. K. .:.n Egnncfipméu.

.fi.
5 x. ..
a. . 3.31:3". 3,?

a. nah
. is

.x: in»

7.
.IV. .3.

 

 

:1;
z , 7... .
;.:. : .3. . 1....” x
S, :2:

u...
.9... e

 

 

 

 

‘
‘ 23:1:
.3 :. vi...»

x ‘ vv}
:3? .2»; 5‘ l

 

 

 

V,

..\ 4..
1 {xiilfib .- . 1...
z r; If

 

.fi: 3'. .1 t: .7
.r Hair.
2. .13.

 

”a; IIIIHIHHNIll!lllllllllllllllllllllllllllllllll

812 6999

This is to certify that the

dissertation entitled

Differential Outcomes of Functional and
Neuropsychological Approaches to Naming
Treatment Among Adults with Chronic

Nonfluent Aphasia
presented by

Jacqueline Jay Hinckley

has been accepted towards fulfillment
of the requirements for

__P_h_..D_._— degree in Audieleg—y— and
Speech Sciences

 

Major professor

(3A0? P. J)“ me

Date (43 (8 ‘ Ciq

MSU is an Aflirman’w Action/[q ual Opportunity Institution 0-12771

 

 

 

 

 

LIBRARY
Michigan State
University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECAUJED with earlier due date if requested.

 

DATE DUE

DATE DUE '

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1198 animus-p.14

DIFFERENTIAL OUTCOMES OF FUNCTIONAL AND NEUROPSYCHOLOGICAL
APPROACHES TO NAMING TREATMENT AMONG ADULTS WITH CHRONIC
NONFLUENT APHASIA

BY

Jacqueline Jay Hinckley

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Audiology and Speech Sciences

1999

RON—nun“...-
t- u- .- ... ..
do. 0 gnu-“I ‘

‘BRAfK‘ Av‘
Ar’J-t :- H
.0. s 5\V¢¢\a:-

‘w

WI 079216
I»- i

$8.5, a,
‘I

ABSTRACT

DIFFERENTIAL OUTCOMES OF FUNCTIONAL AND NEUROPSYCHOLOGICAL
APPROACHES TO NAMING TREATMENT AMONG ADULTS WITH CHRONIC
NONFLUENT APHASIA

BY

Jacqueline Jay Hinckley

We know relatively little about which type of aphasia
therapy is appropriate for which type of aphasic patient.
Two predominant treatment approaches, functional and
cognitive neuropsychological (CN), should have differential
outcomes. Functional treatment should result in large
improvements on the trained tasks and other closely related
tasks, and CN treatment should improve underlying skills
that are broadly generalizable. The present study was
designed to compare the outcomes of functional and CN
treatment on the naming performances of adults with chronic
nonfluent aphasia as it is measured on various standardized
language assessments as well as in functional abilities.
Twelve adults with chronic nonfluent aphasia were randomly
assigned to each of the two treatment groups. The two
subject groups were comparable in terms of age, time post
onset, socioeconomic status, and aphasia severity. They
participated in cognitive assessment and depression
screening prior to participation. The pre/post assessment

battery included selected standardized measures of

‘ .
533;:93‘3352
;:~flfl"r ¢
U¢3Lvugse .

guano; Cr.

‘
a .
5~4¥bo 45a;

”R
I."

.ch 12:61
inerapy we
5"-

...e seagn

‘
T. "'5“ \vr‘p
‘4 dLyy

. 1 be.

a. .
“V‘N‘Q *

‘ ‘ "v‘
.HQC“‘ V‘Ug.

comprehension, naming, and functional communication,
discourse samples, and participation in a specially designed
functional task, involving a role-play of catalog ordering.
Each subject participated in twenty hours of speech/language
therapy weekly for five weeks, all of which was derived from
the designated approach, functional or cognitive
neuropsychological. Results were consistent with original
predictions. Functionally trained subjects improved notably
on the trained task, and a task closely related to its
performance, oral naming. CN trained subjects improved on
broad measures of language abilities, including standardized
functional communication assessment and caregiver ratings of
communication. Among the functionally trained subjects,
those with lower cognitive abilities in new learning and
nonverbal problem solving did better than those with better
cognitive abilities. There was an important correlation
between cognitive abilities and treatment time required to
achieve criterion. These findings are important for

clinical decision making and ongoing clinical research.

Copyright by
Jacqueline Jay Hinckley
1999

To my mother, who taught me to persevere

‘F‘a

 

.

I wound
P‘CTTWY ‘ mm
M. V a‘ CV}.

V‘vuV-

V

“-1 TN 8'.- ‘P
FA”; .5

...ee c‘r

‘ l
9.3.5.339 52C
QT Sup“ Y‘-
u... “My,
- .

‘V

:3: data col

ACKNOWLEDGIENTS

I would like to thank my teachers who served on my
doctoral committee: Tom Carr, John Eulenberg, Antonio Nunez,
and Ida Stockman. Special thanks are extended to my
committee chair and advisor, Janet Patterson, for all her
guidance and input. I am also grateful to Holly Craig, for
her support in every way, including access to her facility
for data collection. This project could not have been
completed without the willing cooperation of the
participants with aphasia and their families, to whom I am
indebted. Finally, I thank Ken, who was an endless source

of help, understanding, and encouragement.

vi

Y
.
I'C

u-q
"
-1".

‘
¢a‘

.
‘

E“
7"IAV‘ n
:VUD

Cczoa

3

~n n
‘5‘“...

Ugui
‘I
an
«a
H
— -.
‘W.

Iowa

‘qvo
‘.

as '1'- v-
r.

OUfi

 

.|.- .....t . r r :0

i i .. r C / o w. 3 Ch 2.8 a a P

C. s . ‘4 S Q 1-. it e O C e D. r a Co F. a. ?m ml ma W to. f.
7‘ EFIU ~RU S to; e e D C S Vs :1! e «G O .1 .1“ 5“ TIM Nu; M. x» e
1 emceemrom. P ewuucCFme 9%. 0.0 or

k . a ‘.
Co 8 S I “W. 0‘ m; .. are.“ ON Dc 0; C D.
I. \-

‘.n DA

TABLE OF CONTENTS

LIST OF
TABLES .................................................. ix
CHAPTER 1
APHASIA THERAPY EFFICIENCY ........................... 1
Comparing Aphasia Treatment Types in Group Studies ..... 3
Towards a Theory of Aphasia Therapy .................... 9
Explaining Therapy Efficiency: A Cognitive Science
Approach .............................................. 15
Cognitive Neuropsychological Approaches to Oral Namingifl!
A Functional Approach to Aphasia Therapy .............. 30
Research Questions ..... ' ............................... 35
Predictions ...................................... 37
CHAPTER 2
METHODS
Design and Rationale .................................. 41
Subjects .............................................. 41
Eligibililty Criteria ................................ 41
Subject Selection .................................... 45
Subject Description .................................. 46
Assessment of Cognitive—Linguistic Indicators ......... 50
Pre/Post Testing ...................................... 51
Pre/Post Standardized Language Assessments ........... 52
Pre/Post Functional Task ............................. 55
Development of Catalog Ordering Task ............. 56
Concurrent Task Condition ....................... 65
Scoring for the Catalog Ordering Task ............ 67
Pre/Post Measure of Social Validity .................. 67
Pre/Post Discourse Tasks .............................. 68
Order of Presentation ................................. 69
Treatment Procedures .................................. 69
Cognitive Neuropsychological Treatment ............... 71
Functional Treatment ................................. 84
Reliability ........................................... 87
CHAPTER 3
RESULTS
Comparability of the Two Groups: Demographics ......... 89
Comparability of the Two Groups: Cognitive Battery ....89
Comparability of the Two Groups: Language Testing ..... 9S
Pre/Post Differences ................................. 106

vii

 

 

Patterns
Cog: ive
Cognitive
Improvere
Cognitive
Discourse
”511.9733. FC'.
32 CUSSICI:
General E
Different
Generaliz
Possible
Possible

Patterns of Transfer to Discourse .................... 126
Cognitive-Linguistic Influences on Treatment Success.132
Cognitive-Linguistic Influences on Rate of

Improvement .......................................... 136
Cognitive-Linguistic Influences on Transfer to
Discourse ...................................... 138

CHAPTER FOUR

DISCUSSION
General Effects of Treatment ......................... 139
Differential Outcomes of the Two Treatments .......... 139
Generalization to Discourse .......................... 142
Possible Influences of Cognitive-Linguistic Ability..142
Possible Explanations for Patterns Observed .......... 143
APPENDICES ............................................. 149
BIBLIOGRAPHY ........................................... 186

viii

LIST OF TABLES

Table 1. Summary of the primary characteristics of
functional and cognitive neuropsychological approaches to
language treatment, as they are particularly related to the
current project.

Table 2. Subject description information.

Table 3. Ordinal position of each routine and embedded
subroutine among the actual calls to representative catalog
ordering companies.

Table 4. Average sentence length (ASL) in words per
sentence and number of sentences for each of the three
actual examples of phone ordering

Table 5. Subject description for the three aphasics placing
catalog ordering phone calls.

Table 6. Overall accuracy and error types for the three
subjects with aphasia in actual calls and in the simulated
catalog ordering task.

Table 7. Description of additional treatment goals and
strategies for each subject.

Table 8. Description of assessments and performances used
to determine the nature of the naming disorder for each CN
subject.

Table 9. Cueing hierarchy for each CN subject.

Table 10. List of trained and untrained items for each
subject in the CN treatment.

Table 11. Treatment strategies used for each functionally

trained subject for the catalog ordering task.

ix

 

 

 

Table 12.
Table 13.
c sprehen.
Table 14.
.....--,. 53.

f-
to

we»-.. a

table 15.

‘fl‘ctiOn .. W

I

I
(a
r v

Table 12. Performance of subjects on the cognitive battery.
Table 13. Pretest performances on standardized measures of
comprehension for both subject groups.

Table 14. Pretest performances on standardized measures of
naming for both subject groups.

Table 15. Pretest performances on functional communication
tasks for both subject groups.

Table 16. Concurrent task baseline reaction times for each
subject.

Table 17. Difference scores on standardized measures of
comprehension for the two subject groups.

Table 18. Difference scores on standardized measures of
naming for the two subject groups.

Table 19. Difference scores on functional communication
tasks for both subject groups.

Table 20. Pretest and posttest performances of the
functionally trained subjects on the three phone-related
items on CADL—z.

Table 21. Percentage of the total change on the CADL—2
attributable to improvements in different response
modalities for each subject.

Table 22. GET; scores for both subject groups.

Table 23. Difference scores for task duration in minutes on
functional communication tasks for both subject groups.
Table 24. Differences in accuracy and task duration for
catalog ordering with concurrent task performance for both

subject groups.

 

 

table 25. Dif
catalog order:
Iable 26. Fe:
pcsttestizg fc

Table 27. 3.2.".

Table 28. 3:5
:he three disc

lele 29. Eif

Table 25. Difference scores for reaction time on the
catalog ordering task for both subject groups.

Table 26. Percentage of untrained items produced at
posttesting for each of the CN subjects.

Table 27. Number of CIUs at pretest for each of the three
discourse types for both subject groups.

Table 28. Difference scores on number of CIUs for each of
the three discourse types for both subject groups.

Table 29. Difference scores for catalog ordering
performance under quiet and concurrent task conditions for
each subject.

Table 30. Amount of treatment time required to achieve

criterion for each subject.

xi

 

 

_ ’Iv“"l'0

 

In his C

reviewed the

r
2. Are
worth :h
money?

3. What
fixariOus i

J

 

Barley discus;

measurement 0;

ConduCting re:

VA
U:

Ports have 1

therapy On 1111.

relation Sh i ps

"i
V ‘6.
usea‘“

en: 8616

an i. : _
was 1:1 a:

a ”‘1 H
mandated SC

Chapter 1

APHASIA THERAPY EFFICIENCY

In his oft-cited paper published in 1972, Darley
reviewed the literature on aphasia therapy effectiveness and
suggested that the field of speech/language pathology pursue
the following three questions.

1. Does language rehabilitation accomplish measurable

gains in language function beyond what can be expected

to occur as a result of spontaneous recovery?

2. Are the language gains attributable to therapy

worth the necessary investment of time, effort and

money?

3. What are the relative degrees of effectiveness of
various modes of treatment of aphasia? (pp. 4-5).
Darley discusses the critical effects of subject selection,

measurement of change, and definition of therapy in
conducting research designed to address these questions.
Over the course of the intervening twenty—six years, many
reports have been published regarding the effects of aphasia
therapy on ultimate communication performance, and the
relationships between subject characteristics and type of
treatment selected.

Robey (1998) subjected many of these studies to a meta—
analysis in an effort to determine the weight of the
accumulated scientific evidence. The meta-analysis was

intended to examine questions that are hauntingly similar to

those raised by Darley in 1972.

 

 

Do differ~
correspor.
of treat:-
What is ::
severely :
individua '.

less resg.
itbey's result

average. Sm:

 

 

it. oath the ac

.-...clusions a:

‘
D
“79:52]. and I

The star:
L M I

6f: 5L
.ECcS Of tre

“‘W Indeed

"V

animation t

Sufti -
.L‘Cxlerlt h‘

“w”.

«’5‘? I
“‘*YSls. ~

‘

“.3 do
es 20+ ,
L
‘9‘. 1
M""‘E‘Ow
‘ ,fi-
(net‘-
C“;
\‘~:‘
t pr0;~ ‘
A‘E
CG“;
«fined C1" 1
is. I

Do different amounts of treatment bring about
corresponding magnitudes of change? Do different types
of treatment result in different magnitudes of change?
What is the magnitude of the treatment effect for
severely aphasic individuals (or mildly aphasic
individuals)? Is one or another form of aphasia more or
less responsive to certain treatments? (p. 173)
Robey’s results suggest that aphasia therapy is effective on
average. Significant outcomes result when therapy is begun
in both the acute and chronic stages of recovery. These
conclusions are essential to the ongoing practice of aphasia
therapY, and respond to Darley’s (1972) statement that
speech/language pathologists must demonstrate effectiveness
of therapy in order to continue having a role in aphasia

management.

The startling conclusion of Robey’s (1998), however, is

that “there are too few studies for examining differential

effects of treatments for different types of aphasia” (p.

184). Indeed, only one treatment type, stimulation-
facilitation treatment, was targeted and described in a
sufficient number of studies to be included in the meta-
analysis. Although this treatment approach is commonly used
by many aphasia therapy practitioners, it is an older
technique originating with Schuell and colleagues (1964),
and does not reflect the full range of the newer
developments in aphasia therapy. Our inability to match
client profiles with appropriate treatment choice is a
continued challenge in our field, and Robey's (1998; 1994)

meta-analyses once again emphasize this need.

 

\I-JIEIII‘

 

We are
. r “
:tat perta-

t V . m
na'i'e 5:10on

ccmlete t;

Cf producti

individuall
detriment I
Care {9301;}
Finally. ti
When We hai
fmCthn a:
neurobeh av;

DeriOdS t1"..-

We are facing a convergence of contradictory forces
that pertain to the practice of aphasia therapy. While we
have shown that aphasia therapy in general is effective as
measured by standardized language battery performance, we
have not shown that these improved performances are
meaningful in terms of the ability of people with aphasia to
complete typical daily activities or to change their level
of productivity. We also have not demonstrated conclusively
that we know, beyond an implicit process based in clinical
experience, which treatment approach is best for which
individual. This lack of scientific evidence is a major
detriment to our field during a period of shrinking health
care resources and competition for rehabilitation dollars.
Finally, these negative influences are occurring at a time
when we have a rapidly growing body of knowledge about brain
function and plasticity after injury, which suggests that
neurobehavioral change after injury likely occurs for longer
periods than previously thought. This extended period of
recovery is more highly affected by neurochemical and
behavioral individual differences than the medical community
has typically assumed. Consequently, research designed to
address these issues is critical to the future of adults
living with aphasia, the practice of aphasia therapy, and

the literature on aphasia rehabilitation.

Comparing Aphasia Treatment Type: in Group Studies
The first study to compare different types of aphasia

therapy was published in 1970 (Sarno, Silverman, & Sands)

 

 

and found
of severel
three tree
were comps
education

ton-progra

Writif‘. , a
Programmed
choose the

iCEITs _ s.

and found no difference between the final performance levels
of severely impaired adults with aphasia assigned to each of
three treatment groups. Thirty-one adults with aphasia who
were comparable in terms of age, time post onset, sex, and
education were assigned to either programmed instruction,
non-programmed instruction, or no treatment. Programmed
instruction involved the practice of six vocabulary items in
a pre-established routine involving visual recognition,
writing, auditory comprehension, and oral production. Non-
programmed instruction allowed the individual clinicians to
choose their own approach to training the same vocabulary
items. Subjects in each treatment group received up to 40
hours of therapy. The results of the study showed that
there were no significant differences between the groups in
terms of retention of a speech behavior one month after the
therapy. The study was frequently discussed as having a
negative influence on the perceptions of aphasia therapy
effectiveness overall.

Shewan and Kertesz (1984) compared Language-Oriented
Therapy (LOT) with stimulation-facilitation therapy in a
group of 50 adults with aphasia randomly assigned to one of
the two treatments. All language modalities were addressed
in a structured approach of stimulus presentation, response,
and reinforcement in sets of 10 stimuli at a time during
LOT. A less defined and more clinician—driven approach
characterized the stimulation-facilitation approach.

Although subjects in both of these treatment groups

 

 

I ‘- : ‘3 9!
s;g:-ficar.t-,
after treatmer

4 :5:
nc s.gn-..car.t

q

1

.a the unrest

P

‘ .. V" n
aruOut taae tre

enzjects perf:

 

7 1-
.r. a .iteratur

swamarize the

I

treatment and
‘0

IT.

significantly improved on a standardized aphasia battery
after treatment compared to a no-treatment group, there were
no significant differences between the two treatment groups,
and the investigators were unable to make any conclusions

about the “treatment of choice”, or whether particular

subjects performed better in one treatment versus the other.
In a literature review published in 1986, Shewan wrote: "To
summarize, there is little evidence comparing types of

treatment and what there is allows us to conclude little”

(p. 40).

Conventional speech/language therapy consisting of
language drills has been compared with supportive counseling
(Hartman & Landau, 1987). Sixty patients with aphasia were
randomly assigned to two groups, either speech/language
therapy or counseling. Both groups received therapeutic
services two sessions weekly for six months. The
investigators found no significant difference in final
language performance at posttest between the two types of
therapy. This study, however, had a decidedly different
goal than those designed to compare different types of
language intervention with each other. It also did not
control for the support that subjects might have received or
perceived during their speech/language therapy
participation.

A similar design and treatment rate were applied in a
comparison of two forms of treatment for auditory

comprehension (Prins, Schoonen, & Vermeulen, 1989).

 

 

.Ji

{initiation t
treatment Pro
presented a 53
stirrali. Ana
the subleCtS 5
revealed r10 6-
cotprehensior‘.

Holland 5
of six ratchec
received two c
pair of subjec
aphasia. One
conversational
received 45 mi
daily, A star:
Smiles were 0

C.
a..er hospital

Stimulation therapy was contrasted with a systematic
treatment program for auditory language comprehension, which
presented a specified hierarchy of word and sentence
stimuli. Analysis of the pre/post performance results of
the subjects randomly assigned to one of the two groups
revealed no differences in outcomes of auditory
comprehension for the two different treatment types.
Holland and colleagues (1983) compared the performances
of six matched pairs of acutely aphasic subjects who
received two different types of aphasia treatment. Each
pair of subjects was matched on age, type, and severity of
aphasia. One member of each pair received 15 minutes of
conversational therapy daily, while the other subject
received 45 minutes of conversational and didactic therapy
daily. A standardized aphasia battery and conversational
samples were obtained pre/post therapy, and at one month
after hospital discharge. At the time of discharge, the
subjects who had received the conversational treatment only
showed significantly better performance on the aphasia
battery. At one month post-discharge, however, there were
no differences between the performances of the subjects
receiving the two different treatments. This pattern was
also demonstrated in the pragmatic-linguistic analysis of
the discourse samples collected at the same time points
(Murray & Holland, 1995). The authors suggest that during

the acute phase of hospital rehabilitation, 15 minutes of

 

‘ x;- ;<" J

 

A;

:on'.'ersati Ita'
arounts of di

Springer
ori nted learr
the trai ing :
The criterion
required prod-
sutjects mice:
receiving L02.
stimul tion 13:
With larger d;
50:35, but sti:

$31 ,
~v-«0k-up.

lea

 

Particular hm
associated w‘ p.

With such

Stildies aVai lat

.0 lack a means
tl‘e '
. right patie
Su'yect
J groups

a; u
renemOre’ 0'

A
N
5

u S in"
.C..Oed in a

.aQtOr

S the

Of

3.. .,
ayDIOacheS t
aller- .
A

conversational therapy is just as efficient as additional
amounts of didactic therapy.

Springer and colleagues (1993) compared linguistically
oriented learning (LOL) to a general stimulation approach in
the training of wh— question production and prepositions.
The criterion task was a short scripted dialogue which
required production of the targeted items. Twelve aphasic
subjects underwent each of the two treatments, with six
receiving LOL training first, and the other six receiving
stimulation first. The LOL training was most consistent
with larger direct effects of production of the targeted
forms, but stimulation was associated with larger effects at
follow-up. Unfortunately, the authors do not suggest any
particular hypotheses for the different outcome patterns
associated with the two different treatments.

With such a limited number of treatment comparison
studies available for review, it is apparent why we continue
to lack a means of determining the right type of therapy for
the right patient. The group studies reviewed all included
subject groups with mixed aphasia types and severities.
Furthermore, only the linguistic status of the subjects are
described in any of these studies, and other cognitive
factors have not been taken into account in relation to type
of treatment. Some studies have not based their treatments
on approaches that are well—defined in the literature, or
alternatively were intended to be tests of newly developed

treatment approaches. Finally, none of these studies

 

provide 2:31.302:
expected to I'

leC'IT.StaIlC€S .

treatments. i
that one treat
for a variety
analysis and d

h

v certain tre

 

The tuner:
”A:
SNBIeS have ll

+
Jeatxent eC C:

c~ ‘*
““91 Sample r

3': V

. ill-Suni'ec.‘
r: in

“starch d Si n

it
‘n‘
kw"
“‘es to 05‘,
r‘
‘d
111d:
‘
eff: .

‘v
.‘Q~ ~
Cc .:.ent t
fps
3“

provide hypotheses about why certain treatments might be
expected to be more efficient than others, or under what
circumstances. These shortcomings relate directly to our
inability in clinical aphasiology to explain the nature of
therapy and the proposed mechanisms for change within our
treatments. The nature of the mixed group designs implies
that one treatment could be found to be better than another
for a variety of aphasic subjects, and precludes the
analysis and discussion of particular individual responses
to certain treatments with varied outcomes.

The numerous factors that may influence group design
studies have led some to pursue the investigation of
treatment efficiency through the use of single—subject and
small sample research designs (Holland, 1975). Smaller
subject designs are advantageous because a larger corpus of
data on each subject can be generated and analyzed for
within-subject effects. A limitation of single-subject
research designs, in particular those intended to address
the question of which treatment for which patient, is that
it is very difficult to generalize from single subject
studies to other patients. These studies may reflect the
nature of individual differences as much or more than the
efficiency of treatment. Although we have many single
subject studies now addressing the comparison of various
treatment types, we are still challenged by our inability to

summarize across these data and generate working principles

 

 

that Can be a:

Cfactice.

search f0r “‘3
1995). A more

. 1 ‘ -
theoretical Dc

 

individual cine.
and outcomes.

addressed the

tirectly. Pet
reta~theory wo
carbination of
process with d

In other words

we to incorp
to use with wit

to accommodat

Sat
~~of ».
apPIOdC}
Par
4 1",
r' 9:38. w

that can be applied more broadly within our clinical

practice.

Towards a Theory of Aphasia Therapy

The limitations of the research designs applied have
been cited by some as the primary stumbling block in the
search for treatment efficiency data (Wertz, 1993; Wertz,
1995). A more fundamental problem is the lack of a
theoretical basis for investigating the issues in matching
individual characteristics to particular treatment regimens
and outcomes. A recent series of articles in Aphasiolggy
addressed the need for a meta-theory of aphasia therapy

directly. Petharam and Parr (1998) suggest that such a
meta-theory would.2make the process of selection and
combination of approach explicit, and directly link this
process with definitions of clinical expertise” (p. 441).

In other words, a meta-theory of aphasia therapy should be
able to incorporate and address the issues of what therapy
to use with whom and when. This meta—theory should be able
to accommodate the diversity in the field, from social and
psychological issues in aphasia, to cognitive
neuropsychological treatments.

