instance, because of inability to relate the nonverbal
tactual instructions to the visual input. In other words, the child or subject would not have
known what to do. It stands to reason that a child with that degree of impairment would have
been untestable on any of the tasks attempted by this investigator. It is, however, important
to consider the severity of an impairment when considering the feasibility of an ERP study
with children diagnosed as autistic.

Thus, one should not get the impression from the present study that this population is
easy to test. The investigator had worked with the subjects on a regular basis for almost a
year before attempting any testing. This approach was time consuming and required the
cooperation of many parties, but it was essential for successful testing of unmedicated
autistic subjects. To make the subjects as comfortable as possible, their aide accompanied
them to the laboratory. The aide sat in the same room with the subject and also watched the
subject closely when the investigator had to leave the child for a moment to attend to the
computer. During school hours, the investigator picked up the autistic subjects and their aides
from school to take them to the Neuro-Audiologic laboratory and drove them back to school.
The support of the school administrators, the AI classroom teacher, the aides, and the speech-
language pathologist for the AI children in the district was absolutely essential for this
procedure. All these parties gave their support to this study.

The positive ﬁnding that a CNV can be elicited in autistic subjects is encouraging. It

has future potential for investigating attention while certain tasks are performed online. The

87

CNV paradigm might be used not only in the study of autistic behaviors but also in the study
of other clinical populations such as children with speciﬁc-language impairment (SLI). With
regard to the present investigation, inclusion of more autistic children is needed for purposes

of ensuring reliability of the data.

Addressing the Clinical Conceptual Questions

Three research questions were posed to investigate clinical conceptual issues. One
question concerned whether it was possible to elicit a CNV in the guided condition. Guiding
as a means of intervention is a hotly disputed issue in the ﬁeld. Opponents of Affolter and
colleagues’ theoretical framework claim that the brain does not process input given to the
child during guiding. The reasoning is that guiding is a passive act for the individual who is
being manually led to perform a functional task. The person guiding might be actively
involved in performing the task, but not the person that is being guided.

Affolter and colleagues, on the other hand, argued that the child can be actively
involved in performing the task while being guided. When a child is active and attentive,
prerequisites for learning are fulﬁlled. The present study was designed to shed some light on
this controversy. The ﬁndings are interpreted in View of what was learned about guiding and

the role of resistance as follows:

Elicitﬁg a CNV in the Guided Condition

Subjects 1, 2, and 3 were guided on two experimental tasks. The difference between
the tasks was the novelty effect, which will be elaborated on later. Only Subject 2 established
CNVs in both guided experimental tasks on his second trial. Subject 3, Trial I, achieved a

very large NPS, and Subject 1, Trial 2, produced a very large NPS as well. Hence, a CNV

88

can be elicited in the guided condition, although it was observed in the same subject in both
the novel and the familiar tasks. It should be noted that Subject 2 was the most severe of the
three autistic subjects, and he was very attentive throughout his second trial, more so than
throughout his ﬁrst trial. The largest NPS, for Subject 1, Trial 2, fell short of the requirement
for minimum duration of the CNV by only 32.9 ms, but the criterion for average amplitude
was met. It needs to be pointed out that Subject 1’s overt behavior suggested that he was
very attentive throughout his ﬁrst trial but much less so during the second one. Nevertheless,
he still established a large NPS on the task in question. Most of Subject 1's data on the ﬁrst
trial were not stored, and all attempts to retrieve the saved data failed. Subject 3 failed to
establish a CNV in the guided condition and did not meet the preset criteria of this study.
However, the conﬁguration of her NPS was similar to the CNVs established by Subjects 1
and 2.

It appears reasonable to assume that CNVs can be elicited in the guided condition in
more than one subject. One reason why Subject 2 established a CNV on every guided task
might be that he beneﬁted the most from the guided input, even when the task was familiar,
as he was the most severe of the three subjects. Although Subjects 1 and 3 established large
NPS on the novel task, they might not have shown a CNV because they realized what the
investigator wanted them to do more quickly than did Subject 2. One must keep in mind that
guiding is only a means to enable the severely impaired child to learn how to perform a
certain task independently. Consequently, the frequency of guiding that children need will
vary, depending on the degree of their impairment, and so might the emergence of the CNV.

Both CNV and largest NPS, which failed to meet CNV criteria, are assumed to result
from cortical activity and indicate attention, as described earlier. Thus, it is reasonable to

assume that all of the subjects were attentive while being guided. If they were attentive, then

89

they had to be actively involved in the task, even though they were not performing the task
independently.

The Wilcoxon matched pairs signed-ranks test revealed a signiﬁcant difference
between the area of the CNV/largest NPS and the overall area of negativities, suggesting that
the area of overall negativities was signiﬁcantly different from the area of the CNV/largest
NPS. Consequently, considerably more brain activity was observed outside the area of the
CNV/largest NPS than inside it. Taking this fact into account, it seems reasonable to use the
measure of overall area of negativities as another indicator of cognitive activity during the

entire sampling time.

Novelty Effect as a Factor in the Guided Condition

To test for signiﬁcant differences between the two guided conditions, the Wilcoxon
matched pairs signed-rank test was used. Comparison of the duration of the CNV/largest
NPS in the two tasks (novel versus familiar) showed a signiﬁcant difference. The subjects’
responses to guiding were tempered by the novelty of the task. The duration of the
CNV/largest NPS was signiﬁcantly longer during the novel task than the familiar task for all
subjects in all trials. The average amplitudes were not signiﬁcantly different in the two tasks.

A signiﬁcant difference also was found when the area of the CNV/largest NPS was
compared in the two tasks. The area of the CNV/largest NPS in the novel task was
signiﬁcantly larger than the area in the familiar task. The overall area of negativities, on the
other hand, did not reveal a signiﬁcant difference between the two tasks.

Hence, larger CNVs/NPSs were found in the novel task as compared to the familiar
task. This trend held for all three autistic subjects. Bloom (1996) pointed out that the level

of effort is closely connected to the arousal level and determines the interest or level of

90

engagement that the child brings to the task. In her research, Bloom found evidence that
effort and attention were increased in novel situations in contrast to familiar ones. This
ﬁnding provides support for the ﬁnding in the present study. All subjects exhibited larger
NPS with regard to the dependent variables of duration and area in the novel task than in the
familiar task, regardless of whether CNV criteria were met.

One needs to consider, though, that autistic subjects as a group might exhibit a
different behavior toward novel stimuli than do normal controls or children with other kinds
of impairments. The ﬁndings of the Kemner et al. (1994) study are relevant here. These
researchers investigated whether autistic children differed from normal children, those with
attention-deﬁcit disorder, and dyslectic children. The researchers used tasks involving
processing of visual and/or somatosensory stimuli. The autistic group showed larger N1 and
P3 (indicators of cognitive functioning and attention used in ERP studies) to novel stimuli in
the visual and somtosensory modality. Although Kemner et al. used different measures, the
results were clearly consistent with the outcomes of the present investigation. The clinical
observation that autistic children react more strongly than normal children to changes in their
environment might also have contributed to the ﬁndings regarding the novel task in the
present study.

In sum, then, the present investigator found that at least one autistic subject
established a CNV on two different guided tasks, and two other subjects came close to the
parameters deﬁned for a CNV. Hence, it is possible to get a child's attention under
guidance. The effect, however, might not always reach the dimensions needed to be deﬁned
as a CNV. Given that it is possible to establish a CNV under guided task conditions

(Squeezing a Soft Ball and Picking Up a Light Glass), and when the subject performs a task

91

independently (e.g., Push Button Task), the next question is whether there are differences in

brain activity between the two conditions.

CNV Differences between Guided and Unguided Conditions

To investigate this question, it was necessary to compare responses under identical or
similar task conditions. In this study, squeezing a ball was compared in the guided and
unguided conditions. Picking up a light glass was compared to picking up a heavy glass in
both the guided and unguided conditions. Two CNVs were observed in the guided condition
and two in the unguided condition. Subject 1 established CNVs only in the unguided
conditions. Subject 2 established CNVs in the guided and unguided conditions. Subject 3
established three very large NPS (Trials 1 and 2) in the unguided condition. Her largest NPS
fell short of the minimum requirement for duration by 26.6 ms. The criterion for average
amplitude was exceeded by 0.365 volts.