Some of the authors in this debate embrace the World
Health Organization’s (WHO) framework of impairment,
disability and handicap as a means to integrate the diverse
set of approaches applied in aphasiology today (Petharam &

Parr, 1998; Jordan, 1998). The WHO framework has also been

 

 

 

advocated I

u
M 1 ‘ - 7"

c:.;..;l;n-ca U

I” 4‘” ya.-

lu-Jaobntt at L
b

I .' 5'
are defin c

' ' ' a
individual

appr p"at<

..

It»...

\‘iulv ~

«Le‘ 9"” it
a; “I

t'. '

n
“HQA

\Vrfi

advocated by ASHA in its development of functional
communication assessments (Frattali et a1, 1995). The
impairment level of analysis includes analysis and treatment
of specific language component deficits. Disability issues
are defined as the effect of the impairments on the
individual's ability to function in life roles that are
appropriate for age, gender, and social status. Society’s
reaction to the disability and the resulting limitations on
the individual’s lifestyle are incorporated into the
handicap level of the framework. Using this framework,
cognitive neuropsychological approaches are seen as mapping
onto the impairment level of the framework, and other social
approaches to aphasia are linked to the disability level of
the model.

This model is advantageous because it facilitates
communication between many different schools of thought and
approaches. It provides a common map on which advocates of
various techniques can point to the level of functioning
that is being addressed. This encourages interaction
between researchers in different sub-specialties of
aphasiology at a time when the combination of various
approaches at the different levels is seen as advantageous
(Sarno, 1991; Lesser & Algar, 1995).

The WHO framework provides a gross conceptualization of
different levels of functioning. It fails to meet the
requirements set out by Petharam and Parr (1998) for a

theory of aphasia therapy, and does not explicitly set out

10

 

|
any ll’ecthX

suggest how t.
interact, or 3

1
carageo .

the relations}
assessment nee.
charge among a

batteries, the

(Porch, 1967la

 

 

 

:5”
MIA) were ad:

 

any directional or predictive relationships. It does not
suggest how the different levels of functioning might
interact, or how these interactions should be assessed or
managed.

Recently, Ross and Wertz (1999) set out to investigate
the relationships between impairment— and disability-level
assessment measures for determining initial severity and
change among adults with aphasia. Two standardized aphasia
batteries, the Porch Index of Communicatiye Ability (PICA)
(Porch, 1967)and the Western Aphasia Battery (WA§)(Kertesz,
1982) were administered as measures of impairment, and the
Communicatiye Abilities in Daily Liying (CADL) (Holland,
1980) was given as a measure of disability. Initial severity
level as determined by the two impairment measures were
related, and these were also related to the disability
measure. The authors found differential patterns of change
on all of the measures after a study period of 2 to 9
months. Change on one impairment measure was not related to
change on another, and change on either impairment measure
was not related to disability-level performance on CADL.
The authors conclude that one measure is not an adequate
substitute for another, and that the tests are measuring
different kinds of performance. Although the WHO model and
the results of this study tend to suggest that these levels
do not interact in relation to change over time, it is
unclear how this model might guide us in choosing between

different therapy regimens.

ll

 

Solution,
treatment for
approaches, w:
goals, styles
clinician mo;
and assessment
bechoosin t?

are used and;

 

then, must be
Ir‘acigemerit dec

should offer p

aid other rel a

issue in her I:

i-L, , ‘
“*9 deficit is

I

3*? t .

not addressed

IEStS 0n the C

“a
/I
(.13) I

This turn

oi.-

.thal apha Q

"v
A Offs).-
I K
In»! ‘-
“‘Qltri

Solutions to choosing the most efficient type of
treatment for a particular individual are left to specific
approaches, which bring with them their own assessment
goals, styles, and instruments. Consequently, an individual
clinician typically chooses an approach to aphasia therapy
and assessment in advance, and may explicitly or implicitly
be choosing the approach based on the assessment tools that
are used and/or available. A theory of aphasia therapy,
then, must be capable of incorporating assessment and
management decisions from various schools of thought, but
should offer predictive ability unavailable in the WHO model
and other related frameworks. Byng (1994) underscores this
issue in her position that a description of the nature of
the deficit is quite different than questions about therapy

choice: “The decision about whether these impairments should

be treated and how represents a different set of questions
not addressed by the account of the deficit; that issue
rests on the development of a theory of therapy” (pp. 272-
273).

This turn towards the WHO model as a meta—theory in
clinical aphasiology follows a period of enthusiastic
pursuit of cognitive neuropsychological models to describe
aphasic impairments and as a basis for the design of
treatment. Initially, the cognitive neuropsychological (CN)
approach offered an in-depth means of assessing an
individual patient’s pattern of deficits, and was seen as an

important new development in the practice of aphasiology

12

K‘l'

 

 

(5:29 et 31'

their lack of
who does not c
1992) and for
applies to so?
1993). The re
treatment and
approaches for
teen achnowled

subsegaent at:

 

Blew-Au-

.uu», lgsas)
The street.
or engage Dr
F n‘n; ‘
sayllgq‘es to ‘
;':O L. ‘
an; mtlon tn‘

«I
118’:

3338d in t.
ML
other aSpeCtS ‘

functioning.

a: -
MICE

&.

*CUltY aCCI

i 1. .
n aphasiology

(Byng et al, 1990). These approaches were criticized for
their lack of applicability to the average aphasic patient
who does not display an unusual dissociation (Goodglass,
1990) and for their inability to provide a taxonomy that
applies to subgroups of the aphasic population (Kertesz,
1990). The replication problem in CN approaches to
treatment and the inability of CN to constrain treatment
approaches for identified language component breakdowns have
been acknowledged by advocates of the CN approach, with
subsequent attempts to address these issues (Mitchum &
Berndt, 1995).

The strength of the CN approach is its identification
of language processes in the system and the ability of its
techniques to pinpoint areas of deficit. The theoretical
information that it is based on, however, is typically
discussed in terms of the language system alone, and ignores
other aspects of the individual’s cognitive system and
functioning. It is also rightly criticized for its
difficulty accommodating more functional and holistic issues
in aphasiology (Weninger & Sarno, 1990).

External pressures stemming from shrinking health care
resources have brought an urgency to our need to address the
ability of individuals with aphasia to live productively.
The functional perspective to treatment has won over third
party payors as a reimbursement standard, because payors
want to be assured as much as is possible that treatments

are affecting performances on activities of daily living.

13

 

 

Consegcent ly.

cur: nt.y the
facilities.
the conn'tnica
performance <
Fun tio:
the pragmatic
rcre spec‘ ‘4.

-&‘\

{‘2’ F‘ I
u¢4’35) py‘;(
"" n ‘

3‘“ iseo con‘
~ .- 1' .
generajlzatl<

between the 1

Mix: -
Clwynathr‘s
aSPECts Of t"

emphaSi Zed b

‘0 l L
‘SOC Vie»
, .
‘ELT'V‘H
Merger-e,”

Consequently, the functional approach to treatment is
currently the most commonly employed in rehabilitation
facilities. A functional approach to treatment emphasizes
the communication skills required for independent
performance of typical daily activities.

Functional approaches, in particular those emphasizing
the pragmatic aspects of language, have been combined with
more specific linguistic deficit approaches (Lesser & Algar,
1995; Holland, 1991; Holland et al, 1983). Lesser and Algar
(1995) applied a CN approach to assessment and treatment,
but used conversational analysis as a basis for determining
generalization of the treated behaviors to communication
loetween the two subjects and their caregivers. This is an
:interesting approach, but it remains unclear how such
(numbinations should occur, when they should occur, and what

aspects of the different goals in therapy should be
enuohasized by the different approaches. It is apparent that
true functional and CN approaches are two predominant
assessment and treatment paradigms in aphasiology today.
Piunctional approaches are advocated by those taking a
lublistic view of living with aphasia and reinforced by
reimbursement practices, and CN approaches are touted by
neuropsychologists, linguists and speech/language
Pathologists who are attracted to its detailed
identification of deficit areas. A CN approach also

Contributes information to our understanding of normal

14

 

 

 

1agiage prOC

of brain iI‘J‘J

whining T1“
Cognitfwl
provide a the
approaches.
neural struct;
various tasks.
The levels of
processing, t}:
corpleted, and
out these perf
issues that n»
the neasnrenenl
becomes a crit;
These methodol:
theoretical fc;
e‘v'aluated, and
the requiremen:

I‘ll c.
‘41 PE gheram a? r:

 

language processing through description of the performances

of brain injured individuals.

Explaining Therapy Efficiency: A Cognitive Science Approach

Cognitive science approaches have the potential to
provide a theoretical basis for both of these treatment
approaches. Cognitive science analyzes the mental and
neural structures and processes used in the performance of
various tasks, in conjunction with a complete task analysis.
The levels of analysis include general domains of
processing, the algorithms by which the functions are
completed, and the neurophysiological substrates which carry
<out these performances (Marr, 1982). Cognitive science
eassumes that mental events require time to occur, so that
iihe measurement of speed and efficiency in task performances

becomes a critical component of the cognitive analysis.
ITnese methodologies have the potential to provide the
tkueoretical foundation on which aphasia treatments can be
eveiluated, and matched to individual patients, thus meeting
thus requirements of a theory of aphasia therapy as described
in? Petheram and Parr (1998).

The expectation of intervention is that the client will
learn new information of some type, or learn to retrieve
information more efficiently, and be able to use this
information in a variety of contexts outside of therapy. In
Order to characterize these processes more specifically, we
must take into account the ability of patients to attend to

the important stimuli within a context, maintain in working

15

 

 

:enofl’ the CI
the relevant
and gene: ate
Some e"?i
executive fun
patients. in 3
S‘IZjeCtS with
been shown to
or auditory W
free recall c:
retrieval of I
lstergaard & l
list learning
show- performar

retrieval abi ‘

.rom t..e perfc
easterior les

- team the
. i
~3ttiona1 con
earl-f
:«dng th
. 3'
{Gr'ussr ‘
an a.
C .
tke a a.
.. y- '
.«Ofil
.es 0
«to
v DOSteY

memory the critical elements for task completion, retrieve
the relevant information from long term memory, and select
and generate the appropriate response.

Some evidence already exists suggesting altered
executive function and working memory in some aphasic
patients, in particular for adults with nonfluent aphasia.
Subjects with anterior lesions and nonfluent aphasia have
been shown to be impaired at recognition memory for visual
or auditory words compared to normals, and are even worse in
free recall conditions that require reconstruction or
retrieval of items from verbal memory (Beeson, 1993;
(Dstergaard & Meudell, 1984; Cermak et al, 1984). In a word
list learning task, anterior patients were more likely to
:ihow performances consistent with impaired LTM storage and
retrieval abilities (Beeson, 1993; Risse et a1, 1984).
Learning curves for new vocabulary items were decreased
Ccnnpared to normals, and also different than those generated
frxam.the performances of adults with left hemisphere
Posterior lesions (Tikofsky, 1971; Risse et al, 1984).
Allihough these nonfluent aphasic patients showed the ability
IX) learn the meaning of a new vocabulary item presented in a
functional context, they were significantly impaired at
learning the syntactic features and usage of the new word

(Grossman & Carey, 1987). Less evidence is available for
the profiles of working memory and learning performances

among posterior, fluent aphasic patients. This is likely

16

 

I!

 

due to the C31
comprehendifié

Obs r‘t’ai
perforrance 6
consistent w:
in normal 813‘-
oi word list
regairing eii
(Eiapur et al
have been in;
relationship
1996).

Much of
in nonfluent
thOtheses a]
decreased me:
aphasia is 3.
mm. resu;

Olltput - Othi

due to the difficulty ensuring the posterior patients are
comprehending task instructions.

Observations about impaired memory and learning
performance among adults with anterior lesions are
consistent with neuroimaging results of memory performance
in normal adults. Left frontal area was activated in tasks
of word list learning (Grasby et al, 1993) and in tasks
requiring either deep or shallow encoding of single words
(Kapur et a1, 1994). Broca's area and surrounding regions
have been implicated in silent speech, which may have a
relationship to working memory abilities (Habib & Demonet,
1996).

Much of the evidence on memory and attention disorders
in nonfluent aphasia is used in support of nonlinguistic
hypotheses about the nature of of the deficit. The
decreased memory abilities in nonfluent, Broca’s—type
aphasia is seen as supportive of the need to use economy of
effort, resulting in strings of substantive words in oral
output. Other linguistic theories of the nature of Broca’s
aphasia, such as the central syntactic deficit hypothesis
and the application of Chomsky’s Government and Binding
theory, have resulted in a variety of experimental
treatments (for a review, see Kearns, 1997). The opposing
linguistic and nonlinguistic views about the primary origin
of the Broca’s symptom complex led Goodglass and Menn (1985)

to write: “Parsimony as a metatheoretical principle in

neurolinguistics is dead f’(p. 19). These hypotheses about

17

 

 

the nature
treatment e
for dirEECt
approaches
Some a
incorporati
activities ,
cognitive i:
(1986) prov
convergent / <

facilitate 1

daily conte)
activities i
investigatic

An earl
problem of i

Impaired n at

Ste?“ .
“berg p :.

shape recogn

errors We

the nature of the disorder also have completely separated
treatment efforts, without an underlying model or rationale
for direct comparison of assumptions about different therapy
approaches originating from the two theories of deficit.

Some aphasia interventionists have advocated the
incorporation of cognitive stimulation into treatment
activities, in acknowledgement of the potential role of
cognitive impairments in the disability profile. Chapey
(1986) provides specific activities based on
convergent/divergent thinking tasks which she suggests will
facilitate the use of language and the deployment of
linguistic skills in the service of actual functions in
daily contexts. Unfortunately, her model and suggested
activities have never been the source of a data-based
investigation.

An early attempt to apply a cognitive approach to the
problem of identifying the processes responsible for
impaired naming incorporated picture naming, a modified
Sternberg picture recognition, and modified Sternberg random
shape recognition (Mills et a1, 1979). Reaction times and
errors were measured for ten normal and ten aphasic adults.
Effects of the uncertainty or certainty with which pictures
could be named was shown to vary significantly among the
aphasic subjects, based on the accuracy and efficiency of
responses. The authors also showed that in the picture
recognition task, in which naming was not required, the

aphasic subjects responded more slowly and less efficiently

18

u:

 

 

than the no:
impairments
deficit in ‘
itself. ‘3‘“
maired 8.31
relation to
i: a-by-iter
score. Thif
work as well
retrieval p1

Improve
denonstratec‘
therapy that
lh'iegeLCr‘n.
irrprovetnent
Suggested ti“.
Skill that i

.mprovernent

than the normal controls. The results suggested that naming
impairments among adults with aphasia are primarily due to a
deficit in the retrieval process rather than in the lexicon
itself. They also showed that these mild-moderately
impaired aphasic subjects improved their performance in
relation to repeated exposures. This was evidenced on an
item-by-item analysis rather than in the overall accuracy
score. This observation has been supported by more recent
work as well (Howard et a1, 1985). It suggests that the
retrieval process in naming is susceptible to improvement.

Improvement in the process of retrieval was
demonstrated to be the result of intensive speech/language
therapy that incorporated a variety of cueing techniques
(Wiegel—Crump & Koenigsknecht, 1973). The generalization of
improvement to untrained items during a treatment study
suggested that it was the process and use of a specific
skill that improved during intervention, rather than
improvement on specific trained lexical items only.

The individual patterns of improvement were emphasized
in the work of Laiacona et al (1996). They found different
degrees of consistency in picture naming at baseline, and
changes associated with treatment were different across two
anomic patients in their study. One subject had an initial
presentation of low consistency, delayed responses, and
self—corrections, and improved in consistency and response

time over the course of treatment. The second subject

19

 

 

irproved i:
nane after
Other
perform no
acalculic 1
the author:
reorganizec
particular
arithmetic
@1595 Of p]
1‘3""? only l
Ci’pes Worke
EfflClency'
Can reVEal
53'8th and
training.
the COnside
0f flaming Cl
IESanSeS
Anothe
IS a Second

he Subl ect

improved in overall naming ability, and was better able to
name after delay or self-correction than prior to treatment.

Other interesting individual patterns of errors and
performance were observed during the rehabilitation of two
acalculic patients (Girelli et a1, 1996), which suggested to
the authors that the nature of the semantic network was
reorganized over the course of training. However, this
particular training showed effects only on the type of
arithmetic problems trained, and did not generalize to other
types of problems. The assumed reorganization appears to
have only been applied to the particular facts and problem
types worked on in the treatment. Analysis of accuracy and
efficiency, along with the qualitative analysis of errors,
can reveal much about the initial status of the cognitive
system and the presumed changes occurring over the course of
training. These sorts of analyses also lend themselves to
the consideration of individual differences in the patterns
of naming disorder presentations, as well as individualized
responses to treatment types.

Another commonly used methodology in cognitive science
is a secondary task that places additional task demands on
the subject. Tasks that depend on the same set of resources
should both show a decrease in performance when the two
tasks are performed in competition. This logic has recently
been applied to the investigation of attentional resources
needed for picture description by mildly aphasic adults

(Murray et al, 1998). In this study, subjects described a

20

 

4.. ._ ' u"-

 

ict' re ora
udder vario
completin
times, accu
secondary t.
injects ha
resources tj
experiment.
resource alj
for aphasic
be assigned
Analys;
Compe’Cition

V

Degiming tc

exec‘ples deg

picture orally while performing a tone discrimination task
under various instructions that altered the importance of
completing one or another task. Analysis of oral response
times, accuracy, error types, and reaction times in the
secondary task suggested to the authors that aphasic
subjects had reduced availability to the attentional
resources that were needed to perform the two tasks in the
experiment. The results were interpreted in the context of
resource allocation theory, in which the primary challenge
for aphasic adults is a lack of cognitive resources that can
be assigned to the completion of various tasks.

Analysis of the time course, nature, efficiency, and
competition of tasks are promising technologies that are
beginning to be applied in aphasiology. However, as the
examples described above underscore, this kind of cognitive
analysis has so far been used to describe the nature of the
cognitive-linguistic deficit in aphasia. This is, of
course, an important endeavor but one that is different,
according to Byng (1994), than determining which treatment
is effective for whom. It is proposed that these
methodologies can reveal additional information about the
nature of change associated with particular therapies and
the unique responses of individuals to those interventions,
just as this general research logic has provided data for
models of individual differences in complex skill
acquisition and word reading (e.g., Ackerman, 1992, 1990,

1988; Bernstein & Carr, 1996).

21

 

 

Further:
adults wit. 4
certain para:
brain :1:qu e:
learning abii
potential for
rehabilitatic
treatment (Le
brain damage
It is possib]
function may
sublects wit?

a5 aPhasia.

anitiVe Neu
cognitiv

{:9 meat hat,-

whiCh has as

a‘d prOCESSeS

3» ‘
UseaKdOYmS 1“

'7 . 1
"0mm mOdEls

’f‘ 1
lkOitheart eh

Furthermore, taking into account cognitive profiles of
adults with aphasia may reveal important relationships to
certain parameters of intervention. Among traumatically
brain injured subjects who are not aphasic, measures of new
learning ability and executive function correlate with
potential for and rate of improvement in cognitive
rehabilitation, and total amount of time required for
treatment (Lamport-Hughes, 1995). Measures of severity of
brain damage were not predictive of success in treatment.
It is possible that measures of learning and executive
function may also relate to treatment performance among
subjects with other types of brain injury and sequelae, such

as aphasia.

Cognitive neuropsychological Approaches to Oral naming
Cognitive neuropsychological (CN) approaches to
treatment have grown out of the experimental literature
which has as its goal to identify the language components
and processes within the normal system by investigating the
breakdowns in language disordered populations, and to apply
normal models of language processing to language disorders
(Coltheart et a1, 1994). Treatment is seen as a means of
testing the accuracy of the assessment procedures and
hypotheses about the nature of the language system. The
language models are also seen as an effective means of
focusing treatment efforts specifically to the area of

breakdown, thus enabling treatment to be more efficient.

22

 

 

A I:
with the
langiage
by Kay ar.
stimuli a
activate
mating, t
to the ph
phonologi
stored (in
a Single
Oral read
(PhoZexfic

In 5-
IZ’ZOSe tha
t"HDiCal V
to eithEr
althors h
CCIreSpOm
Ellis et a
generfll o:
ir-pairmen‘

Word Sela-c

1... ..
”MESH t"

A CN approach to naming disorders, for example, begins
with the identification of an appropriate model of the
language components involved in naming. In the model applied
by Kay and colleagues (1992), pictures or other visual
stimuli are initially recognized as whole entities that then
activate the semantic system and process meaning. For oral
naming, there is a direct pathway from the semantic system
to the phonological output lexicon and buffer, where
phonological forms of words are retrieved and temporarily
stored during neuromuscular programming. This model assumes
a single semantic system for the various types of input.
Oral reading can occur either through letter-to—sound rules
(phonemic decoding) or by way of the semantic system.

In such a model, deficits in picture naming, such as
those that might be observed during administration of
typical visual confrontation naming tests, can be attributed
to either a semantic or phonological level impairment. Some
authors hypothesize additional levels of impairment that
correspond to particular behavioral deficits. For example,
Ellis et a1 (1994) suggest four deficit types for naming. A
general or specific semantic deficit is associated with
impairment of the semantic system, and a frequency—dependent
word selection deficit is associated with a breakdown
between the semantic system and the speech output lexicon.
Phonological approximation errors should be observed with
impairment of the speech output lexicon, and feature errors

with transpositions are associated with the phoneme level.

23

 

 

 

of processing
inpairment a:
there are no :1
Distractors a
phonological
patient’s err
:‘etermining t

Treatmer.
as remediatir.
associations .
processes, or
§€rform the c:
aPproach reg;
S'JCCessful fC:
Printed wOrdS
Fitient than r

of the orthogr

Table 1 S

 

In the CN approach, complete assessment of these levels
of processing are critical. Various tasks for naming
impairment are presented to the patient to determine whether
there are modality—specific strengths and weaknesses.
Distractors are presented that are based on semantic or
phonological relationships to the target item. The
patient's errors then become an important part of
determining the locus of the impairment.

Treatment of the identified deficit in general is seen
as remediating the impairment by increasing the strength of
associations between the stimulus types and the targeted
processes, or by increasing the individual’s ability to
perform the operation through another, related route. This
approach requires the identification of routes that are
successful for the patient. For example, oral reading of
printed words might be a more successful activity for the
patient than naming pictures, because connection is by way
of the orthographic input lexicon.

Table 1 summarizes the major characteristics of the CN
approach. The aim of CN—based treatment is to identify and
remediate the particular deficit, rather than any function
of the individual. This relates to the impairment level in
the WHO framework. The treatment interaction is centered
around the type of stimuli presented, including
consideration of word frequency, imageability, regularity,
and other psycholinguistic features. The cueing hierarchies

are developed based on analysis of the language system and

24

 

a

, MLQKWGQFTrEQ Efl .m...UCmv\.w...nv~m.--..~ i. Iii ll-f. --
I. ; muUwUQIQflUHuva . H HQCOWHWQ ”UmvmfilenUHNrduUfi‘ . H
«H‘UHMVQHOIU’mnmozé gH-HHZMVOUI --I I. . l-§OHBU§l . . -. -i iwgaaommwm o H.

   
 

 

 

 

 

 

 

 

 

mmW£rUH~AV.HflHfi~H~ U-hhvpldnwméud H.”U.me.HAJ£nV\AanO\Nw.—mvflh o>flUWfl~mfiUU “VG” Wfivn~0ufidvUfi-dsN HO \AHN-VCEZJufl .H. -Hufirfi

25

.Ammma .xomHm w mfi>m “mama .mcxm Hmuwflv .uoomonm unouhsu on»
on Uwumamu >HHMHSUHOHMQ mum xmnu mm .quEumeu mmmsmeH ou monomoummo Hmofimoaonohmmondmc

m>fluflcmoo bum Hmcofluocdm wo moaumflumuumnmno mymafium mnu mo >Hmfifidm .H OHQUB

 

 

 

 

mmmhu Gofiummnv mo Gofimcmnwngfioo
“Amcfipmwn Homo mGflUSHoGHV

mnflnopuo

 

 

 

 

 

 

 

oceans 6H03\mnnu0fim .m HflME paw muonm mo hmam maom .m mflmdfi .m
1&mdflmdu
pmufldowu mo musumc .wmcommmn Mmmu vmummuou
mo memm .homusoum ..m.ov Cannes Umwuwmmumucfl xomnpwmm
mommauownmm mo maoflmGoEHU mcfl>aomufimanoum .mcfl>aom
wouwmnmu no omnmucmo .e lamaoouo no omuouqmo .v monaoamnazH .w
Emanonm
mnu mcwucw>Edoufio Mo mdowfi Ho
numamnum mo omum mo moaufiampoa xmmu\amom m>ofl£oo ou
msofluo> mcflms >3 ufloflmop mo .mofluflaopofi mdofiuo> mcflpdaosfl
mono Ga ooGMEH0mHoQ m>onQEH .m .mcme unmhwwwflp me .m mnkuadem .m
Abommn
thuflaflnmmfipv meu Umummuou
o>mfi£oo ou mnumcmnum m.ucmflao
Abommnuusmfihfloqaflv ms» no pmmon mwfimmumuum
mono uaoflwmv wumflpwewm .m whoummcmgfioo smflanoumm .m HAdZOHHdm .N
Umuflmmnmam ma ooGo>wHwH
Ummonuufloflwon .H HMQOmHmm "meonuhufi>fiuo¢ .H SHd .d
anecdomowmmdmpnz! E .9. 560.. .0 1.5239023 mean—3.500

 

 

 

 

monoooummo ucmsumonu amoemoaonohmmonsoc o>Huflcmoo paw Hmcofluocsm mo kHMEEdm .H OHAUB

26

. o '

  

 

 

w._w..u,.....s. ~s.