The Wilcoxon matched pairs signed-ranks test was used to compare the results in the
guided and unguided conditions. No signiﬁcant differences were found when the areas of the
CNV/largest NPS and the average amplitude of the CNV/largest NPS were compared for
squeezing a soft ball in the two conditions. One has to keep in mind that Task 4 (Squeezing a
Soft Ball) was familiar to the subjects by the time they completed it in the guided condition.

However, a signiﬁcant difference was found when the area of the CNV/largest NPS
in the guided condition was compared to the area of the CNV/largest NPS in the unguided
condition. The latter was signiﬁcantly larger in the unguided condition than in the guided
condition. When the overall area of negativities in the guided condition was compared to the
overall area of negativities in the unguided condition, a signiﬁcant difference was also found.

The overall area of negativities was signiﬁcantly larger in the unguided condition than in the

92

guided condition. This outcome is consistent with the ﬁndings of Carr et al. (1996), as
described earlier. The researchers found that more brain areas were active when the subject
lifted a ﬁnger independently than when he or she was guided to perform the same task. One
might speculate that the independent planning and execution of a motor task, for instance, is
more complex and requires the allocation of more cognitive resources than a task condition
designed to facilitate performance by guiding. In contrast, no signiﬁcant differences were
found between guided and unguided conditions with regard to duration and average
amplitude in Task 2 (Squeezing a Soft Ball -- Unguided) and Task 4 (Squeezing a Soft Ball -
- Guided).

For Tasks 5 (Picking Up a Light Glass -- Guided) and Task 6 (Picking Up a Heavy
Glass -- Unguided), the Wilcoxon matched pairs signed-ranks test was used to test for
signiﬁcant differences between the guided and unguided conditions. On the four dependent
variables -- duration of the CNV/largest NPS, average amplitude of the CNV/largest NPS,
area of the CNV/largest NPS, and overall area of negativities -- no signiﬁcant differences
were found between the guided and unguided conditions.

When the area of the CNV/largest NPS was compared to the overall area of
negativities in the guided condition as well as in the unguided condition, signiﬁcant
differences were found. This suggests that there was more brain activity outside the
CNV/largest NPS than inside it.

It is interesting that the area of the CNV/largest NPS and the overall area of
negativities were signiﬁcantly larger in the unguided condition when the task was familiar,
whereas this difference disappeared when the task became novel.

The ﬁnding for Subject 1, who established CNVs only in the unguided condition, is in

agreement with the ﬁnding of Carr et al. (1996). The researchers compared an active

93

performance condition in which the subject moved his hand independently with a “passive”
sensory baseline condition in which the same movement was produced by the experimenter
manipulating the subject's hand. Activity in the somatosensory cortex, proximate parietal
association areas, and the motor cortex was measured using ﬂVIRI. The active performance
condition produced more activated pixels than did the passive condition.

In contrast, Subject 2’s data were not consistent with the Carr et al. (1996) ﬁndings.
This subject established CNVs in both conditions. However, as Thelen (1995) stated, the
perceptual system extracts only the information it needs. Subject 2 might have needed to
focus almost as much in the guided condition as in the unguided condition, whereas Subject 1
might have found the task requirement in the guided condition less taxing.

Subject 3, on the other hand, did not establish a CNV in either condition. Like
Subject 1, she produced larger NPS in the unguided condition than in the guided condition
with the exception of the novel task, Task 5 (Picking Up a Light Glass), where she had a
much larger NPS in the guided condition than in the unguided condition during her ﬁrst trial.
In her second trial, she showed the opposite pattern. One might assume that she learned in the
ﬁrst trial and was able to use her knowledge fully in her second trial in the unguided
condition.

With regard to Affolter and colleagues’ (1980) claim that the input to the brain is the
same in both conditions, the following was observed in the present investigation: The
measurement outcomes did not differ statistically when the subject was faced with a novel
situation. When faced with a familiar situation, the subject appeared to exert more mental

energy in the unguided condition, at least in terms of the area of the CNV and overall area of

negativities.

94

The developmental stage of the child also appears to be a decisive factor. The most
severely impaired subject (Subject 2) was the only one who established a CNV in the guided
condition on two different tasks. Guiding appeared to be most beneﬁcial to him. It needs to
be emphasized, though, that Subject 1 established large NPS approaching the criteria deﬁned
for a CNV in the guided condition. As mentioned earlier, Subject 1 might have been able to
make some hypotheses about the goal of the tasks more rapidly than did Subject 2. Subject 1
might beneﬁt more from guiding when a task becomes more cognitively challenging.

Subject 1 showed more negativity in the unguided condition. Subject 2 showed
increased activity in the unguided condition on his ﬁrst trial and increased negativity in the
guided condition on his second trial. He appeared to be more variable. Subject 3 showed the
same variable pattern, although she had difﬁculty sustaining the duration and average
amplitude of her NPS long enough to establish a CNV.

The assumption that the input given to the brain is the same in the two conditions was
not sustained by the present data. The information appears to get to the brain. But how much
a child beneﬁts from guiding appears to depend on the child’s developmental stage and the
difficulty of the task. When performance of a certain task is beyond a certain child’s zone of
ability, guiding may become necessary to help the child move ahead. If a certain task is well
within a child’s zone of competence, the unguided condition will elicit greater negativities as
the child is in charge of planning and executing the task.

However, the question of interest to the clinician is whether the children can learn
under the guided condition; the answer to that query has to be in the afﬁrmative. To compare
performances in the guided and unguided conditions in more detail, a more reﬁned method of
analysis might be necessary. A method that allows EEG segments to be analyzed in the

alpha bands that have been shown to reﬂect attentional processes while using more than one

95

channel might be a possible way to proceed, provided that undesired artifacts can be

controlled.

Change of Resistance

The fourth research question dealt with the role of resistance in human perception.
Affolter and colleagues (1980, 1993, 1996b) proposed that tactual input is coded in varying
degrees of resistance. If this assumption is correct, one should be able to observe differences
in brain activity when the amount of resistance varies. Differences in the CNV elicited from
autistic subjects were investigated when the event stimuli varied in the amount of resistance.
The Wilcoxon matched pairs signed-ranks test was used to determine whether any
differences occurred between the condition offering weak resistance and the condition
offering strong resistance. No signiﬁcant differences were observed on any of the four
dependent variables -- duration of the CNV/ largest NPS, average amplitude of the
CNV/largest NPS, area of the CNV/largest NPS, and overall area of negativities -- during the
entire sampling time. However, for both conditions, signiﬁcantly more negativity for the
entire area was recorded, suggesting (as mentioned before) that there was more activity
outside the CNV/largest NPS area than inside it. The statistical outcomes were surprising to
the investigator, but a failure to observe a difference does not mean that the subjects were not
able to perceive a difference.

A look at the individual data for each autistic subject revealed an interesting pattern.
When they squeezed the hard ball, the two autistic boys showed increased duration of the
CNV/largest NPS on all measures. The only dependent variable that was lower when the
hard ball was squeezed was the average amplitude of the CNV/largest NPS. When the soft

ball was squeezed, Subjects 1 and 2 showed the opposite pattern. With the exception of

96

average amplitude of the CNV/largest NPS, all values were lower for the latter condition,
which represented weaker resistance. The autistic girl, on the other hand, showed mixed
results on her two trials, so no clear pattern was established.

According to the guided-interaction paradigm, changes in brain activity were to be
expected when a child squeezed a soft ball versus a hard ball. The assumption is that
perceptual impairment exists because the brain cannot easily identify the input in terms of
changes in resistance. Hence, one would expect the impaired child to exert more of an effort
when squeezing a soft ball as compared to a hard ball. Visual inspection of the waveforms
revealed differences in the two conditions (Figures 5 and 6). However, no statistically
signiﬁcant differences were found. The only signiﬁcant differences found were when the
overall area of activities was compared to the area of the CNV/largest NPS in the two
conditions.

The fact that no signiﬁcant differences in the two conditions were found for the four
dependent variables for the autistic subjects was surprising. It was assumed that the nervous
system would have to search harder for information when weak resistance was encountered.
However, a Wilcoxon matched pairs signed-ranks test comparing the condition offering weak
resistance with the condition offering strong resistance revealed signiﬁcant differences for
the normal pilot subject (I = 0). The I—value had to be less than or equal to 1 to be
statistically signiﬁcant. The normal pilot exerted signiﬁcantly more mental energy when
squeezing the soft ball as compared to the hard ball.

The results for the autistic subjects can be partially explained by the fact that they
wanted to achieve the same effect when they squeezed the hard ball as they had experienced
when they squeezed the soft ball. They were keenly aware of the change of resistance

offered by the hard ball. The investigator noticed that all three subjects (Subjects 1 an 2 on

97

both trials, Subject 3 on Trial 1) spontaneously enlisted the other hand when they felt the
hard ball to help them search for more information, whereas the normal pilot was aware of
the differences in touch when he felt a hard surface versus a soft surface. As a result, more
effort exerted with two hands instead of the one hand used to squeeze the hard ball resulted
in greater negative readings on three of the four dependent variables. The normal pilot

subject, however, kept to the instructions and used just one hand.

Probing Differences Between Autistic and Normal Subjects

This study included only subjects with a diagnosis of autism. Hence, an important
question, whether the kinds of data observed in this study would differ from what would be
observable in normal children, still needs to be addressed. In the present study, this issue
could not be explored as a research question. However, some preliminary normal pilot
observations were made that might provide directions for future research.

A typically developing 10-year-old pilot subject was tested. The investigator
evaluated him on the test Clinical Evaluation of Language Fundamentals (CELF 3).
Performance in the normal range was observed. The researcher also read his progress reports
from school, which showed grade-level academic achievement.

There were two differences between the testing procedures used for the normal pilot
and the autistic subjects. The number of trials (the number of times one action of a particular
task was repeated and averaged with the EEG apparatus) was 16 for the pilot subject instead
of 6 as it was for the autistic subjects. The response box for experimental Task 1 (Push
Button Task) was still being constructed when the pilot subject was tested. Consequently,
that task was not performed. Results for the normal pilot subject are presented in Table 5 for

Task 2 (Squeezing a Soft Ball — Unguided), Task 3 (Squeezing a Hard Ball — Unguided),

98

Task 4 (Squeezing a Soft Ball — Guided), Task 5 (Picking Up a Light Glass — Guided), and
Task 6 (Picking Up a Heavy Glass — Unguided). The normal pilot investigation was aimed at
answering several questions that paralleled the main research questions concerning the

autistic subjects. These questions are identiﬁed in each of the following discussion sections.

Research Question 1 — Is it possible to elicite a CNV in the Normal Pilot Subiect
under the task conditions of the autistic subiects of this study?

This question was aimed at the impaired population. Normal children have been
known to establish a CNV, as described earlier. The normal pilot subject was not tested on
the Push Button Task, which was designed to test whether the autistic subjects were able to
establish a CNV for reasons already stated. However, it seems reasonable to emphasize that
the normal pilot subject established a CNV. His CNV had a duration of 601 .676 ms and an
average amplitude of -0.8605 volts. The normal pilot subject’s CNV exceeded the 400 ms
criterion for duration by 201.676 ms and the -0.500 volts criterion for average amplitude by
-0.3605 volts. He established the CNV on Task 2 (Squeezing a Soft Ball — Unguided), see

Table 6. It is also interesting that a subject was able to establish a CNV on this task.

Research Question 2 —— Is it possible to eliciti_nga CNV in the guided condition:
Performance of normal pilot subject in the gmded condition as compared
to the autistic M
The normal pilot subject did not establish a CNV in the guided condition. He met the
criterion for average amplitude on both guided tasks. However, he only reached only 67.9%

of the 400 ms criterion for duration on experimental Task 4 (Squeezing a Soft Ball —

99

.208 N223 02082228“ 082226 202 802-28220 >720 822: 952m .2

.208 mm? 22028820 202 802-28220 >20 85 362m «
.928 25828> 9,882 2:88:00 ..

 

 

100

2NN.N8 NN§N2 NES- osNNN 82:8: 8 3:20 8,822 : n2: 822:
NNNNNV N82 2 N22. 822NNN 82:0 N 820 22222 a 22: 8202
888 28.2.2 8N3- 2NN. 2 N 82:0 2. 22m 28m 8 82888
8.82. 822. 2 ON 8%.:- NNN8N 82:8: N 228 88 22 888:8
N858 N82 2 N 2N82222- .N2$.28 82:8: N 22m 28m .2. 82888
28.8280 282.3220 N222 8N? >20 22: 8% >28
32 22805 «02¢. 2.0.2203 020822222862 .w>< 882 8022825 8022220800 02822- 80>m2

 

2822.20 82:8:->20 N82 82 222223
222282222 8 223 2222 228m 2:. 28.38 22:22 22 82 288...:

 

28.38 22:82 22 an: 82252 8 282

Guided), and only 56.8% of the 400 ms criterion on experimental Task 5 (Picking Up a Light
Glass — Guided).

One of the autistic subjects (Subject 2, Trial 2) established a CNV on both guided
tasks. In addition, on experimental Task 5 (Picking Up a Light Glass — Guided), the values
for the autistic subjects were higher. Subject 3, Trial 1, met 80.7% of the 400 ms criterion for
duration, and Subject 1, Trial 2, met 91.8% of the criterion for duration. The normal pilot
subject met only 56.8% of the criterion. The difference between the two guided tasks was the
novelty effect. The autistic subjects appeared to achieve higher values on the novel task than
on the familiar task. The values for the familiar task (Squeezing a Soft Ball — Guided) were
comparable for the autistic subjects and the normal pilot subject.

The higher values for the novel task might be interpreted as an indication of increased
cognitive processing in the impaired group under the guided condition as compared to the
normal pilot subject. This might mean that the autistic group as a whole beneﬁted more from
guiding (getting more information in that condition) than did the normal pilot subject.

The fact that the impaired subject showed more brain activity in the guided condition
than did the normal pilot subject is in agreement with Affolter and colleagues’ ( ) claim that
guiding helps perceptually impaired children perceive the information they need. However,
their claim that the input to the brain is the same in the guided and unguided conditions
appears not to be supported by the present data. The more severe the child’s perceptual
impairment, the more the child will proﬁt ﬁ'om guiding. The less severe the impairment, the
less beneﬁcial guiding becomes. Guiding clearly appears to be a tool to help children
perceive the input. However, once they can extract the information from the environment on
their own, they will show greater negativities in the unguided condition, reﬂecting the

complexity involved in mental processing and motor execution of a simple task.

101

Research Question 3: Differences between the Guided and Unguided Condition
in the Normal Pilot Subject

Data for Task 2 (Squeezing a Soﬁ Ball — Unguided) and Task 6 (Picking Up a Heavy
Glass — Unguided) were compared, as shown in Table 6. The normal pilot subject established
one CNV on experimental Task 2 (Squeezing a Soft Ball — Unguided). On experimental Task
6 (Picking Up a Heavy Glass — Unguided), he met the criterion for average amplitude, but he
met only 58.8% of the 400 ms criterion for duration.

Two of the autistic subjects (Subject 2, Trial 1, and Subject 1, Trial 2) established
CNVs on experimental Task 6 (Picking Up a Heavy Glass) in the unguided condition; see
Table 4. The individual data for the autistic subjects show that in more than one instance the
unguided condition yielded higher values for duration and average amplitude of the NPS.
However, there were also instances in which the reverse was the case. More autistic and
typically developing subjects are needed to reveal whether there is a trend or whether the

same range of variability will be exhibited in a larger sample.

Research Question 4: Change of Resistanﬁ
The normal pilot subject appeared to be sensitive to varying degrees of resistance.
Data for Task 2 (Squeezing a Soft Ball — Unguided) and Task 3 (Squeezing a Hard Ball -
Unguided) were compared, as shown in Table 6. The normal pilot subject established a CNV
on Task 2 (Squeezing a Soft Ball -- Unguided) in the condition where weak resistance was
encountered. On Task 3 (Squeezing a Hard Ball -- Unguided), the normal pilot subject met
the criterion for average amplitude, but he met only 87.3% of the 400 ms criterion for

duration.