   

  

».r. ..t u ., .
mwhfiucflam SuCQ meflfiECU

. p;

Cme

UCSU >HQHZ£QUO> UGUQULEU
.@

53% u
!...H‘UH.§OIU¥®&OKDHZ EHRHZOOU-}

$5.me fl mafia. «dz—Hum

I . IIII
...a.~».5_

! . .u . NV
TCTLQC Cf. HULGL 0:0 .LOQGQ\C®Q

W152: ,iflzogou 1:1

 

 

 

 

I
\mmvhmuU UflUmeU .meOSQ .mEHON
H®UHO QUMUUQMQ .WOOHMUQU HQJUUQ
m30flhflu> .mwdnwflhnvmw \AQHQ meOK .0
1 1 - a. _ OHHUE iii».-- .. \
.H -Hnfilh.

nus 0:00»

 

moudou
pommnnaaflxm HmooH Hmnuo mo
CoauouflafiumM Ho unmauflmgfifl mo
coflumflpofimu ofi> monk» oncommmn
paw Hadfiaum awo3uwn cofiumfloommm
mo numcmuum SH wmmmuocH .m

mEoouso admmmooosm

now Conflsvmu mmfimmumuum
new mmso mo coauohmcom
(“How “GOHuMEHomGfl ucm>maon
nuxouaoo mo Ho>mfluumn\xmmu
on» NO ucmfinmfiamaooom

ou Homauflno mmflmoumuum
m>HuflGmoo mo quE>oHQmU
paw mono Hanuxoucoo .m

HQZ‘NU
“Oh IMHZGNUHS .m

 

xmou
OUAuUMHQ GHSOAB oncommmu ucmflao
mo hocmwoflmmm can homnsood .b

mmmmmwfi mcfl>fiooon
chmumfia ha Conflfihoump
mm homsvmpo coauouflcdfifiou .b

ZOHBdfififl>fl oh

 

mxmmu
coflmcoflohgfioo whoufipsm HON
flanfiwuw pwuduUflm “mono: cmuufin3
paw chauoflm noon mcflnfioo

bozo whoadnooo> pwuwmnmu

Eouu mfimufl mafiaeflum .m

Audo nmuuflhz

GOHuMEHowGfi Hmcomuom sufl3 mUHoo
..m.mv ucoflao Ho36fi>flpcfl How
powwoc mm mpumo mac .Homomxcwm
.mpumo ufipmuo .mconm .mEHom
Hmpuo oofluooum .mmonumo Hosuom
msofiho> .mumfinom hoam maom .m

 

 

awooqouohmmombnz 29568

 

 

madHflflhdfl .m

 

gonna?

MBBHBUQIOO

 

 

 

 

“UsuflOOv H OHAUB

27

 

 

the patier.‘
accuracy a:
intended b;

A Cla:
enrolled a

b .
.reatment .

presented a
EPPI-Gpriate
3'95 "110 gse:
phonologice
PrOdUCing .
"hEther the
si‘lbje‘CtS r.
other rece.

P’iSEtQSt SI

9:»
"M13" '
H. 1
.t‘c t]
S\:‘a:t‘_
‘
C h

the patient’s responses are evaluated based on their
accuracy and efficiency for the specific target that was
intended by the presentation of the stimulus.

A classic treatment study by Howard et al (1985)
enrolled a total of 12 subjects in each of two forms of CN
treatment for anomia, semantic and phonological. The
semantic therapy techniques included pointing to a picture
from four semantically related pictures when the word was
presented auditorily, matching the written word to the
appropriate picture in an array of four, and answering a
yes/no question requiring the access to meaning. The
phonological techniques were repeating the picture name,
producing the picture name with a phonemic cue, and judging
whether the name rhymed with another word. Half of the
subjects received the semantic treatment first, and the
other received the phonological treatment first. Results at
posttest suggested that the semantic treatment resulted in
greater generalization to untrained items, but no
differences between the therapy results were noted at the
six week follow-up.

LeDorze & Pitts (1995) report the case of a patient who
was entered into an alternating treatments study, including
semantic treatment and word form (phonological) treatment.
Semantic matching and oral reading tasks were compared with
uncued naming opportunities as a control condition. The
authors conclude that the administration of treatments that

were directed to the patient’s deficits resulted in much

28

 

 

greater its;

Oral
as tasks i:
patients w
These four
generalize

of general

sward et .
931316111 “:1:

greater improvements than activities that were not aimed at
the particular language breakdowns.

Oral naming, oral reading, and written naming were used
as tasks in a phonologically—based treatment applied to four
patients with varying deficit profiles (Raymer et al, 1993).
These four patients all showed improvement and some
generalization, to various degrees, in contrast to the lack
of generalization in the phonologically treated patients of
Howard et al (1985). The authors do not satisfactorily
explain why their results differ from other authors, or if
there are particular characteristics of the subjects that
might facilitate generalization. Specifically, this study
and others like it only report the linguistic profile of the
patient, without taking into account attention, memory, or
other executive function abilities which are often
hypothesized as being related to generalization ability
(Ackerman, 1990). In short, we know little that could help
us predict how the subjects can and will be able to use the
skill that was targeted in the treatment.

Nickels and Best (1996) review the CN treatment
literature on naming disorders and conclude that for some
patients semantic therapy seems to result in more
generalization than other phonological tasks, but this
result is inconsistent. CN approaches in general would
predict that if a level of processing has improved, then all
tasks dependent on that process should also be improved in

performance. Similarly, untrained stimuli should show an

29

 

 

,

4m? V." :7":
.¢-*.0 6

process I

generali:
one of t?
trerapy.
loci of t
have imp:

355688276

uESt 'W 19'
“filers-31 (

improvement when the action required relies on the same
process (Hillis, 1989; 1994). Nickels and Best (1996)
suggest that one of the areas to be addressed in future
research is determining which variables affect
generalization, especially to spontaneous speech. This is
one of the critical issues in the development of a theory of
therapy. CN approaches have helped us identify possible
loci of breakdowns in the language processing system, and
have improved the techniques available for comprehensive

assessment of language disorders.

A Functional Approach to Aphasia Therapy

The goal of aphasia rehabilitation, in a functional
frame of reference, is to improve a person’s ability to
communicate in everyday life. The functional approach
assumes that context is an important factor underlying
language processing, and that adults with aphasia perform
best when materials and goals are personally relevant. The
general characteristics of the functional approach are
listed in Table 1.

A functional approach to treatment is often seen as
incorporating many general aspects of aphasia
rehabilitation, including discussion of cognitive abilities
and the importance of psychosocial state. This is likely
due to its broad perspective and reliance on the client and
his/her social environment to determine the relevant issues

to be addressed. Sarno (1991) suggests that handling of

30

psychol
is crit
A:
of :3.
general
in whici
relevar.
order t!
OLjecti‘
vari us

‘

AS.
analyst

message

psychological, social, and functional communication issues
is critical to the long-term view of aphasia management.

Aten (1986) and Chapey (1988) describe the importance
of the cognitive aspects of functional treatment, albeit in
general terms. Functional intervention is seen as a method
in which real world meanings are stimulated as both inputs
and outputs within the contexts in which they normally are
relevant. The use of various compensatory strategies in
order to achieve message transmission within a context is an
objective of the approach, and requires the deployment of
various cognitive skill.

Assessment from a functional perspective includes the
analysis of the aphasic adult’s ability to communicate
messages in a variety of contexts, in particular ones that
are important to the individual. Although standardized
language testing is used to inform the clinician about the
strengths and weaknesses of the language system, the
emphasis is placed on the development of personally relevant
strategies for particular situations in which the client is
likely to find him/herself in the course of normal daily
events. Consequently, comprehensive interviewing and choice
of what is personally relevant for a particular client is an
important part of this approach.

The interactive nature of communication is emphasized,
and the therapist is expected to take the role of both
message sender and receiver during the course of

intervention. Chapey (1988) writes: “We become more

31

effectiv
commit
as a fac
well as

therapy.
therapeu

Inc

aCttally
aphaSia
perform}:

Bari
He

Was orig

Et al)

effective communicators as a result of using language in
communication with others" (p. 185). The clinician serves
as a facilitator of appropriate compensatory strategies as
well as a conversational interactant in the course of
therapy. Role-playing and conversation are primary
therapeutic activities in this approach.

Incorporating these guidelines, Holland (1980; Holland,
Fromm, & Swindell, 1999) has developed a standard assessment
tool for functional communication in aphasia (Communicatiye
Abilities in Daily Liying). It is based on the premise that
role-playing is, first of all, possible with adults with
aphasia, and secondly, an informative source of information
about the actual communication abilities of adults with
aphasia. In the development of the assessment, Holland
actually observed the communicative abilities of adults with
aphasia in a variety of settings and related their
performance to that sampled by the assessment tool.

Evidence for the efficacy of the functional approach
was originally described in a study conducted in 1982 (Aten
et al). Seven adults with aphasia received functional
communication treatment, which included the development of
compensatory strategies in personally relevant activities.

Differences in pretest and posttest scores on a standardized

aphasia battery (WM) did
not show any change over the seven weeks of treatment.
However, scores on the Communicatiye Abiiities in Daily
Liying test (Holland, 1980) were significantly improved and

32

were mainté
the treatmé
years pOSt
of the per:
approach.
differentie
relative i:
assessment
Evalue
rooted in s
05 the aphe
conversatic
Emotional
based On t‘r
COlllv'ersatic
For example
with Broca'
fCr inform;
this trainj

were maintained at that level for six weeks after the end of
the treatment. These adults with aphasia were all several
years post onset, and the authors argue for the importance
of the personally-relevant, compensation—based treatment
approach. The results of this study also support the
differential outcomes of a particular treatment, and the
relative independence of impairment and disability level
assessment measures.

Evaluation of the effects of this type of therapy is
rooted in successful message communication and the ability
of the aphasic adult to perform various speech and
conversational acts. Most studies of the effects of
functional treatment, then, analyze posttreatment results
based on the ability of the clients to participate
conversationally or perform a greater number of speech acts.
For example, Doyle et al (1991) demonstrated that an adult
with Broca's aphasia could be trained to initiate requests
for information within conversational settings, and that
this training generalized to conversations with a variety of
partners. Interestingly, this report of results does not
include the number or type of words or phrases used to
convey the requests. Indeed, an important part of the
operational definition for request behaviors in this study
was the use of prosodic cues.

The requisite set of skills to accomplish specific
functional and productive activities is also an important

theme of functional approaches. Sarno and Buonaguro (1983)

33

invest

associ

IEld"
CI per

(the l

investigated the language and cognitive characteristics
associated with meal preparation among female aphasic
patients. They found that there was no particular
relationship between meal preparation, severity of aphasia,
or performance on a battery of memory and perceptual skills
(the Parietal Lobe Battery of the Bgston Diagngstis Aphasia
Examinatign). The authors suggest that the independence of
the communication abilities to the performance of a
productive task of daily living such as meal preparation can
serve as a rationale for the pursuit of various instrumental
goals, and has important implications for treatment. They
suggest that adults with aphasia are able to learn a variety
of tasks that are important to daily living despite the
language impairment.

In contrast to CN approaches, functional treatment is
most appealing for its emphasis on activities of daily
living and social communication. However, there is a lack
of systematic assessment of the results of this approach.
Although proponents of functional treatment allude to the
use of cognitive abilities in the completion of the tasks
that are targeted, there has not been a systematic attempt
to identify what these abilities might be, and whether
success in functional treatment might depend on residual
cognitive strengths. This is particularly important since
functional treatment emphasizes the development of
compensatory strategies, particularly those that are self-

generated. This assumes that clients are able to identify,

34

learn,

I

t—

. n

o In fl

*ao‘ ‘
V‘.f«. IAY'

up Q5.
0‘ .

. , I
Pv‘ V n a
L.G-..4

’tat C

learn, and recall the appropriate compensatory strategies,
which are characterized as volitionally used tools, in the
appropriate contexts with effective results. Functional
training assumes an emphasis on broader cognitive abilities
than CN approaches, however, these broad-based abilities may
be more contextually-bound than more specific skills. For
example, mastery of strategies that facilitate independent
ordering in a restaurant may make an important difference in
the context of a restaurant, but those same strategies may
not be deployed in other functional situations. An actual
analysis of learning and use of strategies within the

functional approach has not yet been attempted.

Research Questions

A project was conceptualized that compares the two
predominant approaches to current aphasia therapy. Subject
description and data collection were designed to maximize
consideration of cognitive factors, as potential indicators
of future performance, possible correlates of actual
performance and learning, and relationships of cognitive
abilities to transfer and generalization. Concurrent task
methodology is one means of investigating the role of
attention and other executive functions on task performance
and possible improvement associated with different training
techniques.

The present study was designed to focus on two
different approaches to aphasia therapy, cognitive

neuropsychological and functional. Naming impairments are

35

common to

the apha51

relatively
that can is
Therefore,
within non
subject de
treatment.
The r
follows.
DO f1;
treatments
treatment

V
apn

asia?

SIDI‘ECts t

IQCeiving

common to all aphasia types, and are a critical feature of
the aphasic syndrome. Naming is consistently and
chronically impaired across all vocabulary categories in
nonfluent aphasia compared to all other aphasia types
(Knopman et a1, 1984). Additionally, nonfluent aphasia is
relatively well—studied and has a foundation of literature
that can be used as a basis for discussing results.
Therefore, the study has been limited in scope to naming
within nonfluent aphasia, to focus on variation of other
subject descriptors and performances during and after
treatment.

The research questions addressed by this project are as
follows.

Do functional and cognitive neuropsychological
treatments aimed at single word production yield different
treatment outcomes among adults with chronic nonfluent
aphasia? Specifically:

1) Are there differences in performance between
subjects trained using a functional approach or subjects
receiving cognitive neuropsychological treatment on a
functional criterion task?

2) Are there differences in performance between
subjects trained using functional or cognitive
neuropsychological treatments on standardized functional
communication or standardized psycholinguistic tests?

3) Does treatment for single word production

generalize to discourse? Are there differences in

36

gene.

ftnc

V
pre:

$.41!”

smyv

generalization to discourse between subjects trained using
functional versus cognitive neuropsychological treatment?

Additional issues were identified for exploratory
analysis. What cognitive-linguistic characteristics are the
best predictors of success in treatment of single word
naming? Specifically:

4) Are there differences in cognitive-linguistic
predictors of success, as measured by performance on
functional and standardized tests, for each of the two types
of treatment, functional or cognitive neuropsychological?

5) Are there differences in cognitive—linguistic
predictors that best predict generalization of single word
production treatment to discourse?

Predictions

Based on the previous literature, it was predicted that
subjects in both training types would make improvements
(Robey, 1998; Robey, 1994). However, the nature of the
improvements should differ based on the training types.
Subjects undergoing the functional training should make
substantial improvements on the catalog ordering task. This
improvement should take the form of increased accuracy of
naming production as well as improved efficiency and
accuracy of performance in the concurrent task conditions.
Specific task training has been shown to have these effects
in other language areas for adults with aphasia (Doyle et
a1, 1991; Avent et a1, 1995). These subjects should

similarly show some improvement in their performance on the

37

fnfi' A

)‘i 544's.

CAQL;Z, in particular on some items that are most similar to
items on the catalog ordering task, since improvement on
this test has been reported as an outcome of functional
treatment (Aten et al, 1986). The functionally trained
subjects are unlikely to show improvement as measured by the
subtests of the PALPA or the BN1, since these tasks are
specifically designed to inhibit the use of the kind of
compensatory strategies that would have been trained in the
functional treatment.

Subjects in the CN treatment should make fewer
improvements overall in the catalog ordering task, since the
catalog ordering task is a complex task requiring multiple
strategies in addition to the trained vocabulary.

Observable improvement for this group of subjects might
include increased production of specific lexical items, but
less overall efficiency during performance of the catalog
ordering task. Subjects in the CN treatment group may also
show improvement on the posttesting of the CAQL;2, since it
requires multiple skills deployed in the role-playing of
various situations, and impairment-based training is
typically broadly generalizable (Calvanio et a1, 1993). The
CN treated subjects should show more improvements on their
BN1 and PALPA scores than functionally trained subjects,
particularly the Picture Naming subtest of the PALPA, again
due to the expected generalization of the training
procedures. In summary, the CN subjects are likely to show

some improvements across a wide range of contexts, whereas

38

the functionally trained subjects should show greater
improvement in tasks that are most similar to the trained
context, and less improvement in tasks that are least
similar to the training.

Assuming that most or all of the subjects improve in
some way during treatment, then there should be greater
increase in the discourse measures, particularly in the
Correct Information Unit analysis, for subjects in the CN
treatment than in the context-based functional treatment.
We expect that broader generalization will occur for the
subjects in the CN treatment, based on the theoretical
model. However, subjects in the functional training might
show an improvement in procedural and narrative discourse
samples, since they have been treated within a sequenced
activity. In this case, the procedural and narrative
discourse samples are highly similar to the trained activity
in the functional treatment.

As an exploratory measure, the cognitive profiles of
the subjects in each of the two treatments can be analyzed
to determine whether there is a pattern between cognitive
abilities and outcome in either of the two treatments. It
is unknown whether particular cognitive abilities are
specifically related to outcome in aphasia therapy.
Although it would appear that subjects with greater new
learning and memory abilities should have better outcomes
than subjects with lower learning abilities, this

relationship has not been previously explored in the

39

literatur

general i z

literature. Therefore, relationships between cognitive
abilities and outcome on standardized tests, including

generalization to discourse, were investigated.

40

Design and

A bet

L

3:. a corn;

C

cognitive
Project c3;
criteria w

H

A u
“4 unCe r'f,‘

bl)

The I;
which to a
Predict-firs
therapy ty
of Change.
bEC’Neen gr
385:5:th

e f
p r‘On’?«3r1c

“R
”‘48ch
‘

Edi

Chapter 2

METHODS

Design and Rationale

A between groups design with multiple outcome measures
and a comprehensive battery for collection of
cognitive/linguistic predictors was employed to address the
project questions. Subjects who met all eligibility
criteria were assigned to one of two treatment type groups,
and underwent all pre-post testing.

The between groups design is an appropriate context in
which to address the issues of cognitive-linguistic
predictors and differential outcomes. Since different
therapy types may result in qualitatively different patterns
of change, it is important to investigate these differences
between groups of subjects. Alternating and contrasting
treatment designs are within subject models in which
performance quality as a result of each therapy could

obscure differences between therapies.

Subjects

El ibil i eria

Subjects were 12 adults with aphasia who were enrolled
in a therapy program at the Communicative Disorders Clinic
at the University of Michigan. Initial eligibility
requirements for all potential subjects included a time post

onset of brain injury of at least 3 months, to ensure that

41

the sublec
All subjec
the classi
these clas
performanc
the four a'
ensured tn
to follow 1
tasks. Ore
short phras
and Sentenc

Sev ri
the M
”3916111, 1
Scale, Seve
reqiired to

u 1.
sabie Spee.

liar:

isluerate\sey

A11 Sn:
L

“ERASphEri C

Subjects Wer

0th; .
.1" hi S tOr‘

r1:
H.301der a
I nC

the subjects were beyond the spontaneous recovery period.
All subjects demonstrated a nonfluent aphasia, according to
the classification criteria of the Bgston Diagagstig Aphasia
Examination (BDAE) (Goodglass & Kaplan, 1983). According to
these classification criteria, auditory comprehension
performance must be at the fiftieth percentile or higher for
the four auditory comprehension subtests combined. This
ensured that subjects had sufficient auditory comprehension
to follow the directions and requirements of the project's
tasks. Oral expression was characterized by single words or
short phrases, anomia, and agrammatic production of phrases
and sentences.

Severity of aphasia was determined by administration of
the Bastgn Diagnostig Aphasia Examinatign (BDAE) (Goodglass
& Kaplan, 1983) and application of the EDA; Severity Rating
Scale. Severity ratings for the eligible subjects were
required to be between 1-3 on the 6 point scale (0 = no
usable speech; 5 = difficulties not apparent to the
listener). Ratings of 1-3 correspond to a moderate or
moderate-severe level of impairment.

All subjects had a history of a single left cerebral
hemispheric CVA of thromboembolic or hemorrhagic origin.
Subjects were also required to: be medically stable, have
corrected vision and hearing to within normal limits, no
other history of neurologic, psychiatric, or communicative
disorder, and be speakers of English as a native language.

Any subject with a known history of substance abuse was

42

mil

l'

v
W

8253

ix

are

excluded from participation. All eligible subjects were
also required to have at least a high school diploma to
ensure a premorbid functional literacy level.

Sources for eligibility criteria were the medical
records (time post onset, lesion localization, other
significant medical history, medical stability, visual
status), and a client/family interview (native language).
Medical records were reviewed to ensure that a report of a
neuroimaging procedure, minimally a CT scan, was available.
Subjects were eligible for the study if they had documented
lesions in the primary language zone (left fronto-temporo-
parietal areas). No further criteria for lesion loci were
set for this project, due to the frequent report of
individual differences in lesion site among patients with
similar behavioral profiles (e.g., Knopman et al, 1984;
Basso et al, 1985). A hearing screening using a portable
audiometer with pure tones presented at 1000, 2000, and 4000
Hz at 20 dB HL (ASHA, 1985) was administered and passed by
each subject.

One important component of the functional task
assessment used in the project is the comprehension of wh-
questions, specifically ‘what' questions referring to a noun
or adjective (e.g., ‘what size?’, ‘what color?’ ‘what
item?’). These question types are among the easiest of the
various wh- forms for adults with aphasia to understand
(Gallagher & Guilford, 1977). (These data are especially

relevant since the research was conducted at the University

43

of Michigaf

subject san

qaestions v.
Performance

reqaired f o

-3

he (:1
qaestion it
Hi.“ -

”is lS cor.

Sallagher 5,

taxi-51’ se<

doc I

are 6 piCCuI
squEncEd e.»

f .
or aCults’

lachldéd in

SPGCifiCally

presented or
(Second Ed‘t

& .
199

of Michigan Aphasia Program, and are derived from a similar
subject sample to the one targeted in the present study.)
It is critical that the subjects comprehend this question
type, and yet it is not specifically tested on any of the
BDAE subtests. Consequently, an assessment of relevant wh—
questions was administered to all potential subjects.
Performance at a criterion level of 75% accuracy was
required for participation in the study.

The question comprehension assessment included 12
question items of the syntactic form, Wh + copula + noun.
This is consistent with the procedures presented in
Gallagher & Guilford (1977) and is the form that is relevant
to the experimental task. Photographs of steps in two
familiar sequenced events (making an omelet, going to a
doctor's appointment) were selected from the “Social
Sequences" photo set (Communication Skills Builders). There
are 6 picturable steps in each event sequence. The two
sequenced events were chosen for their presumed familiarity
for adults, and their congruence with the content of items
included in other portions of the pre/post assessment.
Specifically, items relating to visiting the doctor are
presented on the Cgmmanigatign Agtiyitiss of Daiiy Liyiag
(Second Edition) test (CADL-2) (Holland, Frattali, & Fromm,
1999), and a cooking-related item is included in the
procedural discourse protocol. The question stimuli and
task instructions are listed in Appendix A. The examiner

presented each question auditorily with the corresponding

44

photogra:C
item frOIT

picture.

question
is regain
the stimui
average of
flject
Thirt
L‘niversi t3
eligibilitj
Eligibilit3
Study, incl
hemorrhagic
mOderate $6
03891; All
a“lglish as -

C

m'a.
“-Lh

normal

 

photograph. The subject is expected to point to the correct
item from among a four choice array presented along with the
picture. This brief assessment of wh— question
comprehension is appealing because comprehension of the
questions is embedded in a familiar routine, similar to what
is required to perform the experimental task. The length of
the stimuli ranged from 4 to 8 words per question, with an

average of 5.75 words per question.

figtjggt fielsction

Thirty three consecutively enrolled clients at the
University of Michigan’s Aphasia Program were screened for
eligibility. Nineteen of these clients met the strict
eligibility criteria for possible participation in this
study, including single left CVA of thromboembolic or
hemorrhagic origin with resulting nonfluent aphasia of
moderate severity, and were at least three months post
onset. All of these nineteen potential subjects spoke
English as a native language, were premorbidly right handed,
with normal or corrected to normal hearing and vision. All
of the nineteen potential subjects were asked to participate
in the study according to approved consent procedures. One
subject did not wish to participate because he did not want
to be randomly assigned to a treatment type, and two other
subjects thought that the study activities would be too
tiring. Two subjects consented and began the study
activities, but fell ill for reasons unrelated to the study

and were unable to complete the protocol. Two other

45

subj ec

:ne in

u
nnvt~
:WU- $0

projec

b... 1
taELVe

ti

subjects who met all other eligibility criteria performed
the functional criterion task above criterion performance at
pretest and were ineligible to continue in the treatment
portion of the study.