102

The autistic subjects did not establish a CNV on either task. However, the values for
Task 3 (Squeezing a Hard Ball - Unguided) appeared to be slightly higher for the autistic
subjects as compared to the normal pilot subject. The criterion for average amplitude was
met by all autistic subjects on Task 3. Subject 2, Trial 2, exceeded the criterion for duration
by 217.44 ms. Subject 1, Trial 2, met 94.7% of the 400 ms criterion for duration, and subject
3, Trial 2, met 79% of the 400 ms criterion for duration.

The normal pilot subject appeared to exert less mental energy than the autistic
subjects. This ﬁnding is not surprising because the autistic subjects used two hands on almost
every trial to achieve the same effect as they had experienced on the previous task. As a
result, more brain activity was observable. This ﬁnding may imply that the tactile-

kinesthetic system was insufficiently developed in the autistic group.

Implications for Future Research

In this study, altogether six CNVs were elicited in the autistic subjects and the normal
pilot subject. Judging from this value alone, it is reasonable to assume that the CNV-
paradigm is a possible tool to investigate cognitive processes as indicative of attention in
normal and impaired subjects. The fact that autistically impaired children were able to
establish a CNV, meeting the criteria set in this study, is promising. It shows that some of the
autistically impaired subjects were capable of exhibiting attention patterns similar to his or
her unimpaired peer. It also shows that the CNV paradigm might be a usable tool for the
impaired population in general. As described earlier, there are studies showing that impaired
children are not able to establish a CNV. Of course, more impaired and unimpaired subjects

need to be tested on a CNV paradigm to conﬁrm or disconﬁrm the present ﬁndings.

103

As the ﬁndings for the normal pilot subject were derived from only one normal
control subject in one trial, more normal control subjects need to be tested under the same
conditions and with the same number of trials as the autistically impaired group. The
inclusion of a normal control group is essential to investigate whether the ﬁndings of the
normal pilot subject would be corroborated by the ﬁndings of a normal control group.

The number of the autistic subjects was small. As the data of the autistic group
demonstrated, autistic subjects did not show the same patterns of performance throughout the
trials. If more autistic subjects were added to this investigation, one might be able to observe
if there are certain tendencies in the autistic population depending on the level of impairment
and the developmental stage of the subjects.

Although CNVs were elicited with the laboratory equipment available to the present
investigation, some changes in the equipment would be beneﬁcial with regard to future
outcomes. For instance, if one could vary the interstimulus interval, tasks could be designed
that would add a little bit more complexity and give the subjects more time to process the
stimuli.

The dependent variable of overall negativities recorded during the entire sampling
time to evaluate brain activity might be a useful measure for between and in-group
comparisons, in particular, at some point in the future when more subjects can be added to

the investigation and reliable comparisons with normal control subjects can be drawn.

Conclusions
The present study addressed the methodological question whether a CNV could be
elicited in autistic subjects. It also addressed three clinical conceptual questions based on the

feasibility of the ﬁrst methodological issue.

104

Research Question 1: Is it gossible to elicit CNVs in autistic children?

As a CNV was established in one autistic subject on the Push Button Task, which was
designed to address this question, it was concluded that autistic subjects can establish a CNV.
Since an autistic subject was able to establish a CNV in the present study, one might assume
that the paradigm can be used to test other populations such as language-impaired children.
In this study, the CNV proved to be an adequate tool to investigate cognitive processing in
terms of attention in the autistic population. Thus, the CNV paradigm might render itself to

investigate research questions addressing cognitive processing.

Research Question 2: Is it possible to elicit a CNV when the autistic subject is manually
glided to do an event?

The study found that a CNV indicative of attention was established under the guided
condition by Subject 2, while Subjects 1 and 3 established large NPS approaching the
dimensions as deﬁned for a CNV. This ﬁnding will contradict the assumption that the
information in the guided condition does not reach the brain.

As described earlier, the normal pilot subject achieved lower values on the tasks in
the guided condition than did the autistic subjects. The autistic subjects responded to the
novelty effect by exhibiting higher values on the novel task in the guided condition. This
effect was not found in the normal pilot subject. The normal pilot subject achieved higher
values when performing the familiar task, presumably because it was the ﬁrst task in the

guided condition. The novelty effect in the second guided task might have been too weak.

105

Research Question 3: Is there a difference between the CNV elicited when the autistic
subject is manually guidedﬁand unguided to do an event?

Possible CNV differences between the guided and unguided conditions were
investigated. An equal number of CNVs was observed in both conditions. Affolter and
colleagues’ claim that the information gets to the brain in both conditions was partly
supported by the fact that no signiﬁcant differences were found between the guided and
unguided conditions in the autistic subjects.

However, there appears to be one difference. For the guided input to be beneﬁcial,
some aspects of the task may have to be challenging for the guided individual. Once the
nervous system has learned the task, more energy may be invested in independent planning,
executing and monitoring of the task. This assumption appears to be supported by the fact
that the most severe of the autistic subjects in the present study established CNVs under both
guided conditions, whereas the somewhat higher functioning subjects were either more
inclined to establish a CNV in the unguided condition or established a large NPS that was
similar to a CNV in conﬁguration but lacked the deﬁned dimensions.

The fact that the most severe autistic subject established a CNV when he was guided
to perform the tasks yields some support for the claim made by Affolter and colleagues that
the tactile-kinesthetic input is an important additional modality for gaining the child’s
attention for learning. One might assume that this subject’s nervous system responded to the
guided input best because his tactual-kinesthetic system was not yet sufﬁciently developed,
and he, therefore, needed the input. Consequently, the T-K input was most beneﬁcial at this
point in his cognitive deve10pment.

It appears that the effect of guiding on attention will depend on the severity of the

impairment of a subject and on the complexity of the task. It is reasonable to assume that

106

even a mature unimpaired nervous system might respond well to partial guiding, provided the
individual is not able to perform a certain aspect of the task independently, for instance,

drawing some facial features.

Research Question 4: Is there a difference between the CNV elicited from autistic
subjects when the event stimuli vary in the amount of resistance?

Nervous system changes in response to stimuli varying in the amount of resistance
were examined to test the assumption made by Affolter and colleagues that the brain codes
degrees of resistance. However, no signiﬁcant differences in measures of brain activity were
found when the two experimental conditions (Squeezing a Soft Ball — Unguided versus
Squeezing a Hard Ball - Unguided) were compared on the four dependent variables.

Nevertheless, a comparison of the data of the normal pilot subject on Task 2
(Squeezing a Soft Ball -- Unguided) and Task 3 (Squeezing a Hard Ball -- Unguided)
revealed higher values for the task offering weak resistance. The normal pilot subject exerted
more energy when the soft ball was squeezed as compared to the hard ball.

The results of the autistic subjects can be partially explained by the fact that they
wanted to achieve the same effect when they squeezed the hard ball as they had experienced
when they squeezed the soft ball in the previous task. On ﬁve of the six trials, the autistic
subjects enlisted the other hand when they felt the hard ball to help them squeeze harder to
achieve the same effect as before. The confrontation with this new situation increased the
subjects' attention. It also increased the autistic subjects’ muscular effort which led to greater
negativities when the hard ball was squeezed. One might conclude that the nervous system of

the autistic subjects was still in the process of coding the various degrees of resistance,

107

whereas the normal pilot subject’s nervous system had already coded for the varying degrees

of resistance.

Summary

The present study was motivated by the difﬁculties arising from treating children with
severe impairments along the PDD/autistic spectrum since this population does not respond
to conventional treatment paradigms with the success one would expect. According to the
theory stated by Affolter and coworkers, one might assume that conventional treatment
paradigms require children to use their visual and auditory input at a supramodal level. The
prerequisite for this level is that the children have gained a multitude of experiences
following nonverbal explorations.

The representation of events at the nonverbal level is also a prerequisite for
meaningful language to occur. To be functioning at this relatively high level, a child must
have experienced a fairly normal development in infancy and early childhood. In other
words, the child must have been able to explore his environment by using his tactile-
kinesthetic input as a primary source of information to pave the way for intermodal
integration and serial organization of perceptual input.

On the contrary, children with severe impairments might have problems either with
the extraction of adequate T-K information from their environment, or with the integration of
perceptual input from various modalities (vision and audition might not get aligned
adequately with other senses), or with organizing perceptual information serially (events
might not get sequenced appropriately). Assuming the hypotheses made by Affolter and
colleagues are correct, it stands to reason to intervene at exactly the level where the deﬁcits

occurred. For most severely impaired children, this would mean intervention at a nonverbal

108

level to allow the child to solve functional problems, explore his environment, and represent
events mapped onto his experiences.