The twelve remaining subjects all consented to the
project procedures and completed all activities. These
twelve subjects all reached criterion performance on the wh-
question task (minimum 75% accuracy). The average

performance for these twelve subjects was 85% accuracy.

mm

The overall average age was 51 years (range = 24—67),
and the average time post onset was 26.5 months (range 4-
102). All but one of these subjects were male, and all
subjects were Caucasian. All but one subject was of high or
upper middle socioeconomic status (SES) based on education
and most recent occupation, as determined by The Four Factor
Index of Social Status (Hollingshead, 1975) (Appendix B).
This index weights scores for previous occupation and
educational level to generate SES on a five point scale,
from 1 (major professionals) to 5 (menial laborers).

Nine subjects had experienced thromboembolic strokes, and
eight subjects had lesions documented by CT or MRI in
frontal—parietal regions of the left hemisphere. See Table
2. Nine of these subjects had persistent right hemiparesis.
All subjects passed a depression screening, the Qstiattig
Dap;sssign_$sals — Short Form (Parmalee & Katz, 1990; Brink

et a1, 1982; Yesavage et a1, 1983) (Appendix C). This

46

 

 

 

_ .lll‘iilli
GZHB: flog-Hm M§OH§
array—Eel! hm ho A mg» mu «mg » QH
‘HBMM .1 moo. -imlMBM. 20H mmflflil “an: mam OER H30 H06 80...“.

( mam-)

 

. COMUQEIHOMNHNH cojflbnwfihhvmmvfiv Uhvhv 30:173. t." -Hn'.“

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.8 3.3 3.8 3.2: 8.3V 76.!
m.~ m.m m.« 93 v.3 duo:
oflamnom
H a $93 + Hmflmumdouomsme o m R 2 mm mm
m m 3333 m a ma r em mm
m m $333.35; 0 m NH m mm E
m a Hmfimumoéuomsme o m ma 2 2. mm
m m 3663 Hancoum o m m 2 mm mm
m N $383.38; 0 m mm 2 m... E
.ng
as
has-ONE
saunas arouse
BE .3 ho Anna—.293 m Augv OH
598.2 -36 goon 2033 HE man one unzuo nod son—BB

 

 

.GOaudEHOmGfi Gofiumwhommp uomhndm

. N OHAHB

47

UZHB‘I
fibHKI>In
‘H a:

 

 

thmQO

 

2U
HZOMBQ «MIHZOIC «hmdflfie DH
mDUOJ ZflmnNMl-ihO Hmflfi nflu 0&8 KHQZHO “0‘. BUHHNDQ

A-U. ugogv N “nan-Uri.

 

 

 

 

 

 

 

 

 

Am.ov Hm.Hg Hm.ov Hw.hnv Hm.MHV H.v.uv
m.H m.m n.H «.mu n.om ado:
HouHoumm
N m nonconm an N 5H 2 em mzo
mHHmem Hommn
+ HouHmHmm
H m nonconm m H m S we mzo
HmuHoucm
m w noncoum o N m 2 mm wZU
chHonm
m m (oucoum o H NOH 2 mm mzo
HmuHmumm
nouogfimu
H N noucoum o H w 2 pm Nzo
HouHmqu
m v 68959 0 H Hm 2 mm Hzo
meomnmom
oBfiasu
zo
GZHH‘K
ufiufisuu uiomam Huufizolv Hmmdflhg DH
«Ha: banana mg ,3ng ho HE man can. Kwanzaa H01 FORGE

 

 

 

 

 

 

 

 

 

 

8.283 N 039—.

48

F

 

«OZHB:
sinus—E! «aouam A “5» «mg » “H
53““)!!! - .Hhmdmnol mglzo.wmuq “O "E Nana. 0“.“ “use "0‘ .UUHHJQDM.

Aura-haw“! v N 9::qu

.HmcmumHH wnu on ucmhmmmo on uoc HMS meuHDUHMMHU I mQMUHUGmn zomme wHQHGHmUmHU
n a “monoumHmmo on
“QHmn nuHB oHQHmmom
“GOHmmoHme Humucwfimohm deounu GoHumochEEou
o .mcHumm huHHm>mm Mdam on» Go pwmmm m

HMEHGHS

n m “GOHmenwugfioo Ho nomwmm CH honosHm mo mmoH 080m

Ho mHuuHH QuHB mEmHQOHQ haphho>m HHm umoEHc mmsome Goo quHucm
mH muoonndm HmHHHEom uzonm GoHucmHobcou
n H “GoncoanQEoo Ho zowomm oHQcms 02
.H0 XHUGmmmmv mwquu on» Go muddfiduoum m>Hummoc mo Hmnfifiz.

N

 

.muuommn Hm: Ho\pcm so no memm m
.m>HmDHUUO
.Hm xHUGwaamv meGH Houoom usom mflu ha poGHEHmqu mSuoum UflfiocoomOHoom u mmm

o “UHmmnHoEEom u m

 

 

 

 

 

 

 

 

 

 

 

 

A . U . e v
Am.ov AM.HV Ho.ov AQ.ONV Aw.MHV lfldflfl
H." m.m 0.H m.mN Hm HHGHObO
«05.52
uhflfléuu «uiomam HuflBZQF. Hundflwv nH
flan: saunas nmpuoq 2965 no as swam one go no: Bushman

 

8.2.8 v N 033—.

49

 

instrument
screening
advantageo
standard a
auditory p
the twelve
depression
form), and

is borderl.

Assessment
SUb'j- ec
:ingdistic

l’esults car

F‘

”*9 910b\a1
MCUrik e: a
useful too.
:nC .
““Omlatlo“

The
”d .
IEqmlrEE
21:01 6 f
ro

instrument has been successfully used as a depression
screening with older adults. It is particularly
advantageous for use with adults with aphasia because the
standard administration technique allows either visual or
auditory presentation of stimulus items, or both. Eleven of
the twelve subjects scored within the normal range on the
depression screening (0-5 negatives on the 15 item short
form), and one subject scored 6 negative statements, which

is borderline for a mild depression.

.Assessment of Cognitive-Linguistic Indicators

Subjects participated in additional cognitive and
linguistic assessment to determine whether these assessment
results can be used as indicators of success in treatment.
The ngbal Aphasia Neurgpsygholggical Battsry (QANB) (van
Mourik et al, 1992) (Appendix D) is an appropriate and
useful tool with which to collect cognitive performance
information on adults with chronic aphasia. The battery
includes assessment of the following cognitive abilities:
attention in a visual cancellation task, memory in a task of
delayed recognition of object and facial pictures, and a
task of nonverbal auditory recognition of environmental
sounds. The battery also includes administration of the
Raysa’s Cglgrsd Prggrsssiys Mattigss (RQPM) (Raven et al,
1979). The RQPM is a measure of nonverbal problem solving,
and requires the subject to point to the correct response
choice from among those presented visually. The gang also

includes assessment of visual perception through

50

adri istrat
memory was
pointing ta
Subtes
Langiage Pr
Coltheart,
Repetition,
The Pointin
groups of w
Sequence.
resaponse fr
gIO‘ups matc
event that
abilities.
performaIlCe
S‘3e1’g, 199
lobe injUry

C
.or Patient;

administration of the Devslgpmental Tsst Qf yisual
Pergsptign (Hammill, Pearson, & Voress, 1993). Working

memory was assessed via the administration of digit span and

pointing tasks.

Subtests of the Psygholingaistig Assassmsnts gf
WW (PALPA) (Kay. Lesser 8:

Coltheart, 1992) were administered, specifically Digit
Repetition, Digit Recognition, and Pointing Span subtests.
The Pointing Span subtest requires the subject to listen to
groups of words and point to corresponding pictures in
sequence. The Digit Recognition subtest requires a yes/no
response from the subject to indicate whether pairs of digit
groups match. These latter two subtests are useful in the
event that subjects have decreased oral repetition
abilities. Finally, rule-learning was assessed by
performance on the Wisggnsin Card Sgrtiag Task (WEST) (Grant
& Berg, 1993). This assessment tool is sensitive to frontal
lobe injury (Milner, 1963) and is particularly appropriate
for patients with nonfluent aphasia who have anterior

lesions.

Pre/Post Testing

The pre/post measures included standardized language
assessments, performance of a functional task under
conditions of quiet and concurrent task, a measure of social
validity, and discourse measures. The standardized measures
relate to previous literature on outcomes of aphasia

therapy, and are representative of the theoretical

51

orientations
criticism of
their lack c
the functior
commicatic
performance
to know whe'.
adainistere:
as perceive:
changes the
inStrument .
adainistere
measure of
communicati
mm
Standa

tr .
ea Lfitent a

 

orientations of the two training approaches used. One
criticism of standardized assessment measures however, is
their lack of ecological validity. The administration of
the functional task provides a naturalistic measure of
communication performance, which can be correlated to
performance on the standardized tests. It is also critical
to know whether changes observed on assessment measures
administered in the clinic relate to communication behaviors
as perceived by caregivers, or whether caregivers perceive
changes that may not be observed in the clinic, so an
instrument designed to collect this information was
administered. Finally, the set of discourse tasks provide a
measure of generalization from the treatment to
communication performance in other contexts.
PrelPost staadargized Langgage Assessments

Standardized assessments representative of the two
treatment approaches, functional and cognitive
neuropsychological, were administered to all subjects. A
standardized measure of functional communication, the
Cgmmuaisatiga Astiyities gf Daily Liying test (Second
Edition) (CADL;2) (Holland, Frattali, & Fromm, 1999) was
administered. This test involves role playing, presentation
of pictured stimuli and auditory questions about common
events to measure the ability of adults with aphasia to
respond in an informative and socially appropriate manner.
A three point scoring system is used to measure the

communicative adequacy of the response, regardless of

52

modality ‘-
:est demo:

the test '

‘Ffln '7!
1:50) 1
§.
EWLDStlt
'7 a
:0 Tre

aphasia b
reliabili
rePorted

Sele
Subfleets.
DIOCBSSiE
aim is tc
that can

Q in:

t1“.

e appr:
patient v.
C1escl’ipt

«or

:50 ' ‘1
St 11K

task WE:

S‘ul‘l‘y.

03 ..
sgbte

Mu
‘1‘}; S S‘. 1

modality used in the response. The original version of the
test demonstrated its validity by correlating performance on
the test with direct observation of the proportion of
communication acts appropriately completed in samples of
actual, daily behavior and in family interviews (Holland,
1980). The correlations between these two measures for
noninstitutionalized adults with aphasia was approximately
.70. The CADL was also highly related to other standardized
aphasia batteries. High measures of internal consistency
reliability and interexaminer reliability have also been
reported (above .95 on all measures).

Selected subtests of the PALPA were administered to all
subjects. This test is derived from an information
processing, modular approach to language processing. Its
aim is to provide a collection of standard assessment tools
that can be used to complete an in-depth assessment of the
deficits of a patient's language processing system. The
PALPA includes 60 subtests, from which the examiner selects
the appropriate subtests that will reveal the individual
patient’s specific impairment. Some of these subtests have
descriptive statistics, and others do not.

For the purposes of this study, a few subtests that are
most likely to relate to the skills required in the training
task were chosen to be administered to all subjects in the
study. Auditory comprehension was assessed by performance
on subtest 55, Sentence-Picture Matching: Auditory Version.

This subtest presents four types of sentences, including

53

reversib
gapped,
stinulus
three li
Descript
In addit
a‘ninist
COnprehe
statistj
Ree
IEVEI s:
reQuire<
PlCture
adminis
normal
6830nst
pointin
Tc

subtest
to mam:
famili.‘
and mu
Orally
0f Sti

items
the S

diSSgc

q

C

reversible, non-reversible in both active and passive voice,
gapped, and converse relations. The examiner reads the
stimulus sentence aloud and the subject points to one of
three line drawings that corresponds to the sentence.
Descriptive statistics are provided for control subjects.
In addition, Spoken-Word Picture Matching (subtest 47) was
administered as a measure of single word auditory
comprehension. This subtest also includes descriptive
statistics for control subjects.

Reading comprehension was assessed at the single word
level since written sentence level comprehension is not
required in the targeted functional task. The Written Word—

Picture Matching subtest of the PALPA (subtest 48) was

 

administered, which reports mean performance levels for
normal control subjects. It requires the subject to
demonstrate comprehension of a single written word by
pointing to the matching picture from an array of six.

To assess naming performance, the Picture Naming
subtest was administered. This subtest requires the subject
to name pictured items which are matched for frequency,
familiarity, concreteness, age-of—acquisition, letter length
and number of syllables. The subject is also required to
orally read the written words corresponding to the same set
of stimuli, repeat the same words, and spell these same
items to dictation. The presentation of these tasks using
the same stimuli enables the examiner to observe

dissociations between these language processing skills.

54

.c at.
V1. wt

Finally, all subjects were administered the Bostgn
Namiag_Tsst (BN1) (Kaplan et al, 1983) as a standardized
measure of visual confrontation naming before and after
treatment. This test is a common reference point among many
reports in the literature and has a long history of
application with extended norms (Ferraro et al, 1998;
Ferraro et al, 1997; Flanagan & Jackson, 1997). It is also
commonly used clinically.

grazPost Functional Task

The non-standardized functional task chosen for
pre/post assessment is two forms of a catalog ordering task,
phone ordering and mail ordering. The ability to complete
catalog ordering was chosen as a measure of functional
communication ability for several reasons. Firstly, catalog
ordering is an activity specifically listed by groups of
patients, asked to identify communication situations that
are important to them and also problematic since the onset
of aphasia (Lomas, Pickard, & Mobile, 1987). Independently
being able to do shopping in any way is considered an
instrumental activity of daily living, and appears on
several instruments that measure functional independence in
communication, or activities as they relate to quality of
life (Lomas et a1, 1989; Frattali, 1993). Catalog ordering
can be completed in two modalities, auditory-verbal (phone
ordering) and reading-writing (mail order form). There is a
routinized script for completing catalog ordering, and such

a script can be analyzed in terms of cognitive and

55

adults ‘
rught r
interac
increas-
Finally
orderinq
motivat.

DQ‘

Th4
establi:
the 0rd€
three Sc
business
transcrj
Tri’IScri
E' SUbr
thIEe tr

Opening

x”) and 1
The C108
informat;
be billm

Script d‘

L

rthestec

linguistic components. Sarno (1991) suggests that many
adults with aphasia speak best when on the telephone, which
might relieve some anxiety associated with face-to-face
interactions and also restricts the language choices and
increases language redundancy typically used on the phone.
Finally, most adults have had experience completing catalog
ordering, and independent shopping is likely to be a highly
motivating task for adults in rehabilitation.

Development of the Catalog Ordering Task

The first step in the development of this task was the
establishment of a representative sequence of events within
the ordering script. This was accomplished by transcribing
three scripts taken from three nationally known mail order
businesses. Each catalog ordering phone call was
transcribed, and the sequence of events analyzed.
Transcripts from each of the three calls appear in Appendix
E. Subroutines that did not vary in sequence across the
three transcripts were the Opening and the Closing. The

Opening consisted of an introduction (“thank you for calling

x?) and the polite form of a question.(“may I help you?”).

The Closing sequence included all of the following types of
information across all three transcripts: total amount to
be billed, delivery information, and thanks for calling.
Across the three transcripts, the remainder of the
script during which specific types of information were

requested from the caller varied in sequence. There were a

56

*7
ad

in

total of seven different subroutines which appeared in the
three transcripts, six of which appeared in all three of the
transcripts. These six subroutines and their embedded
subroutines are listed in Table 3. For each subroutine, the
ordinal position in which it appeared in each of the three
transcripts is listed.

Based on the actual sequence of events in the three
transcripts, a prototypical sequence was derived to be used

in the catalog ordering task of this study. The Opening
includes an initial request (“may I help you?”), followed by

the six routines that were common to all three transcripts
in this order: catalog number, phone number, name (last,
first), address (zip code, city/state, street address),
credit card number and expiration date, and ordering the
target item (item number, quantity, color, size). Finally,
there is a Closing sequence that includes total amount to be
billed, delivery time, and thank you.

The range and average sentence length in words for
items transcribed from the actual examples appear in Table
4, and served as a guide for the development of the phone
ordering task. The overall average sentence length across
all three examples was 6.6 words per sentence (range = 3—
20). The average number of sentences per ordering event was
18. The syntactic forms in the actual transcripts were also

counted. The most frequent question types in the actual

transcripts were wh—questions (“what”), embedded questions

(“delivered to home or office”) and ellipticals (“and the

57

 

F“ /-

Tabl

 

 

 

 

 

 

 

 

 

 

Routine Positions

1. Address 3, 4, 3

Zip code 1, 1, 3

City/State 2, 2, 2

Street address 3, 3, 1
2. Cataloginumber 2, 1, 0
.3.Credit card 5, 5, 4

Card number 1, 1, 1

Expiration date 2, 2, 2
l4.Name l, 6, 2

First name 1, 0, 0

Last name 2, 1, 1
5 . Order item 6, 3 , 5

Item number 1, 1, 1

Quantity 2, 2, 0

Color 3, 3, 3

Size 2, 4, 4
6. Phone number 4, 2, 1
7. Place of delivery 0, 0, 6

 

Table 3. Ordinal position of each routine and embedded
subroutine among the actual calls to representative catalog

ordering companies.

58

 

Table
senten

actual

 

 

 

 

 

 

 

 

 

ASL No. of sentences

(tunnel

Example 1 7.07 14
(3-17)

Example 2 7.5 21
(3-20)

Example 3 5.2 19
(3—11)

Overall 6.6 18
(3-20)

Table 4. Average sentence length (ASL) in words per

sentence and number of sentences for each of the three

actual examples of phone ordering.

59

 

7a

a:

a:

.C.

 

Cr

CE

C\

V.-

EC

 

zip code?”). Consequently, these forms were emphasized in

the development of the task.

The phone and written versions of the catalog ordering
task appear in Appendix F. The phone version has an average
sentence length of 6.8 words, close to the average number of
sentences in the sampled transcriptions. Vocabulary used in
the phone version and the mail order form version are
similar but not equivalent since the nature of each of the
two tasks requires slightly different vocabulary. See
Appendix G for a list of the vocabulary items used in the
phone and mail order versions of the task.

Target stimuli for the task were chosen from a familiar
category (clothing) that is frequently ordered from
catalogs. The top ten items mentioned in response to the
category name by normal subjects was used as a basis for
selection of the target items (Battig & Montague, 1969).
From among these ten items, the items that were equivalent
across gender were selected as the six target items (for

example, “dress” and.“tie” were omitted). The six selected

target items were as follows: pants, coat, shirt, hat,
shoes, sweater. These target items were randomly assigned
to each presentation of the task for each subject.

An actual catalog was selected to be used as the
experimental stimulus, and appropriate examples of each of
the six target items were selected. Color names were chosen

for each item according to plausibility, while maintaining

an emphasis on frequently used color names (e.g., “green”

60

was cho:
require:

examirxz

incorre<
created
price.
correspc
Simulate
task. 3
include
number.
Res
fOllowin.
phOne nm
{COlOrs'
m1111391“, j
these Cat
Chronic a
(30059163

to haVe 6

Val;

-L¢(

was chosen over “taupe”). The catalog ordering task

requires the presentation of an incorrect item by the
examiner, and the six target items were assigned in random
order to be used as the incorrect item for each target. The
incorrect price required in the catalog ordering script was
created for each item by adding 50% of the actual listed
price. A table listing the target items and their
corresponding colors and prices is presented in Appendix H.
Simulated credit cards were also created for use in the
task. Twelve digit credit card numbers were created to
include each of the digits 0 through 9 in each credit card
number.

Responses generated during the task represent the
following categories: personal information (name, address,
phone number), object names (clothing items), adjectives
(colors, sizes), number and letter names (credit card

number, item number), and action (“order”). Exemplars from

these categories have been assessed among adults with
chronic aphasia in both comprehension and expression tasks
(Goodglass et al, 1969). Adults with nonfluent aphasia tend
to have equal difficulty in naming across all categories.
Validity of the task was assessed by comparing
performance on the phone ordering simulation to actual
performance on an audiotaped phone call to a catalog. Three
adults with aphasia were recruited to call actual catalog

companies and place an order at the request of the examiner.

61

(3

'(‘1

’(3

r)

'(1

(f)

'(‘1

l

(I)

n)

’4-

r1:

The descriptive information for these three subjects is
presented in Table 5.

These phone calls were audiorecorded and transcribed
(see Appendix I). Performance on the actual call was
compared to the first performance of the simulated phone
version task in the quiet condition. Accurate responses
were totaled and errors characterized. The error types

listed in Table 5 are failures to communicate the complete

information requested, including no responses or saying “I
don’t know”, related errors which included providing all or

part of an address when item number or phone number was
requested, and unrelated errors would have included the
production of words or phrases that were not related to the
catalog ordering task. As suggested in Table 6, performance
levels and nature of errors occurring during the actual
phone calls were similar to those of the same subjects
during the phone ordering task. Subject 1 was much more
successful in the simulated phone task than in the actual
phone ordering. One reason for this was because she chose
to open and use a communication notebook as a self—cueing
device during the simulated catalog ordering task, whereas
she did not choose to use any of these cues during the
actual phone call even though she had equal opportunity to
do so across both situations. This is particularly
interesting since she did the simulated catalog ordering
first prior to performing the actual phone call. Secondly,

her failure to communicate her phone number during the

62

Sub

Table E

aphasia

 

 

 

 

 

 

 

 

 

Subject Age (years) Aphasia type- Aphasia
BDAE severity
Classification (BDAE Severity
Rating)
1 52 Broca’s 3/5
2 23 Broca’s 3/5
3 13 Anomic 4/5
Table 5. Subject description for the three subjects with

aphasia placing catalog ordering phone calls.

63

 

kl'rr t‘To‘F'TC’C tI‘FIlfilt

'

'1'"

r'lll‘l‘

 

 

 

.xmou mcHHoUuo monumo poucHQEHm mnu GH paw mHHoo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ho5uom CH onanc nuHB muomhndm money man HOu mommy House pew hocusoom HHon>o .u OHAUB
o o o wooH o o o NOOH m
o N o wmm o o N wmm N
o m N wa o N m wbH H
poumH UmuoH noun
umHGD pouoHom deH Hum >omusoo< nous: pmumem mHDHHmm homusuonm Inn—um
momma Honum HHmnm>O momma Honum HHMHw>O
lads unauovuoluoHdudU HHIO Hanan:

 

 

 

 

64

actual call, which was the first item asked, caused a fatal
error in the actual ordering event, and she aborted the
phone call. In the simulated task, her first error was not
fatal to the completion of the entire sequence, which
provided her with more opportunities later in the task to be
successful. The type of errors she made in both
circumstances was similar.

Concurrent Task Condition

Both forms of the functional task were performed under
two conditions, one of which was designed to assess
functional abilities under cognitive load. During one
condition subjects performed the functional task assessment
in a quiet environment. The subjects were also required to
perform the task in a concurrent task condition, during
which a secondary reaction time task was presented to the
subjects. Significant differences in lexical decision and
picture description performance have been reported among
aphasic subjects under quiet, focused attention, and divided
attention conditions, similar to the one proposed here
(Murray et a1, 1997; Murray et al, 1998).

A 1000 Hz tone was computer-generated using Sound
Blaster software on a Power Macintosh computer. PsyScope
software (Cohen et a1, 1993) was used to set up the reaction
time task. The stimulus duration was set at 150 msec and
the response interval at 3000 msec. The response interval
was determined by reviewing reaction times reported in

previous literature for performances of adults with aphasia

65

fi‘

in divided attention tasks (Murray et al, 1997; Murray et
al, 1998), and then creating a response interval long enough
to accommodate these times. Randomized intertrial intervals
(ITI) were set to range between 5000 msec and 11000 msec,
with an average ITI of 8000 msec. A relatively long ITI was
chosen based on the experiences of Murray et al (1998) as
well as on empirical reports of response latency in picture
naming for adults with aphasia. In Murray’s experiment, ITI
was set at 1000 msec, and this created a task that was
extremely difficult even for subjects with a mild aphasia
(Murray, personal communication). Consequently, a longer ITI
was chosen for the present experiment to allow subjects with
a broader range of severity levels to participate, and to
accommodate the more complex catalog ordering simulation
that was the target. Furthermore, adults with aphasia have
been observed to require 2—5 seconds in response time for a
simple picture naming task (Mills et al, 1979). Therefore,
5000 msec was chosen as the lowest ITI. A single switch
mounted on a metal foot stand was connected as an input
device. This response option was chosen since many of the
subjects had residual hemiparesis, with limited use of the
upper extremities. Instructions to the subject during the
concurrent task condition emphasized the importance of
completing the ordering task while responding to the
presented tones as quickly as possible. These instructions

appear in Appendix F.

66

Scoring for the Catalog Ordering Task
A three point scale, similar to the one used in the
CADL-Z, was developed and applied to scoring for accuracy of

both phone and written versions of the catalog ordering

task. A score of‘iW on the three point scale was given for

complete, accurate responses. A score of “1”‘was assigned
when the response was related to the stimulus item or the
target, or was incomplete. A score of “0”*was assigned if

the response was unrelated to the item, inaccurate, or if
there was no response. A list of the specific criteria, and
examples for each item in the catalog ordering task, are
given in Appendix J. Total duration in minutes and seconds
for each run of the catalog ordering task was also recorded,
as a measure of general efficiency during task performance.

Pr Me r of i vali it

Another pre/post measure was the Cgmmanisatiye

Bffsgtiysasss_ladsx (CBII) (Lomas et a1, 1989) which
provided a measure of the perceptions of caregivers at the
beginning and end of the treatment program. The CB3; has 16

items to which the caregiver rates the person with aphasia’s
performance on a visual analogue scale, ranging from.“as
able as before stroke” to “unable to do”. At post-testing,

caregivers mark their current perceptions on the pretest
form.with previous markings visible. This administration

procedure has been established as a valid and reliable

67

method. The CETI score is the change score between the

pre/post test administrations.

Pro/Post Discourse Tasks

Spoken discourse is an important tool used in many
different daily living situations, and consequently it is
critical to determine the relationship between discourse
outcomes in aphasia therapy and its relationship, if any,
with other functional task performances. The discourse tasks
are designed to assess the subjects’ ability to apply
targeted communication skills in different situations that
relate to the functional communication task to various
degrees. Three discourse types were collected, expository,
procedural and narrative discourse. Expository discourse was
collected pre/post treatment using the Cookie Theft picture
from the BDAE, or one of a related set of pictures that has
also been developed for this purpose (Craig et al, 1993).
Each subject was asked to tell the examiner ‘everything you
see going on in this picture’. A procedural discourse
sample was elicited by asking the subject to recount the
steps for accomplishing a set of tasks that were either
similar to the catalog ordering task, or were topics
commonly used in the collection of procedural discourse of
adults with aphasia. The subjects were asked to recount how
to order something from a catalog, both before and after the
treatment program. Subjects were also asked to describe the
steps to ordering a pizza, since the task sequence should be

similar to catalog ordering but the required vocabulary will

68

differ. Samples from two additional topics which are common
to many analyses of procedural discourse in aphasia, how to
make toast and jelly, and how to get groceries from the
market, were also collected (Li et al, 1995; Ulatowska et
al, 1983; Terrell & Ripich, 1989).

The narrative sample was elicited by asking the
subjects to retell the Cinderella story. The retelling of
this story has been used as a standard procedure in aphasia
discourse research (Saffran, Berndt & Schwartz, 1989).

Subjects were asked to tell the story “as though you are

telling it to your children/grandchildren”. All of these

samples were audiorecorded for later transcription and

analysis.

Order of Presentation

Order of presentation of the pre/post assessment
measures were randomized across subjects as much as possible
within certain constraints. During both pretesting and
posttesting, all subjects performed the discourse tasks
first, followed by all other assessment measures. Within the
catalog ordering task, presentation of the phone and written
versions in quiet and concurrent task conditions were

randomized for each testing period for each subject.

Treatment Procedures
Subjects who met all eligibility criteria were assigned
to one of two treatment approaches. The two treatment types

were commonly used aphasia treatment approaches that are

69

representative of common clinical practice in addition to
being well-represented in the aphasia treatment research
literature. The research treatment sessions were provided
as part of the therapeutic activities that the subjects were
receiving in their program at the University of Michigan
Aphasia Program. All treatment activities for each subject
were representative of the assigned treatment type.

Eligible, consenting subjects underwent all pretesting
as described above during the first week. For the next five
weeks, the subjects participated in therapy sessions that
were based on either the functional or cognitive
neuropsychological treatment approaches. Subjects
participated in 4-5 naming treatment sessions weekly which
were derived from the catalog ordering task, in addition to
15-20 additional treatment sessions that made up the total
program. Catalog-based naming treatment sessions were all
provided by one of two trained therapists, including the
experimenter. All subjects entered into the study had naming
as goal in treatment.

The other 15-20 weekly treatment sessions were oriented
to the assigned treatment. Each subject had a total of
three therapists, and these therapists were trained on the
characteristics of the two treatment approaches. All
treatment goals and activities were audited by the
experimenter to ensure compliance with the characteristics
of the targeted treatment approaches. The parameters of the

two treatment approaches described in Table 1 were used as

70

guidelines. A description of the treatment goals and
strategies for each subject are listed in Table 7.

Due to the coordinated nature of the University of
Michigan’s Aphasia Program, it was possible to ensure that
all therapists were in compliance with the targeted
treatment approach by reviewing documentation such as
progress notes and by direct observation of treatment
sessions. The residential component of the program also
provided an element of control for the activities that
subjects participated in outside of the experiment. All
subjects were exposed to the same living situation and
similar social activities during the course of the
experiment. This element of environmental control is
extremely unusual and advantageous for a treatment research

study.

gggaitivg Nggtopsxgholggigal Ttgatment

The cognitive neuropsychological approach assumes two
main potential sources for naming impairments, semantic and
phonological. Consequently, two training procedures with
their aim to address each of the two levels of impairment
were devised for use in the CN treatment. Subjects who were
assigned to the CN treatment group underwent additional
testing to assess the level of function that was most
severely impaired, and to identify input and output

modalitiesof strength and weakness. The most severely

71

Table 7. Description of additional treatment goals and

strategies for each subject.

72

Table 7. Description of additional treatment goals.

 

 

 

Goals/Strategies
Subjects
Functionally
trained
subjects.
F1 Naming and conversational skills using

multimodality strategies; auditory
comprehension of complex questions and
sentences in conversational contexts

F2 Oral and written production of functional
vocabulary words; use of multiple
modalities to communicate messages; reading
comprehension of single word functional
vocabulary

F3 Naming in conversation and with the use of
categorical and semantic cueing;
comprehension of money and time concepts;
writingpersonally relevant single words
F4 Verb naming and short phrase production in
conversation; auditory comprehension of
complex sentences in conversation; writing
personally relevant single words

F5 Verb naming in conversation using multiple
modality compensatory strategies; writing
short phrases using compensatory
strategies; reading comprehension of short
paragraphs

F6 Naming and use of multimodality
compensatory strategies in conversation;
writinggpersonally relevant phrases

 

 

 

 

 

 

 

 

 

 

Table 7 (cont'd)

 

 

 

 

 

 

 

 

 

Goals/Strategies

Subjects

CN trained

subjects.

CNl Generating oral and written phrases using
HELPSS (Helm-Estabrooks, 1981);
comprehension of when/where questions;
reading comprehension of paragraphs

CN2 Visual Action Therapy; writing personal
information in a drill format; naming and
recognizing gestures

CN3 Oral naming given phonemic and rhyme cues;
single word reading comprehension; phoneme
to grapheme matching; writing words and
nonwords to dictation

CN4 Oral naming given semantic cueing; writing
single words and phrases; reading
comprehension of paragraphs

CNS Oral and written naming of single words
given semantic and phonemic cues; auditory
comprehension of commands; silent and oral
reading of words and word combinations

CN6 Oral and written naming given semantic and

 

phonemic cues; reading comprehension of
sentences; formulating word strings to
picture description

 

74

 

impaired area was addressed by selecting either the semantic
or phonological route to treatment.

Assigned subjects were assessed using the BABBA
subtests to identify details of their linguistic profiles.
Subjects who demonstrated a predominant semantic impairment
showed lower than normal performance levels on single word
comprehension and production subtests. Errors consisted
primarily of the selection of semantically related
distractors. Subjects with predominant phonological
impairment exhibited few semantic errors but made sound
selection errors in oral reading and naming tasks.

Table 8 lists the assessments and performance levels
for each CN subject, which determined the nature of the
naming impairment in each case. Subjects CN1, CNS, and CN6
had a similar profile. All three of these subjects had
relatively good semantic processing of auditory and written
words, as measured by their ability to match words and
pictures. All three had poor oral naming performance,
including poor repetition and oral reading abilities. This
profile implicates the lexical phonological system as an
origin for the oral naming disorder, with minimal
involvement of the semantic system. These three subjects
were all administered the same cueing hierarchy during
treatment.

Subject CN2 had a severe impairment of the both the
semantic and phonologic systems. This subject at pretesting

was unable to sort the pictured stimulus items into semantic

75

Table 8. Description of assessments and performances used
to determine the nature of the naming disorder for each CN

subject.

76

 

 

 

 

 

 

 

 

Table 8. Assessment information.
Assessment/Performance
Subject
CNl Auditory word-picture matching = 90%
Written word-picture matching = 95%
Written picture naming = 45%
Oral picture naming, baseline vocabulary = 60%
Oral reading = 66%
CN2 Auditory word—picture matching = 45%
Written word-picture matching = 45%
Written picture naming = 0
Oral picture naming, baseline vocabulary = 0
Oral reading 2 0
Category sorting ofgpictured items = 0
CN3 Auditory word-picture matching = 95%
Written word-picture matching = 73%
Written picture naming, baseline vocabulary =
100%
Oral picture naming, baseline vocabulary = 33%
Oral reading = 18%
Oral spelling = 13%
Phoneme-grapheme matching = 73%
Spoken word — initial letter matching= 13%
CN4 Auditory word-picture matching = 98%
Written word-picture matching = 100%
Written picture naming = 95%
Oral picture naming, baseline vocabulary = 95%
Oral reading = 95%
Naming to auditory definition, baseline
vocabulary = 17%
Sentence production to pictures = 83%
CNS Auditory word—picture matching = 80%
Written word—picture matching = 88%
Written picture naming = 3%
Oral picture naming, baseline vocabulary = 46%
Oral reading = 13%
CN6 Auditory word-picture matching = 90%
Written word—picture matching = 93%
Written picture naming = 10%
Oral picture naming, baseline vocabulary = 65%

 

Oral reading = 85%

 

77

 

 

 

 

categories (e.g., numbers, letters, clothing). This subject
was administered a treatment program that emphasized
semantic processing. Subject CN3 had a relatively good
semantic system, and was able to correctly name all the
baseline picture items with written words at pretest. He
was unable to orally read any of the items he had written,
nor was he able to orally name these pictured items. His
profile was consistent with a downstream phonological
deficit which affected his ability to match phonemes and
graphemes singly or in words. His treatment focused on
phonological processing and oral reading.

Subject CN4 was relatively good at single word auditory
and written comprehension, and naming pictured items. In
spite of these strengths, he was exceptionally poor at
naming given an auditory definition. His conversational
output at pretest was sparse and limited to single words and
short phrases in spite of his relatively good naming skills.
His sentence production skills were also relatively good,
given a pictured stimulus item. Overall, his performance
was relatively good given visual, pictured input, and poor
under conditions of auditory input. This subject displayed
a specific semantic naming deficit related to the auditory
input channel. Therefore, his treatment focused on naming
the target vocabulary given auditory definitions.

The stimuli were 54 items derived from semantic
categories associated with the catalog ordering task. The

semantic categories included: number names, color names,

78

clothing names, letter names, and fruit names. The category

“fruit” was selected because these items could plausibly be

ordered from a catalog but are not included as targets in
the catalog ordering task. Items in each category were
matched for word frequency. A list of the S4 vocabulary
items is given in Appendix K.

Baseline sessions consisted of the first three sessions
of treatment. During each baseline session, all of the S4
stimuli items were presented in random order and the subject
asked to name each item upon picture presentation. Items
missed during all three baseline trials were selected for
training and probes. Half of the consistently missed items
were targeted for treatment, and the other half were
untrained. The items to be trained were divided into two
training sets.

At the beginning of each treatment session after the
baseline sessions were completed, half of the training set
items were presented without cues. Half of the items were
alternated each day to create two different treatment
probes. Performance on this probe was used to chart
progress during training as well as to determine whether the
subject had achieved criterion. Following the probe,
treatment began when the clinician presented a picture
stimulus randomly chosen from the training set and applied
the cueing hierarchy (see Table 9) when the subject
attempted to orally name the pictured item. The clinician

and subject continued with the cueing hierarchy for each

79

 

 

 

 

 

 

 

 

Subject Cueing hierarchy
CNl 1) Semantic cue
2) Phonemic cue
3) Repetition
CN2 1) Sort pictured stimuli by category
2) Semantic cue
3) Phonemic cue
4) Repetition
CN3 1) Match graphemes to phonemes
2) Phonemic cue
3) Repetition
CN4 1) Match definition to name
2) Describe by semantic feature
3) Repetition
CNS 1) Semantic cue
2) Phonemic cue
3) Repetition
CN6 1) Semantic cue
2) Phonemic cue
3) Repetition

 

 

Table 9. Cueing hierarchy for each CN subject.

80

 

stimulus item in the trained set. The entire set was
repeated with the cueing hierarchy. The probe and the two
passes through the training stimuli with the cueing
hierarchy constituted one training session. A sample
training data form is shown in Appendix L. This procedure
continued for four—to-five sessions per week until a 90%
criterion was achieved. Once criterion was achieved, the
second training set was presented in the same way. Training
on this set of items continued until criterion was achieved
or until the subject’s participation had reached five weeks.
Once the training period ended, the complete post-testing
protocol was administered to measure naming performance on
both trained and untrained items. The treatment cueing
hierarchy for each CN subject is presented in Table 9. The
trained and untrained items for each subject in the CN
treatment are listed in Table 10.

CN1, CNS, and CN6 each received the same cueing
hierarchy, which reflected their similar deficit patterns
(Greenwald et a1, 1995). CN2 was asked to sort training
items into semantic categories, and then these items were
named using semantic and phonemic cues (Howard et a1, 1985;
Le Dorze et a1, 1994). CN3 was administered a phonological
therapy (Howard et al, 1985; Le Dorze & Pitts, 1995) with an
emphasis on oral reading (Cherney, 1995). CN4’s treatment

focused on a semantic procedure, requiring the naming of

81

Table 10. List of trained and untrained items for each
subject in the CN treatment. * = items required for
performance in the phone version of the catalog ordering

task.

82

 

 

 

 

 

 

 

 

 

Table 10. List of trained and untrained items.
Subject Trained Items Untrained
Items
Set A Set B

cui Y, coat*, J, V*, U, gray, Q,
pineapple, R, purple, G, red*, M*,
blue*, T, O K*, banana, H lemon, Skirt,
(7) (7) K*, peach, W.

D*, P, I*
(13)

CN2 Blue*, pink, purple, Letter names
black*, grey, red*, dress, Number names
coat*, skirt, shirt*, (36)
pants*, socks, apple,
peach, pear, banana
pineapple, (9)
orange
(9)

CN3 Purple, red*, Blue*, pink, Brown*, grey,
shirt*, socks, peach, skirt, pear,
pineapple, banana apple
orange (5) (5)

(5)

CN4 Black*, Pink, shirt*, Blue*, credit
city*, size*, red*, phone card*, coat*,
dress, number*, catalog*,
apartment, skirt gray, socks,
(5) (6) state*,

purple,
address*,
pants*
(11)

CNS Black, one, Pink, zero, Purple, F, I,
D, K, R, skirt, pear, M, P, S, U,
coat, peach, Y, N, T, H V, W, Y, 2,
apple (8) dress,

(8) pineapple
(13)

CN6 Gray, D, F, E, G, I, K, O, P, Q, R,
H, M, J L, N S, T
(6) (6) (6)

 

 

 

 

83

 

items given auditory definitions, with a semantic map and
semantic features as cueing devices (Lowell et al, 1995).

{tactignal Treatment

The functional approach assumes that contextual
influences and self-generation of cues are powerful
modifiers of behavior within specific contexts.
Consequently, the functional treatment emphasized practice
within the targeted activity and the application of self-
generated strategies. Subjects who were assigned to the
functional treatment group practiced targeted communication
strategies within the actual catalog ordering task.

The materials for this treatment included various
catalogs, a phone, practice credit cards, pen, and paper.
The same phone ordering script used in the pre/post
assessment was used for the training sessions. This is
consistent with the functional emphasis on practice in
relevant contexts. The treatment probes consisted of half
of the trained items administered without cueing, so that
there were a total of two probes which were alternated. A
baseline performance for this treatment was already
established during the pretesting of the catalog ordering
task. A list of the items that a subject needed to be able
to produce to perform the catalog ordering task during
treatment is given in Appendix M.

Only half of the items used for pre/post assessment on
the catalog ordering task were used for training. These

trained items in the semantic categories of “colors” and

84

“clothing names” were also included in the CN treatment

baseline. Trained items that were required in the catalog
ordering task were compared to untrained items at posttest
to determine whether there was an effect of training on the
vocabulary produced.

After the treatment probe was administered at the
beginning of each treatment session, the clinician and
subject worked through the training script. For each item
that was difficult for the subject, the clinician used a
problem-solving approach to assist the subject in generating
a successful strategy for communicating the information in
the appropriate modality for the task. Each training
session consisted of one pass through the phone version with
cueing and practice of strategies, after the training probe.
A sample data recording form for the functional treatment is
shown in Appendix N. Criterion performance of 90% of
information conveyed appropriately and accurately was
required for completion of training. All of the training
sessions were videorecorded for later analysis. Training
continued until criterion was reached or until the five week
training period ended.

The specific strategies used for each subject are
listed in Table 11. All of the strategies were designed to
achieve the goal of completing the phone ordering task.
Whenever possible, oral reading was used as a strategy to

accommodate specific oral naming deficits, like difficulties

85

 

Subject Strategies

 

F1 Oral reading of keywords in
communication notebook

Subvocal counting; using fingers
during counting

Use of electronic device to cue key
words

 

F2 Oral reading of keywords in
communication notebook

Subvocal counting

Use of electronic device to cue key
words

 

F3 Oral reading of keywords in
communication notebook

Use electronic device to name
letters/oral spelling

 

F4 Oral reading of keywords
Write out information needed in
advance

 

F5 Oral reading of keywords in
communication notebook

 

 

F6 Oral reading of number words
Use of electronic device to name
letters

 

 

 

Table 11. Treatment strategies used for each functionally

trained subject for the catalog ordering task.

86

with number or letter naming or saying personal information.
When this was not an option, other strategies like subvocal
counting for number naming, or use of an electronic device
with speech output were used. This required the subject to
identify the specific needs for a particular response, and
choose the correct compensatory strategy. Each strategy, in
turn, could have multiple steps, in particular use of the
electronic device. In this case, items needed were stored
in a personalized list in advance, and the subject was
required to perform a series of keypresses that selected the
target item and produced speech output. Therefore, each
functionally trained subject was learning a compensatory
strategy that might require a few steps for each response
item, and learned to use these strategies to produce

accurate and efficient responses.

Reliability

Transcription and scoring reliability were
determined by having 12% of all of the discourse samples,
catalog ordering, and standardized testing re-transcribed
and coded by an independent observer. This observer was a
certified speech/language pathologist unfamiliar with the
study or its goals, who was blinded to test time
information. Samples were randomly selected from the data
corpus. Transcription reliability was calculated with a
point-to-point comparison at the word level by dividing the
number of agreements by the number of agreements plus

disagreements. The transcription reliability was 96.6%.

87

Scoring reliability was also calculated by point-to-point

comparison, and this yielded a scoring reliability of 97.7%L

88

Chapter 3

RESULTS

Comparability of the Two Groups: Demographics

Eligible, consenting subjects were randomly assigned to
one of two treatment groups, cognitive neuropsychological
(CN) or functional (F). The demographic characteristics of
the two subject groups were compared. The two subject
groups were similar in average age (CN mean = 50.3 years, F

mean = 51.7 years), time post onset (CN mean = 26.3, F mean

26.8), socioeconomic status ratings (CN mean = 1.3, F mean
= 2.3) and severity rating (CN mean = 1.8, F mean = 2.5).
The CN group was slightly higher in SES and the F group was
slightly less severe as measured by the Severity Rating of

the BDAE. Details for each subject are provided in Table 2.

Comparability of the Two Groups: Cognitive Battery

Each subject was administered a cognitive battery.
Results of the cognitive testing for each subject are
offered in Table 12. All subjects performed well on visual
cancellation (overall mean = 98.4) and recognition memory
(overall mean = 95.3). All but one subject (F1) performed
well on a measure of nonverbal problem solving, the BQBN.

Subjects ranged in performance from 80% to 100% on this

89

Performance of subjects on the cognitive battery.

Table 12.

suds
burnupoa

nods
Butqonsm
3161a

ands no
rarqedez
3151a

-Tuboaex
Azonrpns

I
eqzeauon

ts
nqdeozsa
-tsnsrA

FUNCTION-

ALLY

Subjects

 

90

 

' No. of categories learned on the Wi onsi Card rtjn Test (Grant & Berg, 1993).

 

 

 

793
:2: $23 3.5 :13 :.3 3.3 3.1: 3.93 3.3 35-!
h.N v.fl v m.N «.mm mm 04.0 m.mm $.mm ago
Amt: AN.NV AN.MV Aw.NV «RAH: a.mv club 49.1: Ac.vu H.U.uv
no." N.N N.‘ m.N o.mm Mina om mm mm 50:

 

 

 

 

 

 

 

 

 

 

 

APE—c3 NH 035—.

91

measure, except Fl, who achieved a score of 41%. This is
consistent with Fl’s equally poor achievement on the WCSI,
during which he learned no categories. This poor learning
and cognitive performance was not demonstrated by any other
subject. Fl’s language performance, discussed in more
detail later, was comparable to the other subjects and was
consistent with the diagnosis of nonfluent aphasia. Subjects
also performed relatively well overall on the visual-
perceptual assessment, comprised of selected subtests of the
B113 (overall mean = 89%). Both treatment groups had
comparable performance on the assessment of nonverbal
auditory recognition (overall mean = 95.8%).

Digit repetition span was similar across the two
treatment groups. Three subjects were unable to perform the
task because they could not repeat numbers, and at least one
of these subjects was assigned to each group. Similarly,
one subject in each of the treatment groups was able to
perform within normal limits on the digit repetition span
(span = 7 digits). A wide range of performance was also
observed on the digit matching span task. Five subjects
overall achieved a normal span on this task (span = 6-7),
and three of these subjects were assigned to the CN group.
The remaining subjects could not perform the task (two
subjects), or achieved low scores of 2—3 (five subjects).

Performances on the pointing span task ranged from 0 to
6, and all of these performances were notably below normal

levels of performance (Kay, Lesser, & Coltheart, 1992). The

92

good recognition memory performance of all the subjects in
contrast to the relatively poor recall performance as
evidenced by scores on the digit repetition and pointing
span tasks is a typical profile for anterior aphasic adults
(Beeson, 1993).

The generally poor category learning performance of all
of the subjects, with the exception of CN1, is consistent
with average performances reported for adults with frontal
lobe injury (average number of categories learned for adults
with frontal lobe injury = 3.46)(Grant & Berg, 1993). F1,
however, is notable for being totally unable to learn the
task at all during the entire 128 trial period, and CN1
performed two standard deviations above the mean for frontal
lobe injured adults. Although CNl’s language performance
was consistent with a nonfluent aphasia, CT documentation
suggested a temporo—parietal lesion rather than a frontal
lesion. It is tempting to suggest that his more posterior
lesion might partially explain his different WCSI
performance, but three other subjects also had documented
temporo-parietal area lesions (CN3, CNS, and CN6) and none
of these subjects performed better than the subjects with
more anterior lesions. CN1 was reported to be highly gifted
prior to his stroke by his wife, but this may not be
significantly different than the other subjects, most of
whom were college educated with high level occupations.