If Affolter and colleagues’ hypotheses are correct, one should be able to observe
attention while severely impaired children are performing a problem-solving task at the
nonverbal level. To measure brain activity electrophysiologically, the CNV paradigm was
adopted for the present investigation. One methodological research question and three
clinical conceptual research questions were addressed. The subjects performed 6
experimental tasks. The recorded data was pooled across two trials for two subjects and
across one of the six tasks for a third subject. This amounted to a maximum of ﬁve to six
data points to analyze each experimental variable. The Wilcoxon matched pairs signed-ranks
test was used to test for signiﬁcant differences. The results were as follows with regard to the
individual research questions.

The ﬁrst research question was designed to determine whether the CNV paradigm
could be used to evaluate cognitive processing in autistically impaired children. The present
study showed that autistically impaired children were able to establish a CNV or NPS that
were below the designated criteria set for this study. Three subsequent research questions
investigated clinical conceptual issues.

The second research question investigated whether a CNV could be observed in the
guided condition. The present study demonstrated that a CNV was established in the guided
condition. Hence, it was shown that the subjects received the perceptual input in this
condition. The beneﬁt of guiding appeared to depend on the subject’s developmental stage
and on the novelty and complexity of the task.

The fact that the most severe of the autistic subjects established 2 CW3 in the guided

condition demonstrated that this subject beneﬁted the most from the guided input. This

109

provided indirect evidence for the assumption made by Affolter and coworkers that the T-K
input provided evidence that tactual input from guiding adds critical information capable of
getting the subject’s attention and anchors the other senses.

The third research question investigated possible CNV differences between the
guided and unguided conditions. When task performances were compared in the guided and
unguided conditions, no signiﬁcant differences were found for the novel task. However, two
signiﬁcant differences were found for the duration and area of the CNV in the familiar task
suggesting that more mental energy was exerted in the unguided condition as compared to
the guided condition.

The autistic subjects demonstrated signiﬁcantly more cognitive activity in the novel
task as compared to the familiar task suggesting that the novelty effect was very pronounced
in the autistic subjects as a group.

The fourth research question investigated the role of resistance. When the event
stimuli varied in the amount of resistance, the subsequent changes in brain activity were not
signiﬁcantly different in the two conditions (weak resistance as compared to strong
resistance). Some subject biases were observed. In most trials, the autistic subjects enlisted
the other hand to achieve the same effect as they had experienced in the prior task. The
normal pilot subject, on the other hand, developed a CNV when he squeezed the soft ball but
not when be squeezed the hard ball suggesting that he had to search harder for information in

the condition offering weak resistance.

110

APPENDIX A

Copy of Diagnostic and Statistical Manual of Mental Disorders:
DSM-IV (Fourth Edition). American Psychiatric
Association, Washington, D.C.

 

 

i. Diagnostic criteria for 299.00 Autistic Disorder

A. A toralofsix(or more) items from(l).(2). and (3). with at laSt two
from (1). and one cadt from (2) and (3):
(1) qualiarive impairment in social interaction, as manifested by at
least two of the following: .
(a) marked impairment in the use of multiple nonverbal behaviors
such as eye-to-eye gaze, facial expression, body postures, and
gestures to regulate social interaction
(b) failure to develop peer relationships appropriate to develoo
mental level
(c) a ladt of spontaneous seeking to share enjoyment, intereas,
or adtievements with Other people (cg. by a ladt of showing,
bringing, or pointing out obiecrs of intereSt)
(d) lack of social or emotional reciprocity
(2) qualitative impairments in communication as manifested by at leasr
one of the following:

(a) delay in, or road lack of, the development of spoken language
(not accompanied by an attempt to compensate through
alternative modes of communication such as gesture or mime)

Cb) in individuals with adequate speech, marked impairment in
the ability to initiate or win a conversation with others

(c) StereOtyped and repetitive use of language or idiosyncratic
language

(d) lack of varied, spontaneous make-believe play or social
imitative play appropriate to developmental level

(continued)

 

 

 

____d

 

 

Cl Diagnostic criteria for 299.00 Autistic Disorder

(3) reached repetitive and stereotyped patterns of behavior, interests,
and activities, as nunifested by at last one of the following:

(a) encompassing preoccupation with one or more srereoryped ‘
and restricted patterns or interesr that is abnormal either in
intensity or focus

(b) apparently inflexible adherence to Specific. nonfunctional
routines or rituals

(c) stereotyped and repetitive moror mannerisms (e.g.. hand or
finger ﬂapping or twiSting. or complex whole-body move-
ments)

(d) persiaent preoccupation with parts of objeas

B. Delays or abnormal funcrioning in at leasr one of the following areas,
with onset prior to age 3 years: (1) social interacrion, (2) language as
used in social communication, or (3) symbolic or imaginative play.

C. The di$turbance is not better accounted for by Rett's Disorder or
Childhood Disintegrative Disorder.

 

 

 

 

 

 

Diagnostic and Statistical Manual of Mental Disorders: DSM-IV (Fourth Edition).
American Psychiatric Association, Washington, D.C.

lll

APPENDIX B

Copy of
Human Subjects’ (UCRIHS) Approval
and Consent Form

 

OFFICE OF
RESEARCH
AND
GRADUATE
STUDIES

University Committee on
Research Involving
liuman Subject:
(UCRIHS)

Michigan State University
246 Administration Buildi
East Lansing. Michigan
48824-1046

517/355-2180
FAX' 517/353-2976

{he Michigan State University
IDfA rs mstiiuhona! Diversity:
Excellence in Action

MSU is an allirmarneaction.
emuimoolrmirv instiluiim

  

STATE

SITY

MICHIGAN
u N I v E R

June 28, 1999

TO: Dr. Ida Stockman
326 Northlawn
East Lansing. MI 48823

RE: IRB #98395 CATEGORYZZ-C
RENEWAL APPROVAL DATE:June 28, 1999

TITLE: ELECTROENCEPHALOGRAPHIC (EEG) ANALYSIS DURING A GUIDED
INTERACTION TASK VERSUS UNGUIDED CONDITION IN NORMAL AND

AUTISTIC CHILDREN

The University Committee on Research Involving Human Subjects' (UCRIHS) review of
this project is complete and I am pleased to advise that the rights and welfare of the
human subjects appear to be adequately protected and methods to obtain informed
consent are appropriate. Therefore. the UCRIHS APPROVED THIS PROJECTS

RENEWAL.
This letter approves the change in location for the testing & the addition of three adult

subjects.

RENEWALS: UCRIHS approval is valid for one calendar year, beginning with the approval
date shown above. Projects continuing beyond one year must be renewed with the green
renewal form. A maximum of four such expedited renewal are possible. Investigators wishing
to continue a project beyond that time need to submit it again for complete review.‘
REVISIONS: UCRIHS must review any changes in procedures involving human subjects, prior
to initiation of the change. If this is done at the time of renewal. please use the green renewal
form. To revise an approved protocol at any other time during the year. send your written
request to the UCRIHS Chair, requesting revised approval and referencing the project's IRB#
and title. Include in your request a description of the change and any revised instruments.
consent forms or advertisements that are applicable.

PROBLEMS/CHANGES: Should either of the following arise during the course of the work.
notify UCRIHS promptly: 1) problems (unexpected side effects. complaints. etc.) involving
human subjects or 2) changes in the research environment or new information indicating
greater risk to the human subjects than existed when the protocol was previously reviewed and

approved.

If we can be of further assistance. please contact us at 517 355-2180 or via email:

UCRIHS@piIot.msu.edu.

ely,

     
  

avid E. Wright, Ph. D.
CRIHS Chair

DEW: ab
cc: Christel Berretta

All UCRIHS forms are located on the web: http://www.msu.edulunItlvprgsIUCRIHS!