All portions of the cognitive battery were subjected to

correlational analysis. The RCPM and the WCBI were highly

93

correlated (r = 0.76, p< 0.01). The digit repetition span
task of the BABBA was correlated with the digit matching
span task (r = 0.75, p<0.01) and with the pointing span task
(r = 0.75, p< 0.01). There were no other relationships
between components of the cognitive battery. This suggests
that the assessment measures that comprised the cognitive
battery were tapping relatively separable skills, with the
exception of problem-solving/new learning as measured by the
WEST and BCBM, and the three span tasks.

All subjects performed relatively well on visual
cancellation, recognition memory, visual-perceptual
assessment, and nonverbal auditory recognition. The
cognitive assessments that detected differences between the
subjects are those that measure nonverbal problem-solving,
executive function, and memory span. This suggests that
basic processing of stimuli, like that required in the
visual cancellation, visual-perceptual, immediate
recognition memory, and nonverbal auditory recognition are
relative strengths in these twelve subjects. However, there
was more variability between subjects among the higher level
cognitive functions.

Because the BCBN is part of the Western Aphasia Battery
(Kertesz, 1980), it has been routinely used in the
assessment of subjects with aphasia. The BCBN has
occasionally been observed to be related to auditory
comprehension abilities, although it is not related to other

linguistic abilities (Kertesz, 1988). Correlations between

94

single word and sentence level auditory comprehension and
the BQBN were run to determine whether such a relationship
existed in the current data set. Pretest oral and written
naming abilities, as well as functional communication, were
also analyzed for potential relationships to performance on
BCBN. There were no relationships between any of the
pretest language abilities and the BCBN.

Consequently, inspection of the BCBN, WCBI, and
digit/pointing span tasks together yields the distinctive
cognitive profiles for each subject. Among the functionally
trained subjects, F1, F2, and F6 displayed the lowest
cognitive profile (Table 12). F1 had the lowest performance
on BQBN and WCBT of all of the subjects, and F2 and F6
showed the next lowest performances on WCBT. These three
subjects also had the lowest performances on digit/pointing
span tasks. The other three functionally trained subjects
had moderate levels of performance on WEST, BCBN, and digit
span tasks.

Among the CN trained subjects, CN2 and CN3 had the most
severe cognitive profiles based on their performances on the
WCSI, RCBN, and digit/pointing span tasks (Table 12). CN1
had the best cognitive profile overall for all twelve
subjects in the study. CN2 and CNS sustained deep lesions to

the frontal—parietal-temporal regions.

Comparability of the Two Groups: Language Testing

Pretest performances of standardized language testing,

functional communication measures, and reaction times during

95

concurrent task performance were compared between the two
groups. All subjects were relatively good at understanding
single written words (overall mean = 88.3), with the
exception of CN2 (score = 45%). This subject was equally
poor at spoken word-picture matching and sentence-picture
matching subtests of the BABBA (35% and 0, respectively).
CN2 was the most linguistically impaired of all of the
subjects. On the BBAB, which measures auditory
comprehension primarily through spoken word picture matching
tasks, CN2 performed at the 45th percentile, qualifying him
as a nonfluent, Broca’s-type aphasia according to the Rating
Profile, and meeting the eligibility criteria of this study.
His performance on the standardized BABBA subtests, which
are controlled for word frequency, imageability, and age of
acquisition, were notably worse than his performance on the
BBAB. This is consistent with his semantic impairment. The
remaining subjects performed well for spoken word—picture
matching, and demonstrated decreased performance overall for
sentence comprehension. Specific test scores for each
subject are provided in Table 13.

The subjects overall demonstrated variable performance
on the naming assessments. Generally, the subjects scored
higher on the Spoken Picture Naming subtest of the BABBA
than on the BN1. This is likely due to the controlled
stimulus characteristics of the BABBA, including controls

for word frequency, imageability, and age of acquisition.

96

GonGonoHQEou mo mwudmowfi poNHUHmpcmum no Ammmmudmouwe

.mQDOHm uowwndm 03» OED How

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

OOH anoAOHQEOU hoquu—d

 

nadvdom

 

mwoGMEHOMHwQ umoumum .MH OHAUB

Am.va Hw.mHv HN.MHO Am.mHO A.p.mv

m.hw m.om v.2. an Egg

3.va 35.: 2.113 3.08 36.3

mm m.mm h.mw N.Nm zsaz

Ob ON O. mm 020

mm Om Ow mm mzu

mm ON mw OOH #20

pm Wm Ow Mb MZU

O mm mV me NZU

OO om Ow mm Hzo

.meomnmsm

3545. 20

Am.mv lle.NM Am.mv Hm.mv A.U.mv

h.mh m.mm h.mb m.mm cows

OOH OOH mm OOH Om

ma OOH ox mm mm

mp OOH OO OOH vb

mm mm or ON mm

mm OOH Nb mi mm

Mb OOH ma Immll Hm

flmqflm tqum «mama

I 95:00.02 I unHAoudE HnOHHufloOHOGO .. unwnoudz .2833 $559

OHSUOHm ounUOHm QOHuaOAOHQEOU unfluUHm ESEEEBESL
Ioofiounom IUHOB noxonm huouH65¢ ”dam :UHOS nouuwuk

.mmwmmmmmmmmmm1 muoohnnm

 

 

97

In contrast, the BNT samples picture naming abilities from

high frequency words (e.g., “tree”) to low frequency words

(“abacus”). The specific performances of each subject are

listed in Table 14. Five of the twelve subjects were
severely impaired in oral picture naming, and three of these
were assigned to the CN group (CN2, CN3, and CNS) and two to
the functional group (F2, F6). Three subjects were
relatively good at picture naming (F4, F5, CN4).

There was a similarly wide range of abilities in

written picture naming, as measured by performance on that

subtest of the BABPA. Subjects who were poor at oral naming
were not necessarily poor at written naming (e.g., CN3), and

subjects who were poor at written naming were not
necessarily poor at oral naming (Fl, CN6). However, three
subjects were quite poor at both oral and written naming
(F2, CN2, and CNS). These patterns of impairment are
indicative of the nature of the language breakdown and were
taken into consideration when designing the subsequent
treatments.

Assessment of functional communication on the
standardized CADL—Z test demonstrated relatively preserved
functional communication abilities, despite poor naming and
comprehension performances on decontextualized assessments.
All but one subject (CN2) achieved at least a score of 70%
or better (overall mean = 78%, range = 71-93%). CN2, who
achieved a pretest CABL—z score of 17%, also performed the

lowest of all of the subjects on auditory comprehension.

98

.mQSOHm pomwndm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Quon How OGHEMG mo monummofi UONHpHmpcmum no Emmanunouuwmv moo:MEHouHmQ umoumum .vH OHAHB
HH.mmv Hw.mmv Aw.OmO A.n.mO
H.Nv H.Hm w.mm ado: HHduo>o
3.62 33 3.va H66.
o.wm m.He h.Hm new:
oumuwwoz OH OO mm mzu
oumuoooz m O m mzo
oumuowoz mm mm mm #20
30H me OH OH ”20
30H O O O N20
.85 mm ms mm :6
.meomempm
$5.28 20
Am.ovv HH.hmO Am.bNV H.U.uv
N.mv m.om m.mv BEE
30H OO an O om
oumuwvoz mm 2. mm mm
mucnovoz ON Mfiw rr Wm
wumuwvoz mm a. mm mm
30a m O W” mm
30H O NH He Hm
.meomnmom omzH<me
(minim. 3 5442050sz
ununHusm I undadz I usfiadz

OHuanuqu OHSHDHQ ounuuHm unoa muoofinfim

|0>HOHnuou nauuHHS noxonm unfifidz nounom

 

 

99

This may have impeded his ability to understand the
questions and role—playing required in this test. Summaries
of each subject’s performance on functional communication
pretesting are given in Table 15. Pretest performance on
CABB;2 was related to performances on the phone version of
catalog ordering in quiet (r = 0.73, p<0.01) and in the
concurrent task (r = 0.72, p<0.01). CADL-Z performance was
also related to pretest accuracy on the written version in
quiet (r = 0.85, p<0.01) and with the concurrent task (r =
0.67, p<0.01). These correlations show that the two
functional communication assessments were related to each
other. In addition, the CADL-Z performances were related to

a number of modality-specific tasks, including single word

auditory comprehension (r 0.90, p<0.01), sentence level

auditory comprehension (r 0.87, p<0.01, and single word
reading comprehension (r = 0.96, p<0.01), and naming as
measured by BNT to a lesser extent (r = 0.60, p<0.05). This
suggests that the CADL—2 measures a broad range of
communication skills.

The phone version of the catalog ordering task, being a
functional task requiring a variety of skills, was also
related to a number of other language measurements. The
phone ordering task was less reliant on auditory
comprehension skills than CADL-Z, with a moderate

correlation between pretest phone task accuracy and single

word auditory comprehension (r = 0.62, p<0.05). There was

100

.mmdoum DothSm QDOQ How mxmou coHucoHGdEEoo HoGoHqudm so moocmEH0meQ Dmououm

.mH

CHAIR.

101

.5238 “0:5 5 083:8an Hush .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.288 2 039—.

immune Am.ume nevuue He.ome He.uuo ..u.u.
Hmum mm bmnn n._vv m.mw ado:
mHum 3. menu HO OH. Ozo
Hmnm K. mHnu mH m; £40
vanm 3 menu mp mm #20
”Tm om wHuw mm HS ”20
NmnN O mm" m O SH ”20
EN or bHuw NH. mm HZU
.meomembm

01.225 20

8.33 2..an Hmong 3.93 :23 H.U.3
.vmuh sham MN; N.Om m.wm use:
moum om mmum mm m E
monm wb OHVn N mm on mm
mHnm Hum mo; ma. OM .Vm
OHum Hm Hmnm mm x mm
OHuw Hm. mvum ON .X E
m¢"OH mm mmum on H} Hm

riflmmrmnnflfiv 33 693504 303835 33 hush—50¢ $3 Nuns
aoHusHsn flown-sunn— . meomnmbm 3225.
“Hush. “Hush. 3.22036sz
.nOHnHo> GOUUHHS .flOHmHOKV unonm (ma—nail
anus—uncued
ADHHHOUHO uoHduuuv Ends nofiuouduo HddOHuoflnh UONHUHdUflsum

mxmcu GOHOMUHGSEEOO HmcoHquDu Go mmocmfiuowHon umwumhm .mH OHAIH.

102

dog—Eco “0:5 5 3:58.59 “Rah .

 

 

 

 

 

 

 

 

 

 

 

333 3.4.3 3.33 2......3 fiméwfl 70;:
«mum v.3 onum mufic. mgr Egg
Avonuadfiv 33 #09502 ~00uundflv Gov hound—004 Cay «Lug
nowudunn dowuuudn
xuda xnda
.nownun> nouuuhz .no«uuo>.ononm
uaoauuonud

.uaauoouo noduuuov nuns

nodHOUdNU annouuondh

 

gadgvfiflum

 

 

 

103

no relationship between phone ordering and performance on
the sentence~picture matching task. There was a moderate
correlation between phone catalog ordering and single word
reading comprehension (r = 0.72, p<0.01). Phone catalog
ordering was more strongly correlated with all aspects of
naming than CADL—Z, including BN1 (r = 0.89, p<0.01), oral
naming as measured on EALBA (r = 0.88, p<0.01), and
writtennaming (r = 0.76, p<0.01). Overall, the phone
version of the catalog ordering task relied more on naming
abilities and less on auditory comprehension than CADL-Z.
The written version of the catalog ordering task was
related to performance on the written naming subtest of
EALBA (r = 0.70, p<0.05) and single word reading
comprehension (r = 0.82, p<0.01). Performance on the
written version of catalog ordering was not related to
performance in oral naming. The two versions of the catalog
ordering task draw on modality-specific aspects of naming.
The baseline reaction times were also compared to
ensure that all subjects had sufficient motor skills to
perform the concurrent task and were comparable in speed and
ability. The reaction times for subjects to respond to the
computer generated tones by hitting the foot switch in the
baseline condition are given in Table 16. The functional
and CN groups were comparable on baseline reaction times,

and all subjects were able to perform the baseline task.

104

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects
Functionally
trained subjects. Baseline RT (ms)
F1 436
F2 267
F3 471
F4 591
F5 609
F6 365
Mean 456.5
(s.d.) 1131.31
CN Trained Subjects.
CN1 523
CN2 428
CN3 177
CN4 420
CNS 697
CN6 286
Mean 421.8
(s.d.) (181.3)
Overall Mean 439.2
(s.d.) (152.0)

 

 

 

Table 16. Concurrent task baseline reaction times for each

subject.

105

In summary, the subjects assigned to the two groups
were comparable on basic cognitive skills such as
recognition memory, visual cancellation, visual-perceptual
skills, and on nonverbal problem solving as measured by
BC_M. They varied in their abilities to do digit and
pointing span tasks, and to perform on the WEST. The twelve
subjects were all relatively good at single word
comprehension, but showed impairments of sentence level
auditory comprehension. There was a wide range of naming
performances, which were correlated with performance on the
functional communication assessments. Only written naming
and single word reading comprehension were related to
performance on the written version of catalog ordering.
This general profile is consistent with the typical
abilities of people with nonfluent, Broca's type aphasia.
It demonstrates both the consistency of the diagnostic
category and the wide range of skills that such a
categorization tolerates.

Pro/Post Differences

Difference scores for all of the pre/post assessment
measures were calculated by subtracting the pretest score
from the posttest score. Difference scores render the

performances more interpretable for the purposes of this
study. It is possible to determine “clinical significance”

by inspecting the difference scores within and between
subjects. Clinically significant change for some

standardized measures have been determined as at least one

106

point out of ten possible on a standardized aphasia battery
subtest which is scored correct or incorrect, or at least
five points on a summary of several subtests (Katz & Wertz,
1997). Clinically significant change on the CAQL;2, a test
with a three—point scoring interval for each item, has been
defined as 10 points (Elman & Bernstein-Ellis, 1999). This
method was deemed appropriate for analyzing change in
response to the two treatment methods in the present study.
The following criteria were adopted for analyzing
clinically significant change. A criterion of 5 points was
established for clinically significant change on
standardized measures of comprehension and naming, which all
used +/- scoring. This was set as a more stringent measure
than the 1 point criterion applied in other work. The
previously established criterion of 10 points for difference
scores on the CADL-z and also the catalog ordering task,
which used the same scoring procedure as CADL-2, was applied
to these two measures of functional communication. A
criterion change score of 10 points was also established for
clinical significance on the C511. This was based on
original reports of an average change score of 0 for a
stable group of aphasic adults, and an average change of 11
points on this instrument for recovering patients (Lomas et
al, 1989). Similarly, reported average change for the
Correct Information Unit analysis applied to the discourse
samples for session—to-session stability served as a basis

for determining criterion on the discourse measures. Since

107

an average change of 5 to 7 CIUs was observed between
sessions (Nicholas & Brookshire, 1993) across all discourse
types, a criterion of 10 CIUs was set as suggesting
clinically significant improvement for the pre/post
discourse measures.

Difference scores for all three standardized measures
of comprehension are shown in Table 17. Neither group
demonstrated any particular improvement overall on the
single word measures of reading and auditory comprehension.
Two of the six CN subjects (33%) and one of the functionally
trained subjects (17%) made clinically significant
improvement on sentence-picture matching, but this was not a
pattern of either group overall.

Both subject groups made improvement on measures of
oral and written naming. See Table 18. Five out of six
subjects in each group (83%) made clinically significant
change (a difference score of at least plus five points) on
at least one measure of oral naming. Three subjects in the
functional group (50%) and four subjects in the CN group
(67%) improved in written naming. The CN trained subjects
made the most improvement on written picture naming (mean =
9.67), compared to their performance on the BET (mean = 7.5)
and the EALBA naming subtest (mean = 7.5). The functionally
trained subjects made more improvement on oral naming,

including the EMT (mean = 11.5) and the PALPA naming subtest

 

(mean 13.5) than on written naming (mean = 7.5).

108

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reading Auditory Comprehension

Subjects Comprehension
written WOrd- Spoken Wbrd- Sentence-
Functionally Picture Picture Picture
Trained Matching - Matching - Matching -
Subj 80 t s - PALPA 9gp; my;
F1 5 0 10
F2 0 0 0
F3 —3 -5 -10
F4 0 O 0
F5 2 -2 3
F6 0 0 0
Mean 0.7 -1.16 .5
(s.d.) (2.65) (2.04) (6.4)
CN Trained
Subjects.
CN1 0 -5 -l7
CN2 -5 -10 33
CN3 0 0 20
CN4 0 0 O
CNS 3 10 5
CN6 0 5 -10
Mean -.33 1.67 5.16
(s.d.) (2.58) (6.83) (18.67)
.17 .25 2.83

Overall mean (2.55) (5.02) (13.54)

(e.d.)

 

 

 

 

 

Table 17.

comprehension for both subject groups.

109

Difference scores on standardized measures of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects Boston Naming Spoken written
Test Picture Picture

Functionally Naming - Naming -
Trained PALPA PALPA
Subjects.
F1 25 5 3
F2 3 37 0
F3 5 l4 5
F4 5 5 3
F5 4 —3 16
F6 27 23 18
Mean 11.5 13.5 7.5
(s.d.) (11.27) (14.55) (7.55)
CN Trained
Subjects.
CN1 2 22 22
CN2 3 0 0
CN3 0 5 8
CN4 6 3 3
CNS 15 18 17
CN6 l9 -3 8
Mean 7.5 7.5 9.67
(s.d.) (7.71) (10.13) (8.35)
Overall Mean 9.5 10.5 8.58
(s.d.) (9.44) (12.36) (7.68)

 

 

Table 18. Difference scores on

naming for both subject groups.

110

standardized measures of

 

Difference scores for the functional communication
tasks, the CADL—Z and the quiet phone and written versions
of the catalog ordering task, were calculated and are shown
in Table 19. The functionally trained subjects did not make
particular improvement on the CADL-z (mean = 0.5), but the
CN subjects did (mean = 9.83). The average CADL-Z
performance for the CN subjects was essentially at the level
of clinical significance. Three of the CN subjects obtained
difference scores of +10 or greater (50%), and none of the
functionally trained subjects reached this criterion.

The functionally trained subjects improved notably on
the phone version of the catalog ordering task, which had
been included in their training (mean = 32.3). All of the
functionally trained subjects (100%) reached or exceeded the
level of clinically significant improvement on the phone
version of catalog ordering. None of them, however,
achieved clinical significance on the written version, which
had not been trained. The CN subjects on average improved
more on the written version of the catalog ordering task
(mean = 11.83) than the functional subjects (mean = 6.83).
Four of the six CN subjects (67%) achieved clinically
significant improvement on the written version of catalog
ordering, and two of them attained this level of improvement
on the phone version of catalog ordering.

Performance of the functionally trained subjects on the

phone—related items of the CADL—Z were analyzed to determine

111

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects Standardized Functional Criterion Task
Assessment (Catalog Ordering)
ggggégnal 1y CADL-2 Phone version' 3:33:31
Subjects.
F1 7 47 2
F2 0 54 -9
F3 —4 17 -2
F4 -1 25 6
F5 -5 20 11
F6 6 31 33
Mean .5 32.3 6.83
(s.d.) (5.01) (15.01) (14.52)
CN Trained
Subjects.
CN1 12 -22 12
CN2 14 0 16
CN3 9 —6 17
CN4 5 17 —5
CNS 16 14 3
CN6 3 -6 28
Mean 9.83 -.5 11.83
(s.d.) (5.11) (14.41) (11.54)
Overall Mean 5.17 15.91 9.33
(s.d.) (6.86) (22.16) (12.78)

 

 

Table 19.

tasks for both subject groups.

 

‘ Performance in quiet condition.

112

Difference scores on functional communication

 

whether there was improvement on items that were
specifically related to the functional training. Items 40,
41, and 42 require subjects to identify a phone number in a
directory, dial the phone, listen to a tape recorded
message, and report back to the examiner. The three-point
scoring system is applied to these items according to the
standard scoring procedures. The pretest and posttest
performances for each of the functionally trained subjects
on these three CADL-Z items combined appear in Table 20.
Overall, the functionally trained subjects performed
relatively well at pretest on these items leaving little
room for improvement. Two subjects improved by one point
each and one subject decreased by one point on the selected
items. There was essentially no change on these items for
the functionally trained subjects.

Since the CN subjects improved on both the CADL-Z and
written naming and catalog ordering, an item analysis was
conducted to determine whether improved CADL—2 scores were
attributable to an improvement in writing skills overall.
Since the CADL-Z uses a three point scoring system that
assigns points on the communicative effectiveness of the
response, regardless of response modality, it is possible
that improvement on this test could be due to a particular
modality—specific pattern. The response modality of each
item with an improved score at post test was determined.

Some items were improved at posttest because the

subject was able to completely verbalize the appropriate

113

Table I
functic

items (

 

 

 

 

 

 

 

 

 

 

 

Subject Pretest Posttest
P1 5/6 6/6
P2 6/6 6/6
P3 6/6 6/6
P4 6/6 6/6
P5 4/6 3/6
P6 4/6 5/6

 

Table 20. Pretest and posttest performances of the
functionally trained subjects on the three phone-related

items on CADL—Z.

114

 

to a j
"fire"
respoz
during
receix
pointi
questi

T
each s'
modali-
respon:
patCEn

functi

Percei~
Scores

indiCa

 

functi
5

change

 

response. For example, for the stimulus item "if you saw
this [picture of a house burning], what would you do?", a
full credit verbal response would be»"call 911" , compared
to a partial credit response like writing down the word
"fire". Other items either accept or require written

responses, such as completing a ‘patient information form’
during a doctor’s visit role-play. Furthermore, many items
receive full credit for gestural responses, including

pointing to the correct portion of a bus schedule given the
question."what time does Bus #3 leave Maintown?".

The percentage change for each response modality for
each subject is given in Table 21. The written response
modality did not improve more than either verbal or gestural
responses. Overall, the CN subjects had a more generalized
pattern of improvement across all modalities than the
functional subjects.

The 9311 provided ratings by significant others of
perceived change within the family environment, and these
scores are displayed in Table 22. The CN trained subjects
overall received higher change ratings (mean = 13.67),
indicating more perceived improvement overall, than the
functional subjects (mean = 9.83). Two functional subjects
and five of six CN subjects achieved clinically significant

change on the CETI.

115

r_____._
Subject
Functio

‘Trained
Subject

(Fl

 

modali«

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subject verbal written Gesture

Functionally

Trained

Subjects.

Fl 90 10 0

F2 50 50 0

F3 0 0 100

F4 100 0 0

F5 50 25 25

F6 25 25 50
45 18.3 29.2

Mean

CN Trained

Subjects.

CN1 33 12 55

CN2 15 20 65

CN3 25 30 45

CN4 80 20 0

CN5 42 8 50

CN6 0 0 100
32.5 15 52.5

Mean

Overall Mean 38.75 16.7 40.85

 

Table 21. Percentage of the total change on CADL—2
attributable to improvements in different response

modalities for each subject.

116

 

 

 

 

 

O
J:

D J 0 ‘fl
/ " cf;

 

O
‘4 '
(J1

8‘ £2
5}»

s . d .
Overal

 

Table

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects QETI score
Functionally trained
subjects.
F1 18
F2 7
F3 9
F4 10
F5 9
F6 6
Mean 9.83
(s.d.) (4.26)
CN trained subjects.
CN1 14
CN2 7
CN3 18
CN4 18
CN5 12
CN6 13
Mean 13.67
(s.d.) (4.13)
Overall Mean 11.75
(s.d.) (4.47)

 

Table 22. CETI scores for both subject groups.

117

 

Ef
task du
Differ

subtra

   

for the
differe
change
catalog
six fur
increas
catalo;
had bee
did no:
Dostte:
averag.
SubjECf
Perforr
Robust:

Calcul

task C

 

 

DErfOr

dUrati

 

Efficiency of performance was assessed by measuring
task duration on the functional communication tasks.
Difference scores for duration were calculated by
subtracting the pretest duration from the posttest duration
for the CADL-2 and the catalog ordering task. The duration
differences are given in Table 23. There was essentially no
change between pretest and posttest task durations on the
catalog ordering task for both subject groups. Two of the
six functionally trained subjects (33%) substantially
increased their task duration for the phone version of
catalog ordering. Otherwise, the functional subjects, who
had been trained specific strategies for the phone version,
did not change in task duration between pretest and
posttest. The functional subjects also maintained their
average task duration for CADL-Z. Four of the six CN
subjects (67%) decreased task duration time for CADL-z
performance, suggesting an improvement in efficiency.
Robustness of the training effects was analyzed by comparing
performances on the catalog ordering task under conditions
of load. The accuracy and duration difference scores were
calculated for the catalog ordering task in the concurrent
task condition, and these are shown in Table 24. These
performances were compared to accuracy levels observed
during the quiet condition, shown in Table 19, and task

duration in the quiet condition, shown in Table 23.