112

Consent Form

 

You are being asked to participate in a research study that is based
on Felicie Affolter's theory that input via the sense of touch is
essential for exploring one's environment and hence at the root of
learning. Vision and audition are secondary to the perception of
touch and need to be integrated intermodally. Children with severe
impairments along the pervasive developmental disorder
(PDD)/ autistic spectrum fail to explore their environment
successfully on their own. Consequently, these children start at a
task, but fail at its completion. Affolter believes that this group of
children beneﬁts from a therapy where the children are guided
manually through functional problem-solving tasks with the
structured help of a therapist. The more the children understand the
goal of a certain task, the more they will take over and eventually
will perform the task independently. The question is whether the
brain registers the sensory input under the guided condition and
whether the recorded brain activity can be identiﬁed as
characteristic of attention. Positive ﬁndings in this research study
might change the way we work with this group of children. We
might come to a more functional, problem-solving approach that is
very much "hands on".

The goal of this research is to explore the question whether there is ‘
comparable sensory input registered by the brain in guided and
unguided conditions. Furthermore, it will be examined whether
there is brain activity indicative of attention under the guided
condition. The role of resistance to human perception will also be
examined. The ﬁrst set of two experiments involves squeezing a
hard ball (offering a lot of resistance) and a soft ball (offering little
resistance) to investigate the importance of resistance to human
perception. To offer an environment, where the impaired child ﬁnds
a lot of resistance, is a prerequisite for Affolter's treatment

113

paradigm. While hard and soft balls are being squeezed, brain
activity is being registered. A comparison between the two kinds of
resistance will allow us to draw some conclusions about this basic
assumption in Affolter's theory.

The second set of experiments deals with t0pological changes. The
subject will " move apart" (move apart = one t0pological change) 10
pairs of wooden blocks. The subject will also "move together"
(move together = one topological change) 10 pairs of blocks. These
tasks will be carried out in guided and unguided conditions.

Your son will be tested three times. Language testing at the
beginning of the study will take about 30 minutes and take place in
your home. There will be one visit to the neurophysiology/ audiology
lab at MSU where the experiments will be carried out. Each visit will
last for 30 minutes.

For the duration of these 30 minutes, three electrodes will be pasted
on the subject's scalp. This is not painful and does not cause any
discomfort. The subject's participation is voluntary. If the subject
refuses to comply with the experimental procedures, he may
discontinue his participation in the experiments without any penalty
and at any time. The subject will still be entitled to the honorarium
of 25.00 dollars although he discontinued experimental procedures.

All results will be treated with strict conﬁdence and the subjects will
remain anonymous in any report of research ﬁndings. On request
and within these restrictions results may be made available to the
subjects and their parents or legal guardians.

An honorarium of $ 25.00 will be given to each subject on each visit

to the neurophysiological/audiological lab in the Communication
Arts and Sciences building on campus of MSU.

114

At the beginning of the study, three undergraduate students will be
tested to ensure that the computer and BBC apparatus are working
ﬂawlessly. These students will not receive a honorarium for their
services.

If parents, legal guardians, or subjects have any questions or
concerns, they are welcome to contact the investigator, Christel
Beretta, (telephone number 349-4722) at any time.

You understand that your participation in this research project will
not involve any additional costs to you or your health care insurer.

You give permission for your son to
participate in the above-stated research study.

 

Name of parent or guardian:

Name of undergraduate student:

 

UCRIHS APPFtQVAL FOR
THIS project EXPIRES:

JUN 2 8 2000

SUBMIT RENEWALAPPUCATION
ONE MONTH PRIOR To
ABOVE DATE TO CONTINUE

115

LIST OF REFERENCES

LIST OF REFERENCES

Affolter, F. (1987). WahrnehmungLWirklichkeit und Sprache. Villingen-Schwenningen:
Neckar-Verlag GmbH.
Affolter, F ., & Stricker, E. (Eds) (1980). Perceptual processes as prerequisites for complex

human behavior. Bern: Huber.

 

Affolter, F., & Bischofberger, W. (1993). Die Organisation der Wahmehmung, Aspekte der
Entwicklung und des Abbaus. In F. Affolter & W. Bischofberger (Eds.), Wenn die
Organisation des ZN S zerfallt, und es an gespiirter Information mangelt [When the
organization of the central nervous system breaks down and tactual input is
insufficient] (pp. 24-55). Villingen/Schwenningen: Neckar-Verlag.

Affolter, F., & Bischofberger, W. (1996a, April). From action to interaction as the root of
development: A clinical perspective. Invited lecture presented at the Annual Jane &
Herbert Oyer Lecture, Michigan State University, U.S.A.

Affolter, F., & Bischofberger, W. (1996b). Gespiirte Interaction im Alltag. [Tactual
interaction in daily life]. In B. Lipp & W. Schléigel (Eds.), Mge von Anfanw (pp.
77-99). Villingen/Schwenningen: Neckar Verlag.

Affolter, F ., & Bischofberger, W. (In Press). Nonverbal information processing in children
with language disorders: Towards a new frameworLfor clinical intervention.

Mahwah, NJ: Erlbaum.

116

Affolter, F., Bischofberger, W., & Calabretti-Erni, V. (1996, September). Nonverbal
interaction in babies and brain-damaged patients. Poster session presented at the
symposium: The growing mind--La pensée en evolution, Geneva, Switzerland.

Ayres, J. (1972). Types of sensory integrative dysﬁmction among disabled learners. ”I_'h_g
American Journal of Occupational Therapy, 26, 13-18.

Baron-Cohen, S. (1991). Do people with autism understand what causes emotion? Child
Development, 62, 385-395.

Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge,
MA: MIT Press.

Bloom, L. (1996, March). The integration of exgession into the stream of evegyday activity.
Lecture presented at a symposium: Movement and action: Links to Intelligent
Behavior, Michigan State University, East Lansing, MI.

Bonda, E., Petrides, M, & Evans, A. (1996). Neural systems for tactual memories. Journal of
Neurophysiology, 75, 1730-1735.

Booth, A. (1998). The role of functional information in the development of concepts of
obiects in infancy. Unpublished doctoral dissertation, University of Pittsburgh,
Pennsylvania.

Brothers, L., & Ring, B. (1992). A neuroethological framework for the representation of
minds. Journal of Cognitive Neuroscience, 4, 107-118.

Burgess, A. P. & Gruzelier, J. H. (1997). Short duration synchronization of human theta
rhythm during recognition memory. NeuroReport, 8, 1039-1042.

Cantor, D. S., Thatcher, R. W., Hrybyk, M. & Kaye, H. (1986). Computerized EEG analyses

of autistic children. Journal of Autism and Developmental Disorders, 16, 169-186.

117

Carr, T. H., Bernard, R. A., Goran, D. A., Nordell, B. A., Cooper, T. G., Meyer, R. A.,
Egger, T. G., Backus, M. A., Putnam, C. M., Paonessa, P. D., & Potchen, E. G.
(April, 1996). Somatosensory activity during simple motor taik. Poster session
presented at the annual meeting of the Cognitive Neuroscience Society, San
Francisco, CA.

Carvell, G. E., & Simons, D. J. (1996). Abnormal tactile experience early in life disrupts
active touch. Journal of Neuroscience, 16 (8), 2750-2757.

Celesia, C.G., & Brigell, M. (1993). Auditory evoked potentials. In E. Niedermeyer & F.
Lopes da Silva (Eds.), Electroencephalography: Basic principlesglinical
applications, and related ﬁelds (pp. 937—956). Baltimore: Williams & Wilkins.

Cohen, J. (1969). Very slow brain potentials relating to expectancy: The CNV in average
evoked potentials. In E. Donchin & D. Lindsley (Eds.), Average evoked potentials:
Methods, result and evaluations (pp. 143-198). Washington, DC: National
Aeronautics and Space Administration, SP191.

Cohen, J ., & Ofﬁier, F. (1967). Some clinical applications of the CNV and evoked potentials.
Electroencephalography and Clinical Neurophysiology, 22, 190.

Cook, E. H. (1990). Autism: Review of neurochemical investigation. Sygapse, 6, 292-308.

Dawson, G., Grofer Klinger, L., Panagiotides, H., Lewy, A. & Castelloe, P. (1995).
Subgroups of autistic children based on social behavior display distinct patterns of
brain activity. J oumal of Abnormal Child Psycholmlzl, 569-5 82.

Donald, M. W. (1970). Direct-current potentials in the human brain evoked during time
cognitive performance. Nature, 227, 1057-1058.

Donchin, E., Gerbrandt, L.A., Leifer, L., & Tucker, L. (1972). Is the contingent negative

variation contingent on a motor response? Psychophysiology, 9, 178-188.

118

Dujardin, K., Bourriez, J. L., & Guieu, J. D. (1995). Event-related desynchronization (ERD)
patterns during memory processes: Effects of aging and task difﬁculty.
Electroencephalpgraphy and Clinical Neurophysiology 96, 169-182.

Fenelon, B. (1984). Effects of stimulus coding on the contingent negative variation recorded
in an information processing task. In R. Karrer, J. Cohen, & P. Tueting (Eds.), 13m
and information. Event-related potentials (pp. 194-198). New York: New York
Academy of Sciences.

Fisher, A,. Murray, E., & Bundi, A. (1991). Sensory integration: Theory anQaractice.
Philadelphia, PA: PA. Davis.

Frith, U., & Herrnelin, B. (1969). The role of visual amd motor cues for normal, sub-normal
and autistic children. Journal of Child PsychologL and Psychiatry, 10, 153-163.

Frith, U., & Baron-Cohen, S. (1987). Perception in autistic children. In D. J. Cohen & A. M.
Donnellan (Eds.), Handbook of autism and pervasive developmental disorders (pp.
85-102). Silver Spring, MD: V. H. Winston & Sons.

Gillberg, C., & Coleman, M. (1992). The biology of the autistic syndromes (2nd_ed.). New
York: Cambridge University Press.

Gloning, K., Burian, K., Gestring, G. F., & Haider, M. (1970). Language evoked response
audiometry (LERA) in aphasia. Electroencephalography and Clinical
Neurophysiolggy,;8_, 215.

Gorbachevskaya, N. L., Yakupova, L. P., Kozhushko, L. F., Bashina, V. M., Sirnashkova, N.
V., & Amelin, S. N. (1992). Topographic EEG mapping in child psychiatry. m
Physiolyogy, 18, 404-410.

Grandin, T. (1996). Thinking in pictures. And other rgmrts from my life with autism. New

York: Vintage Books, Division of Random House, Inc.

119

Griinewald, G., Griinewald- Zuberbier, E., Netz, J., Homberg, V. , & Sander, G. (1979).
Relationships between the late component of the contingent negative variation and
the Bereitschaftspotential. Electroencephalography and Clinical Neurophysiology, 46,
538-545.

Harmony, T., Fernandez T., Silva, J ., Bemal, J ., Diaz-Comas, L., Reyes, A., Marosi, E.,
Rodriguez, M. & Rodriguez M. (1996). EEG delta activity: An indicator of attention
to internal processing during performance of mental tasks. International Journal of
ILSYchgphysiology, 24, 161-171.

Ikeda, A., Shibasaki, H., Kaji, R., Terada, K., Nagamine, T., Honda, M., & Kimura, J.
(1997). Dissociation between contingent-negative variation (CNV) and
Bereitschaﬁspotential (BP) in patients with parkinsonism. Elecetroencephalography
and Clinical NeurophysioLogy, 10; 142-151.

Imashioya, H., Dollins, A. B., & Kakigi, S. (1987). Motor response information inﬂuence on
CNV shape and resolution time. The Pavlovian Joumil of Biolggical Science, 22, 1-
6.

Johnston, J. (1988). Speciﬁc language disorders in the child. In N. Lass, L. McReynolds, J.
Northern, & D. Yoder (Eds.), Handbogc of speech-lggpage patholﬁy & audiology
(pp. 685-715). Philadelphia: B. C. Decker.

Kemmer, C., Verbaten, M. M., Cuperus, J-M., Camfferman, G., & Van Engeland, H. (1994).
Visual and somatosensory event-related brain potentials in autistic children and three
different control groups. Electroencephalography and Clinical Neurophysiology,_9_g,

225-237.

120

Klatzky, R., Lederman, S., & Reed, J. (1987). There's more to touch than meets the eye: The
salience of object attributes for haptics with and without vision. Journal of
Experimental Psychology. 116, 356-369.

Klimesch, W. (1997). EEG-alpha rhythms and memory processes. International Journal of
Psychophysiology,_26, 3 19-340.

Klimesch, W., Doppelmayr, M., Russegger, H., & Pachinger, T. (1996). Theta band power in
the human scalp EEG and the encoding of new information. NeuroReport, 7, 123 5-
1240.

Klimesch, W., Doppelmayr, M., Schimke, H., & Ripper, B. (1997). Theta synchronization
and alpha desynchronization in a memory task. Psychophysiology, 34, 169-176.

Klimesch, W., Schimke, H., Doppelmayr, M., Ripper, B., Schwaiger, J. & Pfurtscheller, G.
(1996). Event-related desynchronization during encoding predicts later recall
performance? International Journal of Psychophysiology, 24, 47-60.

Krumholz, A. (1993). Evoked potentials in infancy and childhood. In E. Niedermeyer & F.
Lopes da Silva (Eds.), Electroencephalograghy: Basic principlechlinical application,
and related ﬁelds (pp. 975-988). Baltimore: Williams & Wilkins.

Lai, C., Ikeda, A., Terada, K., Nagamine, T., Honda, M., Xu, X., Yoshimura, N., Howng, S.,
Barrett, G., & Shibasaki, H. (1997). Event-related potentials associated with
judgment: Comparison of SI- and Sz-choice conditions in a contingent negative
variation (CNV) paradigm. Journal of Clinical Neurophysiolgy, 14, 394-405.

Langer, J. (1994). From acting to understanding: The comparative development of meaning.
In W. R. Overton & D. S. Patermo (Eds.), The ontogenesis of meaning (pp. 191-213).

Hillsdale NJ: Lawrence Erlbaum.

121

Lopes da Silva, F. (1993). Event-related potentials: Methodology and quantiﬁcation. In E.
Niedermeyer & F. Lopes da Silva (Eds.), Electroencephalography: Basic principles,
clinical application, and related ﬁelds (pp. 877-886). Baltimore: Williams & Wilkins.

Lopes da Silva, F ., Van Rotterdam, A., Storm van Leeuwen, W., & Tielen, A. M. (1970).
Dynamic characteristics of visual evoked potentials in the dog. H. Beta frequency
selectivity in evoked potentials and background activity. Electroencephalography and
Clinical Neurophysiology 29, 260-268.

Low, M. D., Borda, R. P., Frost, S. D., & Kellaway, O. (1966). Surface-negative, slow
potential shift associated with conditioning in man. Neurolagy, 16, 771-782.

Low, M. D., Coats, A. C., Rettig, G. M. & McSherry, J. W. (1967). Anxiety, attentiveness-
alertness: A phenomenological study of the CNV. NeuropsychologLLS, 379-3 84.

McAdam, D. W. (1969). Increases in CNS excitability during negative cortical slow
potentials in man. Electroencephalography and Clinical Neurophysiology,l§, 216-
219.

McCallum, W. C. (1988). Potentials related to expectancy, preparation, and motor activity. In
T. W. Picton (Ed.), Handbo_ok of electroencepha_lo_graphv and clinical

neurophysiology (pp. 427-534). Amsterdam: Elsevier.

 

McPherson, W. B., Ackerman, P. T., Holcomb, P. J., & Dykman, R. A. (1998). Event-related
brain potentials elicited during phonological processing differentiated subgroups of
reading disabled adolescents. Brain and Langgge, 62, 163-185.

Nakamura, M., Fukui, Y., Kadobayashi, I., & Kato, N. (1979). A comparison of the CNV in
young and old subjects. Its relation to memory and personality.

Electroencephalogaphy and Clinical Neurophysiology, 46, 337-344.

 

122

Nativ, A. (1991). Brain potentials associated with movement in traumatic brain injury.
Physical Therapy. 71, 48-59.

Nelson, K. I. (1986). Event knowledgg: Structure and function in development. Hillsdale, NJ:
Lawrence Erlbaum.

Panksepp, J. (1992). Oxytocin effects on emotional processes: Separation distress, social
bonding, and relationships to psychiatric disorders. In C. A. Pederson, J. D. Caldwell,
G. F. J irikowski, & T. R. Insel (Eds.), Oxytocin in maternal, sexual, and social
behaviors. New York: New York Academy of Sciences.