118

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects Standardized Functional Criterion Task
Assessment 4(Catal g Ordering)
gifgirtlégna 1 1y CADL- 2 Phone vers ion‘ :jisfg
Subjects.
F1 5 2.5 .6
F2 3 6.67 -2.86
F3 —2 -.19 -1.07
F4 —4 -.84 -.04
F5 2 .91 2.52
F6 0 .89 2.46
Mean 0.67 1.65 .27
(s.d.) (3.32) (2.71) (2.08)
CN Trained
Subjects.
CN1 -5 .28 -.18
CN2 —4 —.3 -.2
CN3 —16 -.99 1.11
CN4 -9 .13 1.56
CNS 0 -.03 -.01
CN6 7 .99 .28
Mean -4.5 .00 .42
(s.d.) (7.81) (1.45) (.73)
Overall Mean -1.92 .83 .35
(s.d.) (6.33) (2.06) (1.49)

 

 

Table 23.

Difference scores for task duration in minutes on

functional communication tasks for both subject groups.

Negative values represent an improvement from pretest to

posttest.

 

‘ Performance in quiet condition.

119

 

T. . L: SMHWTISEEEEFWL «WAS EMF: CE um! WWU

eh

 

Phone version

written version

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects Accuracy Duration Accuracy Duration
Function-

ally

Trained

Subjects.

Fl 22 6.06 10 -0.9
F2 33 5.25 43 -0.45
F3 14 -0.1 7 0.72
F4 6 —1 6 -0.83
F5 20 0.62 55 -1.95
F6 50 0.77 25 1.89
Mean 24.17 1.65 24.3 -0.25
(s.d.) (15.49) (2.71) (20.65) (1.36)
CN1 —6 0.2 26 -0.08
CN2 0 0.7 16 -0.8
CN3 0 -1.72 17 0.15
CN4 25 0.86 -11 —1.23
CNS 8 -0.05 11 -0.86
CN6 17 -0.68 28 2.76
Mean 7.33 -0.12 14.5 -0.00
(s.d.) (11.76) (0.96) (14.04) (1.45)
Overall 15.75 0.91 19.42 -0.13
Mean (15.79) (2.36) (17.59) (1.35)
(s.d.)

Table 24. Differences in accuracy and task duration for

catalog ordering with concurrent task performance for both

subject groups.

120

 

Cc
disrupt
conditi
functic
32.3; I
they 11
condit.
subj ec
the ta
improv
= 24.3

T
Signi
concu
averag
bette;
to p0
diffe
accur

the s
I

11ride:

in th

14.5

Sight

COHC

 

Comparison across these tables reveals a slight
disruption of accuracy on the phone version from the quiet
condition to the concurrent task condition for the
functionally trained subjects (mean in quiet condition =
32.3; mean in concurrent task condition = 24.17), although
they improved overall from pretest to posttest in both
conditions. Interestingly, the functionally trained
subjects improved only marginally onthe written version of
the task in quiet (mean = 6.83), but their performance
improved even more so in the concurrent task condition (mean
= 24.3).

Two of the six CN subjects (33%) made clinically
significant improvements on the phone version in the
concurrent task condition from pretest to posttest. The
average improvement for the CN group (mean = 7.33) was
better than their average performance in quiet from pretest
to posttest (mean = -0.5). Comparison of individual
difference scores for accuracy in the quiet condition to
accuracy in the concurrent task condition shows that four of
the six CN subjects (67%) improved under conditions of load
in the phone version. This pattern of improved performance
under conditions of load was also observed in the written
version of the task for the CN subjects overall (mean =
14.5). Five of the six CN subjects (83%) made clinically
significant improvement on the written version in the

concurrent task condition from pretest to posttest (Table

121

24). Their overall improvement on the written version was
similar across both quiet and concurrent task conditions.

Task duration during the concurrent task condition did
not change from pretest to posttest for any of the subjects
in either group, with the exception of two functional
subjects (F1 and F2), who prolonged task performance at
posttest compared to pretest for the phone version under
both conditions (Tables 23 and 24). For the CN subjects,
task duration remained essentially the same from pretest to
posttest in both conditions and both versions of the task.

Another measure of efficiency of function is the
analysis of reaction time data from the concurrent task.
Reaction times to each presented tone were recorded in
milliseconds and mean reaction times for each subject and
each condition at pretest and posttest were calculated.
Difference scores were generated by subtracting pretest
reaction times for each condition from posttest reaction
times. These differences are listed in Table 25. There was
no particular pattern of performance for reaction times for
the two subject groups. Subject CN2 missed a large number
of tone presentations at pretest, however he was able to
perform the concurrent task quite well at posttest. Four of
the six CN subjects (67%) improved their reaction times at
posttest in the phone version of the task, and only one of
the functional subjects (17%) did. In the written version
of the task, three of six subjects in each group (50%)

improved their reaction times. Overall there was a high

122

 

Subjects
Functionally
Trained Subjects.

Phone version

written version

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F1 314 261
F2 44 65

F3 234 350
F4 -460 -543
F5 158 —30
F6 74 -20
Mean 60.67 13.83
(s.d.) (273.95) (313)
CN Trained

Subjects.

CN1 —30 40
CN2 —2352 581
CN3 143 -100
CN4 162 -49
CNS -l41 -510
CN6 -259 -95
Mean -412.83 -22.17
(s.d.) (963.73) (351.49)
Overall Mean -176.08 -4.17
(s.d.) 41719.33) (317.87)

 

Table 25. Difference

scores for reaction time (msec) on the

catalog ordering task for both subject groups. Negative

scores indicate an improvement in speed at posttesting.

123

 

degree of variability across the subjects for reaction time
data, as evidenced by the large standard deviations.
Generalization of the treatment effects from trained to
untrained vocabulary items was analyzed for each treatment
group. Functionally trained subjects were trained on half
of the target items used during pre/post testing on the
catalog ordering task. Only two functional subjects (33%),
F1 and F6, failed to produce items during posttesting on
catalog ordering that had been trained. Each of these
subjects failed to produce a single item each, and did so in
the quiet condition of the phone version. Two of the
functional subjects (33%), F4 and F6, produced items at

posttest that had not been trained and had not been produced

at pretest. F4 produced “blue coat” and.“yellow sweater” at

posttest that had not been trained, and F6 produced.2hat”

which had not been trained. None of the other functionally
trained subjects produced any untrained items.

Training items for the CN subjects had been selected by
identifying consistently missed items on the 54-item
baseline assessment. Half of these items had been trained
across two training sets. The S4-item baseline was re-
administered after treatment. Four of the CN subjects (67%)
produced at least two untrained items at re—administration
of the 54 item assessment at the end of treatment. The
individual percentages are reported in Table 26. In

general, the CN subjects tended to show more generalization

124

CN1
CN2

CN3
CN4
CNS
CN6

 

Table 26. Percentage of untrained items produced at

posttesting for each of the CN subjects.

125

to untrained items than subjects in the functional treatment
group.

In summary, subjects in both treatment groups improved
on written naming and on the written version of catalog
ordering in the concurrent condition. Subjects in the CN
group made clinically significant improvement on the CADL—Z,
the CEII, and in both conditions of written catalog
ordering. They also tended to show more generalization to
untrained items. Subjects in the functional treatment group
made clinically significant improvements on two measures of

oral naming, the BNT and the PALPA, and on both quiet and

 

concurrent conditions of the phone catalog ordering task.

Patterns of Transfer to Discourse

All of the discourse samples were orthographically
transcribed and the total number of Correct Information
Units (CIUs) were calculated based on the procedures of
Nicholas & Brookshire (1993). The sampling contexts for the
Nicholas and Brookshire (1993) data include expository,
narrative and procedural discourse measures. Non-brain-
damaged adults produced no fewer than 50 CIUs on average
across all elicitation contexts, and 49 CIUs was identified
as the cut—off score for non—brain—damaged adults. The
number of CIUs produced in each discourse sampling context
at pretest for each of the subjects in the present study is
shown in Table 27. Only one subject produced a total number
of CIUs that fell within the normal range, subject F4 in the

narrative discourse task. This subject produced a 15 minute

126

.wQDOHm uowmfldw 03u mflu How mwaeu wwudoumwp

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mwhgu mflu mo £000 NOW ammuwum um AmDHUO mufinD GodumfiuomGH uomhwoo mo HmQEdz .hN canda
A.U.mv

:s.mV AN.QO 250.3 AN.wV Ac.omv Hm.mv s02
h.w hum M.m mow §.MN MN.m HHUH0>O
AN.mO ANS: AH.wv Am.hv AOHL A¢.mv A.U.mv
h.# h.@ m.‘ w.h N.‘ O.m “60:
O OH OH m m MH $20

0 O O m N N mZU

w b O ON 0 OH «20

v m N H O O MZU

O O O O O O NZU
vH NH vH mH mH m HZU
.wuomhndm

pwsflmnu

ZU

Amuhv Am.OHv AO.NHV afl.¢v Aw.mwv RNmumv A.U.wv
how b.0H M.NH m.m bwoflfi m.MH fldflz
H O O O H O oh

v m m OH O OH mm
vH mH ON w NHN VH vb
mH wH PN OH ON PH mm
O m O H O mN Nm
mH ON NN w MN HH Hm
.muowaSm

Uwcflmsu

hHHOH. >HHM

uOHduuu nowuooouu dNNMm a bunch omufloondn unknoondn IGofiuoCdm
unknoomwn Hauuuoooum 0>Mushudz Shouunonxfl nuochndm

 

 

127

sample relating the story of Cinderella. She reported that
this had been a favorite story of hers which she had
listened to everyday during her childhood, and she had
retold to her children and grandchildren. The number of
CIUs subject F4 produced in the other sampling contexts was
more consistent with the rest of the subjects in the study.
With the exception of F4's Cinderella story, the number of
CIUs for these subjects across discourse contexts was
consistent with previous reports of nonfluent aphasic
performances (Nicholas & Brookshire, 1993; Yorkston &
Beukelman, 1980; Craig et al, 1993).

Difference scores were calculated by subtracting the
total number of CIUs produced in each elicitation context at
pretest, from the total number produced at posttest.
Difference scores for each subject in each discourse context
are shown in Table 28. Overall, there was a wide range of
variability in performance between pretest and posttest, as
evidenced by the large standard deviations.

None of the functionally trained subjects and one of
the CN trained subjects made clinically significant
improvement on expository discourse. One subject in the
functional group (17%) and three subjects in the CN group
(50%) made clinically significant improvement in the
narrative discourse context. One to two subjects in each
group improved in the various procedural discourse contexts.

Three subjects in the functionally trained group (50%) made

128

.mmdoum uomwnfim anon

HOm women mmusoomap mounu mnu mo 3080 How mDHU wo HmQEdG Go monoum wocmnmmwaa

.QN OHQGB

129

.E .335 we oocmccotom 05 38c“: sou—Eon cavesm och a

5:5“ 03:95 E ewes—.6 « 38:2 8: mace 02.; $5. .3558 w 3 038% 8: co camco—

ofi c0328“. bccuocEwi 2133:8348. .20: .«o Sign.“ a we? $.53. 2.23:5 05 GEES @2095 2... 8360b 0335: 8=EE 2 a 32.55 283:.“ min. .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

95:9: an wish.
Am.¢v A#.wv A¢.mv Am.mv Av.va Am.vv A.U.uv
m.N m.N o.m O.H o.nH Q.m G602
O NI MI O OH 0 wZU
w N m MI NI H mZU
OH v MN O vm NH VZU
vI vl O HI m H mZU
O O O O O O NZU
m vH m N mm m HZU
.muowwnsm
Uwchsu
Z0
A#.mv Am.HHO Am.HHv Am.Nv nam.ovv Ho.mv ~.U.mv
m.w m.OI m.H O.H m.ml mm.MI G602
6H H m H b m mm
OH NH O m vH O mm
wH HI ml v .le O 6m
v m ON MI m m mm
O MI m O m mHI mm
ml NNI OH1 H w ml Hm
.muowhfldm
Umcflmuu
hHHOb % m
uoHduuu nONHoOOHO sNNHm a unsoa unknounwn onunoonwn IGoHuowwm
onuuouan Hdhnuooonm 0>Huduusz Shouauogfim muoohndm
.wDHO How wmhoom wonmummmwo .wn sauna

130

131

 

 

 

 

 

 

Aououfi
Am.wv AN.mv A¢.O.: Ao.mv Aminmv Am.hv n30:
ho. whoa ng Ood mod MN.O HHGHOPO
NHHOb
uoHsusU nowuoooua snafu a sedan. 09550an “Macon .3
Gannon-HO Hsuavoooum 0b.." venue: 032.03 L

 

 

 

 

 

 

 

 

 

a notable improvement in the describing how to order
something from a catalog, and all three of these subjects
improved more than the one CN subject who improved in this
context.

For the expository and narrative discourse samples,
subjects with the largest magnitude of improvement were in
the CN treatment group. This was also true for the
procedural discourse samples with the exception of the

catalog ordering description.

Cognitive-Linguistic Influences on Treatment Success

As previously reported, the two groups of subjects
differed on measures of high-level cognitive abilities
(Table 12) and displayed variable pretest naming abilities
(Table 14). Inspection of difference scores on the pre/post
assessment measures suggested differential outcomes between
the two groups. It is important to consider the potential
role of initial cognitive and language status on the
patterns of outcomes.

The three functional subjects with the lowest cognitive
abilities (F1, F2, and F6) also achieved the lowest pretest
naming scores of the six functional subjects (Table 14).

F1, F2, and F6 consistently demonstrated the most severe
cognitive and linguistic impairments of the six subjects in
this group. These three subjects also had deeper lesions,
<extending into subcortical gray matter, than the other three

functional subjects.

132

After the six week functional treatment, inspection of
difference scores reveals that these three subjects improved
more on oral naming and the phone version of catalog
ordering in the quiet condition than the other three
subjects, although the other three subjects also improved in
these areas. Two of these three subjects, F1 and F6,
improved on the CADL—Z as well (difference scores of +7 and
+6, respectively, although these scores did not achieve the
criterion set for clinical significance). However, the
three functional subjects with milder cognitive-linguistic
impairment at pretest made no change on QADL—Z and in fact
obtained negative difference scores.

The role of cognitive-linguistic factors in the
robustness of the treatment effects can be analyzed by
inspecting the effects of the concurrent task on performance
accuracy. The difference in performance between the quiet
and concurrent task conditions at pretest were subtracted
from that difference at posttest. Positive values reflect
improved performance in the quiet condition, and negative
values reflect improved performance in the concurrent task
condition. These difference scores for each subject are
shown in Table 29.

The performance of four of the six functional subjects
xnas better in the quiet condition than in the concurrent
‘task condition of the phone version and the performance of
'three of these four subjects was substantially facilitated

(F1, F2, and F4). F1 and F2 are two of the three subjects

133

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subjects.

Phone Written Cognitive-
Functionally Linguistic
trained subjects. Rating!)
Fl +25 -8 Low
F2 +21 —52 Low
F3 +3 -9 Moderate
F4 +19 0 Moderate
F5 0 —44 Moderate
F6 —19 +8 Low
Mean 8.17 -17.5
(s.d.) (16.71) (24.54)
CN trained
subjects.
CN1 -16 -14 High
CN2 O 0 Low
CN3 -6 0 Low
CN4 -8 6 Moderate
CNS 6 -8 Moderate
CN6 —23 0 Moderate
Mean -7.83 -2.67
(s.d.) (10.52) (7.12)
Overall mean 0.17 -10.08
(s.dl) (15.72) (18.89)
Table 29. The difference of difference scores for catalog

ordering performance under quiet and concurrent task

conditions for each subject.

134

with the most severe cognitive—linguistic profiles. In
contrast, F6, who also had a more severe cognitive-
linguistic profile, demonstrated a facilitation of his
catalog ordering performance with the concurrent task. The
three subjects who were most substantially aided by the
quiet condition, F1, F2, and F4, all had frontal lesions.

CN1 and CNS made the largest improvements in oral
naming of the CN group, although overall there was slightly
less improvement in naming among the CN subjects than for
the functionally trained subjects. The two CN subjects with
the worst cognitive profiles improved the least on oral
naming. All of the CN subjects made some degree of
improvement on the CADL—Z. There did not seem to be a
relationship between cognitive-linguistic profile of the
subject and improvement on CADL-Z after CN treatment. Nor
was there any apparent relationship between improvement on
the written version of the catalog ordering task after CN
treatment and cognitive-linguistic profiles.

In contrast to the functionally trained subjects, five
of the six CN (83%) subjects showed either no effect of load
in the concurrent task condition, or experienced a
facilitation of performance in the phone and written
versions of catalog ordering. Five of the six CN subjects

had lesions affecting the left frontal lobes.

135

Cog

ave
twc
rel

SU

OI“.~

in

St

DE

(”1'

 

Cognitive-Linguistic Influences on Rate of Improvement

The amount of treatment time, number of sessions, and
average amount of time per session for each subject in the
two treatment groups are given in Table 30. There was a
relationship, particularly for the functionally trained
subjects, between cognitive ability as measured by the WEST
and aggm and the amount of training time required to achieve
criterion. In general, the subjects with the lowest scores
on was: in each group required the greatest amount of time
in total minutes for the training. However, subjects with
the highest cognitive profiles did not necessarily achieve
criterion in the least amount of time.

Among the functionally trained subjects, there was a
statistically significant negative correlation between
performance on the figs: and total number of training
sessions (r = -O.89, p<0.05). There was also a
statistically significant negative correlation between
performance on the BQEM and total training time (r = -O.88,
p<0.05) and between the EQEM and the total number of
training sessions (r = -O.89, p<0.05). These relationships
were not observed among the CN trained subjects. The
observed correlation is a general pattern only, and does not
suggest a direct relationship for each individual between

the two measures of cognitive ability and treatment time.

136

 

_ karanduHUnwnw

finance... wwflmww< a uwaupz Hmuoa \

HUUOE-

WUUUWNNM—W

 

 

 

-\

.mfifiu «noun mdqflE mfiflu Gofimwwm Hmuou mamsvw mafia mafiafimue

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.uowwndm £06m How GoflHwUflHU m>wfl£om ou UmHHSUwH wEHu ucmfiumonu mo uGSOE< .om ufinda
AN.0HV Am.Nv Am.0bdv A.U.nv
m.mH w.HH mmu n60: HHdHo>O
. .u
3.8 F.: 3.32 A mum“
w.wd m.HH «.mmH
wumuwcoz h Ha or wzu
muchmooz m oa mm mzo
wumuwooz ma an Hma V20
304 on ma New qu
309 on ma mam WZU
nmflm H.” m-“ 5..“ Eu
.muothSm
Umcflmuu 20
3.”: 392 2.83 A whats
w.HN b.HH th
304 ma NH mma wm
wumumpoz m Ha mm mm
mummwcoz NH w Nb vb
wumuwcoz on an mmm mm
30A w” .W” mos mm
304 mm M?“ bmm Hm
.muvwhnsw
Amoufldfifiv Umcflwnu
unaudm nowmuon Auousnfifiv haamcofluocsm
owumwsbnfln H00 65%» HO aflownuom no 05MB unwnfidua
Io>wuwfluoo udfiOEd oudhohfl Honfifiz HUHOH Hduoa muumHASm

 

 

137

Cos

peI

pat

p“?1

pe’

miC

f111

m!

moE

COS

Cognitive-Linguistic Influences on Transfer to Discourse
There did not appear to be any relationship between
performance on the cognitive battery and the generalization
patterns to discourse. The subject with the best cognitive
performance, CN1, showed one of the best discourse
performance improvements, but so did CN4, who was in the
mid-range of subjects for cognitive abilities. Among the
functional subjects, F4, F5 and F6 made the most
improvements in discourse. One of these, F6, had one of the
most severe cognitive profiles, and F4 and F5 had mid—range

cognitive performances.

138

Gen

Se:
max
Va:

195

Dr.
Ou
Su
th

im

Chapter 4

DISCUSSION

General Effects of Treatment

The first of the study’s predictions was that subjects
in both treatment groups would show improvement. This
prediction was supported as demonstrated by improvements in
oral and written naming, at least one measure of functional
communication, and on caregiver ratings of the 9511 across
both treatment groups. Furthermore, some individuals made
additional improvements in specific areas such as auditory
sentence comprehension. These results are in agreement with
many other studies showing the effectiveness of treatment of
various types among adults with chronic aphasia (Robey,
1994; 1998).
Differential Outcomes of the Two Treatments

The two different treatment approaches compared in the
present study were predicted to differ in patterns of
outcomes and transfer, and these predictions were generally
supported by the data. It was predicted that subjects in
the functional treatment group would show much more
improvement on the catalog ordering task, which had been

trained, than subjects in the CN group, and this difference

139

was c
both
funct
suffe
of 1c
than
subje
imprc
phone
The 1
imprc
predj
trair
Hamil

subje

was observed. Improvement in accuracy was demonstrated in
both quiet and concurrent task conditions for these
functionally trained subjects. However, these subjects
suffered a disruption in their performance under conditions
of load, and this effect of the concurrent task was greater
than the disruption obtained at pretest. The functional
subjects showed little improvement on CADL-Z, and this
improvement was not attributable to increased ability on
phone related items on the test, contrary to predictions.
The functional subjects were predicted to show little
improvement on standardized measures of naming, and this
prediction was not supported. Indeed, the functionally
trained subjects demonstrated greater improvement on the

naming subtest of the PALPA and the BNT than the CN

 

subjects.

The CN subjects made the little amount of improvement
on accuracy in the catalog ordering task that had been
predicted. However, these subjects appeared to improve in
general efficiency of function as measured by the lack of
disruption on performance made by the concurrent task.

These subjects also showed marked improvement on QAQL;2, and
this improvement had been predicted due to the nature of
this test, which requires the deployment of multiple
abilities and strategies. The CN subjects were predicted to
show greater improvements on the standardized measures of
naming than the functional subjects, and this prediction was

not supported. The caregiver ratings on the QETI were

140

in
of
Th
C}:
01:
th
mc
at
re

81

CC

in

t}

CE

greater in magnitude for the CN subjects, and this
observation tends to support the prediction that CN
subjects’ improvement will be observed in a wider variety of
contexts.

The results of CN treatment were consistent with
predictions. It is quite interesting to note that an
important outcome for this group of subjects was on measures
of functional communication, specifically, QADL-Z and £321.
These measures are typically not administered in studies of
CN treatment, and so this outcome has not previously been
observed. These results are consistent with observations
that CN treatment which targets semantic representations is
most likely to generalize to a broad range of communicative
abilities. The generalized effect of the CN treatment
regardless of cognitive ability of the subject tends to
suggest that this treatment approach is in fact focused on
the language system, and does not require additional
cognitive abilities to support its outcomes.

Only three of the six CN subjects made notable
improvement on standardized measures of oral naming, and
this was contrary to the original prediction. This result
can be interpreted as being consistent with the previous
literature, however, in which some studies show an
improvement on generalized naming, while other studies do
not (Hillis, 1989; Howard, 1986; Howard et al, 1985; LeDorze
& Pitts, 1995; Nickels & Best, 1996; Raymer et al, 1993).

It appears that a more reliable method of detecting changes

141

based on a semantic treatment may be assessment of overall
language abilities, rather than a reliance on modality-
specific tests.

Generalization to Discourse

Although the pattern of transfer to discourse was not
conclusive, the results were suggestive of the initial
prediction. Half of the CN subjects improved on the
discourse task requiring the greatest variety of skills, the
narrative discourse task. Half of the functional subjects
improved notably on the catalog ordering procedural
discourse task, due to the use and generalization of trained
vocabulary to this particular discourse context.

The results generally support theoretical predictions
about the nature of the two contrasting treatments. The
functional treatment resulted in improvements attributable
to the contexts trained, with some generalization to closely
related discourse tasks, and demonstrated more of a reliance
on cognitive abilities. The CN treatment resulted in
improvements in tasks requiring a broader range of
abilities, some generalization in discourse tasks requiring
the most manipulation of cognitive and linguistic abilities,
and less dependence on cognitive abilities.

Possible Influences of Cognitive-Linguistic Ability

The subjects with the lowest cognitive profiles made
the greatest magnitude of change in the functional training.
Subjects with the best cognitive profiles tended to make the

broadest improvements on several measures, including

142

SI

pa

SC
If

16

c:
c]
t:
ft

PC

me

im

on

C0

discourse tasks, in the CN treatment (e.g., CN1). However,
subjects with mid-range cognitive abilities also showed this
pattern of improvement in response to CN treatment.