Peters, J. F ., Billinger, T. W., & Knott, J. R. (1977). Event-related potentials of the brain
(CNV and P300) in a paired associate learning paradigm. Psychophysiology, 14, 579-
585.

Peters, J. F ., Knott, J. R., Miller, L. H., VanVeen, W. J ., & Cohen, S. I. (1970). Response
variables and magnitude of the contingent negative variation. Electroencephalggrapiry
and Clinical Neurophysiolqu, 29, 608-611.

Pfurtscheller, G., Stancak, A., & Edlinger, G. (1997). On the existence of different types of
central beta rhythms below 30 Hz. Electoencephalography and Clinical
NemopthioloM 316-325.

Piaget, J. (1977). Egocentric thought in the child. In H. E. Gruber & J. J. Voneche (Eds.),

 

The essential Piaggt (pp. 119-153). New York: Basic Books, Inc.

Prevec, T. S. & Ribaric, K. (1986). Contingent negative variation in children. In Gallai, V.
(Ed.), Maturation of the CNS and evoked potentials (pp. 269-278). Amsterdam:
Elsevier Science Publishers.

Prevec, T. S., & Beric, A. (1991). Measurement of light touch perception threshold by

contingent negative variation. Experimental Brain Research, 84, 643-648.

123

Purves, W. K., Orians, G. H. & Heller, H. C. (1995). Life: The science of biology (pp. 777-
778). Salt Lake City, UT: W. H. Freeman.

Rebert, C. S., & Knott, J. R. (1970). The vertex non-speciﬁc evoked potentials and latency of
contingent negative variation. Electroencephalography and Clinical Neurophysiology,
28, 561-565.

Reneau, J. P., & Hnatiow, G. Z. (Eds). (1975). Evoked response audiometrg Baltimore:
University Park Press.

Rugg, M. D., & Barrett, S. E. (1987). Event-related potentials and the interaction between

 

orthographic and phonological information in a rhyme-judgment task. Brain and
Language,32, 336-361.
Rugg, M. D., Cowan, C. P., Nagy, M. E., Milner, A. D., Jacobson, 1., & Brooks, D. N.

(1989). CNV abnormalities following closed head injury. Brain 112 489-506.

 

Rutter, M. & Schopler, E. (1987). Autism and pervasive developmental disorders: Concepts
and diagnostic issues. Journal of Autism and Developmental Disorders,l7, 159-186.

Rugg, M. D., & Dickens, A. M. J. (1982). Dissociation of alpha and theta activity as a
function of verbal and visuospatial tasks. Electroencephalography and Clinical
Neurophysiology,_5§, 201-207.

Saper, C. B. (1987). Diffuse cortical projection systems: Anatomical organization and role in
cortical ﬁinction. In V. B Mountcastle & F. Plum (Eds.), ﬂandbookhof Physiology;
(pp. 169-210). Bethesda, MD: American Physiological Society.

Schneble, H. (1984). EEG-Befunde bei Patienten mit autistischem Syndrom. [EEG ﬁndings

in autistic patients]. Zeitung ﬁir Kinder-Juﬂthychiatrie, 12, 358-378.

124

Scholl, H. M. (1981). Language disorders related to learning disabilities. In J. Gottlieb & S.
Strichart (Eds.), Developmental theoryahnd research in learning disabilities (pp. 130-

168). Baltimore: University Park Press.

 

Sheer, D. (1984). Focused arousal, 40 Hz, and dysfunction. In T. Ebert (Ed.), Selfregulation

of the brain and behavior (pp. 64-84). Berlin: Springer.

Singh, V. K. (1996). Short communication: Plasma increase of interleukin-12 and interferon-

gamma. Pathologic signiﬁcance in autism. Journal of Neuroimmunology, 66,143-145.

Singh, V. K., Lin, S. X., & Yang, V. C. (1998). Serological association of measles virus and
human herpesvirus-6 with brain autoantibodies in autism. Clinical Immunology and
Irnmunopathology, 89, 105-108.

Singh, V. K., Warren, R., & Ghaziuddin, M. (1997). Brief communication: Circulating
autoantibodies to neuronal and glial ﬁlament proteins in autism. Pediatric Neurology,
L6, 88-90.

Stark R., & Tallal, P. (1988). Language, speech and reading disorders in children.
Neuropsgzhological Studies, 66, 143-145.

Steriade, M. (1993). Cellular substrates of brain rhythms. In E. Niedermeyer & F. Lopes da
Silva (Eds.), Electroencephalography: Basic principles, clinical applicationshand
related ﬁelds (pp. 27-62). Baltimore: Williams & Wilkins.

Stockman, I. (In Press). Epilogue: Language learning and nonverbal interaction in daily
events. In F. Affolter & W. Bischofberger (Eds.). Nonverbal information processing
in children with language disorders: Towards a new framework for clinical
intervention. Mahwah, NJ: Erlbaum.

Stone, V. E., Baron-Cohen, S., & Knight, R. T. (1998). Frontal lobe contributions to theory

of mind. Journal of Cognitive Neuroscience; 105, 640-656.

125

 

 

Swisher, L., Plante, E., & Lowell, S. (1994). Nonlinguistic deﬁcits of children with language
disorders complicate the interpretation of their nonverbal IQ scores. Langgge,
@eech and Hearihg Services in Schools, 25, 235-240.

Tecce, J. J. & Cattanach, L. (1993). Contingent negative variation. In E. Niedermeyer & F.
Lopes da Silva (Eds.), Electroencephmggaphy:hasic principles, clinical application,

and related ﬁelds (pp. 887-910). Baltimore: Williams & Wilkins.

 

Tecce, J. J ., Cattanach, L., Yrchik, D. A., Meinbresse, D., & Dessonville, C. L. (1982). CNV
rebound and aging. Electroencephalography and Clinical Neurophysiology, 54, 175-
186.

Tecce, J. J ., Savignano-Baowman, J ., & Kahle, J. B. (1978). Effects of visual distraction on
contingent negative variation and type A and B CNV shapes. In D. A. Otto (Ed).
Multidischﬂinary perspectives in event—related potential research (pp. 358-363).
Research Triangle Park, NC: US. Environmental Protection Agency.

Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50, 79-95.

Timsit-Berthier, M., Delaunoy, J ., Koninckx, N., & Rosseau, J. C. (1973). Slow potential
changes in psychiatry. I: Contingent negative variation. Electroencephalography and
Clinical Neurophysiology, 35, 355-361.

Wallace, M. T. & Stein, B. E. (1997). Development of multisensory neurons and
multisensory integration in cat superior colliculus. Journal of Neuroscience, 17, 2429-
2444.

Walter, W. G. (1964). Slow potential waves in the human brain associated with expectancy,
attention, and decision. Archiv ﬁir Psychiatric Nervenkrankheiten, 206, 309-322.

Walter, W. G. (1965). Effects on anterior brain responses of an expected association between

stimuli. Journal of Psychosomatic Research, 9, 45-49.

126

Walter, W. G. (1966). Electrophysiologic contributions to psychiatric therapy. C_u_rg1_t
Psychiatric Therapies, 6, 13-25.

Warren, R. P., & Singh, V. K. (1996). Elevated serotonin levels in autism: Association with
the major histocompatibility complex. Neuropsychobiology, 34, 72-75.

Warren, R. P., Singh, V. K., Averett, R. E., Odell, J. D., Maciulis, A., Burger, R. A., Daniels,
W. W., & Warren, W. L. (1996). Irnmunogenetic studies in autism and related
disorders. Molecular and Chemical Neuropathology, 2& 77-81.

Washer, E., Verleger, R., J askowski, P., & Wauschkuhn, B. (1996). Preparation for action.
An ERP study about two tasks provoking variability in response speed.
Psychophysiology, 33, 262-272.

Waterhouse, L., Fein, D., & Modahl, C. (1996). Neurofunctional mechanisms in autism.
Psychological Review, 103, 457-489.

Weiss, T., Sust, M., Beyer, L., Hansen, E., Rost, R., & Schmalz, T. (1995). Theta power
decreases in preparation for voluntary isometric contractions performed with maximal
subjective effort. Neuroscience Letter, 193, 153-156.

Zeki, S. (1993). A vision of the brain. Oxford: Blackwell.

127

"IIllllllllllllll’lllI“