The subject with the best cognitive profile made
somewhat broader improvements than other subjects in their
respective groups. Response to CN treatment tended to be
less reliant on cognitive ability overall. Subjects with
mid-range cognitive profiles varied in performance. This
pattern of response to treatments suggests that aphasia type
and severity are not the only characteristics that may be
critical when making treatment decisions. Furthermore,
clients with more cognitive impairment may require greater
treatment time to achieve goals. This observation should be
further considered when making treatment decisions.

Possible Explanations for Patterns Observed

Two aspects of the functional treatment outcomes are
not totally consistent with original predictions. It was
surprising to observe that the functionally trained subjects
improved on standardized measures of oral naming, and less
on CADL-Z, than expected. The broad oral naming improvement
could be due to an effect of the type of contexts in which
most of the training occurred. A large percentage of the
total therapy involved conversational contexts, and this has
previously been reported as being associated with a
generalized improvement in naming (Holland et al, 1983) as
measured by standardized batteries. In contrast, other

investigations of functional treatment outcomes have shown

143

an improvement on QADL—Z performance in the absence of
improved standardized test or naming performance (Aten et
al, 1982). The mixed pattern of results could be
attributable to individual differences in cognitive and
linguistic abilities across these studies; unfortunately,
cognitive abilities were not assessed in previous work.

The potential role of cognitive abilities and the
omission of cognitive description information in previous
work is particularly important since in the present study,
subjects with more severe cognitive abilities were more
likely to show improvements on CADL-Z after functional
treatment. It is possible that both CADL-Z and oral naming
improvements are outcomes of functional treatment, depending
on the characteristics of the individuals enrolled in the
treatment.

Functional treatment has been often discussed as
relying on cognitive skills for the development of multi-
step strategies which require cognitive resources to deploy
(Aten, 1986; Chapey, 1988). It is necessary for individuals
using such strategies to recognize the need for the
strategy, recall the entire sequence of the strategy, and
implement it. Functional treatment has also been
characterized as stimulating real world meanings as inputs
and outputs of the training context.

All of the subjects in the present study had sufficient
cognitive abilities to learn these types of strategies and

use them. However, subjects with lower cognitive profiles

144

required more time to learn and master them. Once mastered,
these subjects performed these strategies reliably across
multiple contexts. This accounts for their improvement on
QADL-Z in addition to improvements in oral naming and phone
ordering.

The cognitive resources involved in use of these
strategies are evidenced by the functionally trained
subjects’ response to load in the catalog ordering task.
These subjects experienced a disruption of performance in
the concurrent task condition compared to their performance
in quiet that reflects the effortful processing required in
this task. The functional training appeared to result in an
even greater difference between performance in quiet and
with the concurrent task at posttest than at pretest.
Communication strategies targeted during functional training
required general cognitive resources to implement. It is
unclear whether, with additional training, these strategies
would have become less effortful and more automatic.

The apparent improvement of the semantic and language
system in response to CN treatment is also supported by
pattern of performances under conditions of load. CN
subjects displayed little difference in performance between
quiet and concurrent conditions of catalog ordering task at
posttest. This pattern of performance was equally observed
in the written version of the task on which they had notably
improved. This suggests that their improvement did not

depend on effortful processing, but rather, that the

145

language system became more efficient in a manner that was
more automatic. The large magnitude of improvement on tasks
that required the greatest cognitive resources, such as the
concurrent task condition, role playing in a variety of
tasks in QADL-Z, and the narrative discourse task which
required telling a story from memory, also tends to support
this hypothesis.

Although the wide degree of variability among the
subjects’ discourse performances precluded any particular
conclusion regarding pattern of generalization to discourse
based on treatment type, the results are suggestive. The
large magnitude of improvement for three of the CN subjects
in the narrative task, and the large improvement for three
of the functional subjects in the catalog ordering
procedural discourse, might suggest a pattern that is
consistent with the results of standardized testing.
Improvements in the catalog ordering procedural discourse
task among the functionally trained subjects were due to the
use and generalization of trained vocabulary from training
sessions to the discourse context. Improvements in the
narrative discourse among some of the CN trained subjects
might reflect the richness of the sampling context, which
provides the opportunity for demonstration of the breadth
and flexibility of the improvements made in response to the

CN treatment.

146

Results of this study should be interpreted with
caution due to the size of the sample and the likelihood
that the sample may not be representative of the population.
Although every effort was made to obtain a homogenous sample
based on characteristics deemed important, there was still
quite variable performance on many measures among the twelve
subjects. Furthermore, the analysis of multiple variables
requires sufficient sample size and increasing the study
group will result in the statistical power required for
additional analysis. Small sample size increases the
likelihood of a Type II error, falsely accepting a result of
no difference. Therefore, it will be imperative to replicate
these results with a larger sample.

In addition to the replication of the results with a
larger sample of subjects, additional future questions
should be addressed. This project addressed the broad,
conceptual issues underlying two different treatment
approaches. Future research should attempt to identify the
specific components of each approach that result in
differential outcomes. The amount and duration of treatment
required to obtain the response also needs to be identified.
Finally, the interaction between these treatment approaches
and different types of aphasic subjects will need to be
further addressed.

Overall, the project design and results are promising
first steps to answer questions about the most effective

therapy for particular individuals. Theory can and should

147

serve as a basis for making predictions about the underlying
mechanism and action of a treatment, and the cognitive-
linguistic characteristics of individuals who might benefit
the most. It is from theoretically driven clinical research
that we will be able to address the issues of treatment

effectiveness and efficiency that remain.

148

APPENDICES

149

APPENDIX A

WH- QUESTION ASSESSMENT TASK

Instructions to Subject: “I will be showing you some
pictures. For each picture, I will ask you a question.
Point to the item that best answers each question.” A
single repetition of each item is allowed if requested by

the subject.

 

Stimulgg Target Resnogsg
1. What color is her blouse? Yellow
2. What is he touching? Ear

3. What is the item around the doctor’s neck? Stethoscope

4. What sound does the doctor hear? Heart

5. What is the doctor writing? Notes

6. What is on the door? Sign

7. What is on the stove? Pan

8. What is next to the bowl? Salt/pepper,
milk

9. What is in the bowl? Fork

10. What is in the pan? Butter

11. What do you put the omelet on? Plate

12. What is next to the omelet? Tomato

150

APPENDIX 3

SUMMARY OF THE FORMULA FOR CALCULATING SOCIOECONOMIC STATUS

BASED ON THE FOUR FACTOR INDEX (HOLLINGSHEAD, 1975).

Occupational Score

9 =

(I)
I

Hmwbmmq
II II II II II II II

Higher executives, proprietors of large
businesses

Administrators, proprietors of medium sized
businesses

Smaller business owners, managers
Technicians, semiprofessionals

Clerical and sales workers

Skilled manual workers, craftsmen
Machine operators and semiskilled workers
Unskilled workers

Farm laborers

Education Sccre (level of school completed)

wabmmq
II II II II II II II

Formula

Occupation score x 5 =

Graduate degree

University graduation

Partial college (at least one year)
High school graduate

Partial high school

Junior high school (9th grade)

Less than 7th grade

Education score x 3 =

 

Total:
Ran e com t scores
Social Strata Range of scores

1 Major business and professional 66 - 55
2 Medium business, minor professional 54 - 4O
3 Skilled craftsmen, clerical, sales 39 - 3O
4 Machine operators, semiskilled workers 29 - 2O
5 Unskilled laborers 19 - 8

151

APPENDIX C

PROTOCOL FOR THE GERIATRIC DEPRESSION SCALE - SHORT FORM
(GDS) (PARMALEE & KATZ, 1990)

NMNE/ID DATE

 

EXAMINER

 

PRESENTATION’MODALITIES: Visual only Auditory only
visual + Auditory

Please circle the answer that best describes how you have

been feeling recently. You may read the questions silently
to yourself, have them read to you, or both.

1. Are you basically satisfied with your life?

Yes No
2. Have you dropped many of your activities and interests?
Yes No

3. Do you feel that your life is empty?
Yes No

4. Do you often get bored?
Yes No

5. Are you in good spirits most of the time?
Yes No

6. Are you afraid that something bad is going to happen to
you?
Yes No

\1

Do you feel happy most of the time?
Yes No

8. IDo you often feel helpless?
Yes No

59. Do you prefer to stay at home, rather than going out and
doing new things?
Yes No

10. Do you feel you have more problems with memory than

most?
Yes No

152

11. Do you think it is wonderful to be alive now?
Yes No
12. Do you feel pretty worthless the way you are now?
Yes No
13. Do you feel full of energy?
Yes No
14. Do you feel that your situation is hopeless?
Yes No
15. Do you think that most people are better off than you
are?
Yes No
Sccring
Score (# of negatives)
Normal range 0-5

Mildly depressed 6—10

Severely depressed over 10

153

APPENDIX D

FACE SHEET FOR THE COGNITIVE BATTERY (FROM VAN MOURIK ET AL,
1992)

COGNITIVE BATTERY

SUBJECT ID DATE

 

EXAMINER

 

Global Aphasia neuropsychological Battery

 

 

Task Raw Score %correct
1. Visual Cancellation Task
2. Subtests of the Rivermead Behavicral

Memory Test1

Facial Recognition Test

 

 

Object Recognition Test

Total for RBMT subtest

3. Raven’s chourcd Progressiye Matrices2

 

 

 

 

 

4. Dcuelcpmcntal Test of yisual Perception3

General yisual Perceptual sum of subtests

5. Nonverbal Auditory Recognition

Environmental sounds from QADL‘

 

Wilson, Cockburn, & Baddeley, 1985
Raven et al, 1979

Hammill, Pearson, & Voress, 1993
Holland, 1980

bUN-fl

154

Memory Tasks
6. Digit span (from PALPA“

7. Pointing span (from PALPA)

 

5

Kay, Lesser, & Coltheart, 1992

155

 

 

APPENDIX E

TRANSCRIBED SCRIPTS FROM THREE ACTUAL CALLS MADE TO NATIONALLY
KNOWN CATALOG ORDERING COMPANIES BY THE EXPERIMENTER.

m l 1

Thank you for calling xxxx. This is Bernice. How may I help
you?

May I have the C number above your name on the back of the
catalog?

May I have your phone number starting with the area code?
May I have the first item please?

And the color?

And the size?

OK that’s the plaid turtle neck.
Your next item please.

May I have your credit card number?
And the expiration date?

Your name as it appears on the card?

OK the total comes to $26.50 including shipping and your
turtleneck will be delivered Thursday this week.

Sample £2

(recorded) Please make your selection from the following
items: To place an order, press 1, tom.

Chris may I help you?

Your home phone number with the area code first?
Your first name?

Your last name?

Spell it please.

And your address?

156

And the city?

And the zip code?

Do you have a xxxxx’s charge?

Do you have it with you?

What is the number?

The catalog item number?

And the size?

And the color number?

OK that’s the taupe pantsuit.

Did you want to use your xxxxx’s charge?
Did you want it delivered to your home or the store?

This item will be available on Thursday after 5 p.m. Remember
to bring your xxxx's charge with you.

am 1
This is Jeannie. How may I help you?
Will you be using your credit card?
Can I have the spelling of your last name?
May I have your billing zip code?

May I have thee“S” number — “s” as in “sam? — on the back of
the catalog?

You live in Ypsilanti, Michigan?

May I have your billing address?

Does your address include an apartment number?

Will you be shipping to that address?

May I have your home phone number starting with the area code?
May I have your credit card number?

The expiration date?

May I have your first item number and the quantity?

157

Ms. Hinckley, that catalog has some daily specials in it.
Would you like me to tell you about those?

Thank you. Your total today is $26.50, that includes the
shipping and handling.

Your package will arrive in 10—14 business days and your
credit card will be billed when the order is shipped.

Thank you for calling xxxx. Have a nice day.

158

APPENDIX F

PROTOCOL FOR THE CATALOG ORDERING TASK, PHONE AND MAIL ORDER

FORN‘VERSIGNS

 

159

TOTAL TIME

PHONE ORDER TASK
PRE/POST ASSESSMENT

 

 

 

SUBJECT ID DATE EXAMINER

TEST TIME: PRE POST

CONDITION: QUIET TONE

ORDER OF PRESENTATION (RUN NO.): B 1 2 3 4 5

TARGET ITEM:
(randomized from list)
COLOR:
SIZE:
PRICE:
INCORRECT ITEM:
INCORRECT PRICE:

 

 

 

 

 

 

CREDIT CARD NO:1 2 3 4 5 6

Equipment; RT Task, video recorder setup, stopwatch.

Task Materials Required: Catalog, pen and scratch paper, play
credit card, phone.

Instructions tc Examiner; Read each item with a normal rate

of speech and with normal intonational contours. Your job is
to sound as much like a phone salesperson as possible. Read
each item naturally, allowing 20 seconds after the item for
the subject to respond. If the subject has not responded
after twenty seconds, read the item a second time. Allow an
additional 20 seconds for a response. If the subject still is
unable to respond, then move on to the next item. Repetitions
of any item are allowed if requested by the subject, but this
should be noted in the response column. If the subject
responds incorrectly, move on to the next item.

Instructions tc Subject;

“We are interested in how well you can do activities that
you would normally do at home. We are going to pretend to
make a purchase from a catalog using the phone. Here are all
the things you will need: a catalog, a pen, a credit card,
and a phone. We are going to do this a few different times,
each time ordering a different item. Occasionally you will be
doing the tone task in addition to the ordering task. Some
parts of this may be difficult for you now, but just do the
best you can.”

160

 

 

wumnfid:

ucwfiunomm has
mcfloSHoGH mmwuccm
mcfiHHHQ H50» mo
ummuum one HmQEdG
mzu ma Dong .m

 

mommmam .mmeUUm
mGflHHflQ H50>

mo mumum can huflo
man we boss .5

 

Ramayana mcaaaan
H50» mo wcoo

man was me “as: .o

 

mmcwaammm
mgu one was: umnflm
udom ma umn3 .m

 

mmmmwam . GHHHQO
one can mess ummH
HSOM ma DMQB .o

 

mmcoo owns
on» nufi3 mcfluumum
Hwnfidn OGO£Q $803
MSG» me OMS: .m

 

mmmawumo may “0
xomn ms» no menu
wsu m>onm Hwnfidc

was ma page .N

 

mach mam:
H has .mcflaawo
How 50% xnmse .H

 

 

OHOUW

 

sans
couguau

 

oanonuom uoofindm

 

59H “Gag

 

 

161

 

.Aucdofid
Domuuooawv

ou mmfioo H0000
Hmpuo snow .bH

 

mNuwucan
0:0 one Emufl axon
H90» me 0033 .ma

 

mnfl Donn
mxfla 20% UHDOS

mafia pas: .mH

 

mwxfia

UHDOS 50% xcHnu
00% 00 HoHoo Dons
.Afimufl uomuuoov
men 0.00:»

u Shuom E.H .va

 

.Afimufi Domuuoocflv
m m.umnu .xo .mH

 

wmuflucmsw man can
H0985: Edna umuflm
H50» me 0033 .NH

 

AuumflQSm

ou xomn mumo
coaumnfimxw one
Hmnada choc ufiomno
unmomme .HH

 

wwmmmflm .mumo
cowumuflaxm on» was
nomad: choc uflvmno
950% me 00:3 .oa

 

 

 

 

 

mmmfl> Ho
UnmonmumME .mcflm:
on 50> Hafl3 choc
become 0033 .m

 

 

162

TOTAL TIME
MAIL ORDER TASK
PRE/POST.ASSESSMENT

 

 

 

 

 

 

WRITTEN
SUBJECT ID DATE
EXAMINER
TEST TIME: PRE POST
CONDITION: QUIET TONE
ORDER OF PRESENTATION: B 1 2 3 4 5 6
TARGET ITEM (randomized from
list):
COLOR:
SIZE:
PRICE:

 

CREDIT CARD NO.:1 2 3 4 5 6

Equipment; RT Task, video recorder setup, stopwatch.

Task Materials Required; Catalog, pen and scratch paper,

play credit card, calculator.

Instructicns tc Examiner; Read the instructions below to
the subject. Keep track of the total time for task

completion. You may NOT read aloud any item on the order
form, unless the subject specifically requests you to read a
word out loud. You may NOT, under any circumstance, explain
the meaning of any word to the subject, nor may you provide
any feedback to the subject regarding the accuracy of what
they have written. If the subject indicates that he/she
does not know what to write, say “Skip that item if you need
to. Just do your best."

Instructions tc Subject;

“We are interested in how well you can do activities
that you would normally do at home. We are going to pretend
to make a purchase from a catalog using the phone. Here are
all the things you will need: a catalog, a pen, a credit
card, and a phone. We are going to do this a few different
times, each time ordering a different item. We may be
asking you to do this task while listening to tones and
pressing the foot pedal. Some parts of this may be
difficult for you now, but just do the best you can."

163

SPECIAL CLOTHING COMPANY
1—555-5252
East Lansing, Michigan

Ship to:

 

Last name First name

 

Address

 

Apartment number

 

City State Zip code

Daytime Phone Number: (
Area code

)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Item. Size Quanti Color Description Price Total
number ty
Subtotal
Shipping/Handling
Total
Payment By: Check AMEX
Visa Mastercard

Card Number:

 

Expiration Date:

 

Signature:

 

Your merchandise will be shipped within 24 hours of receipt
of your order, and should be received within 10 days. If

you are in a hurry, please contact us for rush delivery
charges.

164

 

Qral (Phcne)

APPENDIX G

RECEPTIVE VOCABULARY USED ACROSS THE ORAL AND WRITTEN
VERSIONS OF THE CATALOG ORDERING TASK

help

catalog
above

name

back

home

phone number
starting
area code
last name
first name
zip code
billing
address

city

state

street
apartment number
credit card
expiration date
first

item number
quantity
name of clothing article
color

size

next

package
arrive

10 days

ship

Visa, MC
order

written (Order Form)

name

phone number

area code
last name
first name
zip code

address

city

state

street

apartment number
credit card
expiration date

item number

quantity

color

size

package subtotal
arrive total

10 days payment
shipping check, AMEX,
handling signature
description price

165

APPENDIX H

TARGET STIMULI OF THE CATALOG ORDERING TASK FOR

PRE/POST ASSESSMENT

166

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m b . vm 9.1mm o m . m m mdqbwmm 3044M? mm KNEE . w
om . HH BMHmm om . b MNHm MZO 9mm mm Hg .m
oo.om mega om.mH mafia Emma mm 880 .v
oo.N¢H munkmgm oo.mm OH 248 mbH MHONM .m
mm. . mm mmomm om . hm mddbwmm MUfinm mm. ugh . N
m b . mo 9400 om . m N fiquvmm MDAm me BMHNM . H
Afom +v
flUHmm BUHEKOOZH SHBH BUMNMOOZH NUHMm HNHM KOQOU .02 flaflm

 

 

92m muffin/dad mama magnum

mZMBH BWUMSH. mL/HWS

167

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm. . vm 8% cm . mm MEADQWM 3044?». Nw fig . w
om . Ha BMHmm om . b MNHm. WZO 9mm mm Edd .m
00 . om Wham om . ma WOMEN Mqu Na 9400 . .v
m b . m m mmedmgm om . m N m MU<Am mvH mflONm . m
mb.mm mmomm om.¢v NH med0 on. man .N
om.bN B<OU om.mH M<QDUMM ><m0 ow BNHMM .2"
Afom+v
HOHmm BUHKMODZH ZHBH EUHKKODZH HUHMm HNHm “CHOU .02 Hmflm

 

 

Dzm m.DZ¢A mama mag—Um

mZWBH Bmwmdfi w .ZMSOB

168

APPENDIX I

TRANSCRIPTS FROM ACTUAL CALLS MADE BY THREE ADULTS WITH

APHASIA TO CATALOG ORDERING COMPANIES

am 1 l

Salesperscu Adul with h i 'e l
H011y may I help you?

Hi How are you doing? Marge Naber,
I’m stroking, I understand everything but slow down please.
I’m sorry, get it?
Hmmm.

OK, um, um, xxx’s please, OK, right
it out please, let’s see (counting to herself) 9 4
What are you giving.me ma’am your phone number or what?

Umm
Do you wish to place an order?

Yes please.
What is the area code of your phone number?

1 2 3 4
The area code of your phone number ma’am

Um, Cross please, Cross please
Cross? What do you mean ma’am?

Well that’s wrong again. I guess
so I can’t say it. Pardon me I'm sorry thank you very much

thank you very much. Bye.

169

Salesperscu Adult with aphasia (Subject 2)
This is Cathy may I help you?
Hi could I have um um one..um T—
shirt.
What would you like?
The T—shirt.
Oh. OK first I need your phone number with the area code.
OK um m603 xxx-xxxx (correct).
OK. And your name please.
Alexis
And the last name.
Kxxxx (correct).
OK and is there anyone in the home that usually orders?
Uh no.
Is there are you the head of the household?
Yeah
The only we ask is people don’t like so many catalogs coming
to their home, so I have an Arlene showing here.
Really? My um mother.
So that would be your mom.
Yeah.
And she lives there doesn’t she?
Yes.
OK so now on this order, do you mind if I just leave it in
Arlene’s name or do you want me to change it to your name?

No that’s fine, Arlene is fine.

170

OK could you verify the address and zip code.
Um again?
I don’t think you told me the address and zip code did you?
Oh OK OK. Um, 130 xxx Drive,
Manchester, NH.
And the zip code is 02134?
Yeah.
Do you want this order to come to your home or the store?

Um I think um my house.

 

If it comes to your home, do you have a charge card to put
it on?

Yes.
OK. For home delivery? Are you under 18 or over?

23
26?

23.
Oh 23 you know how to do things. May I have the catalog
number of your first item?

OK um I want um embroidery T.
OK, the number is on the plate there should be a letter in
front of seven numbers. Do you see that anywhere?

Yeah it’s um and the numbers?
Yeah.

It’s um 6 or um 8 8 4 2 dash 7 6 9
2.
And the first letter was a C like Charles?

Um..

171

At the beginning of the number?

UmmC or um Carol, yeah.
OK and what size?

At um Misses’ medium
Misses’ medium, and which color?

At um the 37 navy um stripe.

OK a navy stripe medium T shirt.

Hmmm.
OK and how many?

Just one.
OK and the next item?

That’s it.

And which credit card would you like to use?
Um, visa.
And the visa number?
Or um xxx’s.
OK you’d rather do xxx’s.
Yeah.
OK and the xxx’s number?
Um xxx xxx xx x xx (correct).
OK and oh boy after we did all that I found out the T shirt
in that color is not available.
Oh.
we only have the ocean stripe in that size. Is medium the
only size you can try?
‘Um OK it’s other color is the same

thing this time?

172

Do you like that color?

Sure.
OK good. I’ll put you for that. OK good, then an ocean
stripe medium T-shirt will be coming to you. The total is

$12.49, and thanks for shopping xxx’s catalog.

Salesperscu Adult with aphasia (Subject 3)

This is JEan may I help you?

I wanna place an order please.
OK what is your home telephone number with the area code
please?

xxx-xxx-xxxx (correct).
And your first and last name?

David Dxx, D - X —X (spelled
correctly).
Alrighty and for verification your street address and zip
code.

323 Cxxx Rxxx xx.
OK and for verification your zip code?

43xxx (correct).
Thanks and the catalog number of your first item.

Uh, 3504227D.
The first letter is C as in Charles?

Yeah.
OK what size?

Extra large.

173

Color?
Black.
H0w.many?

One.

OK that’s a pocket tee shirt in the extra large black.

.your next item?
Nothing else.
OK is this cash or charge?
Uh, charge.
XXx’s?
Yeah.
Account number?
Xxx xx xxxx (correct).
OK delivered to your home or office.

Home.

And

OK unfortunately that item is not currently available in

the black I only have navy or white.

Navy please.

OK and I have that on order. That will be delivered to your

home in 3—4 working days with a shipping charge of $2.50,

and thank you for shopping xxx’s catalog.

Alright. Thank you.

174

APPENDIX J

SCORING GUIDELINES AND EXAMPLES FOR THE CATALOG ORDERING

TASK
Catalog Ordering Task
Scoring Criteria/Examples
Score. Characteristics

 

O No Response (NR)
"I don’t know" (IDK)
Unintelligible response
Completely unrelated to task
Completely inappropriate
Related to task, but not to item
answers "blue" when asked for size

 

1 Related to item, but incorrect

says "red"for'1blue"
Partially correct

Misses one digit on credit card number
Incomplete Gives
credit card no. but not expiration date
Correct information provided in inappropriate
modality Says Wmastercardilbut doesn’t circle it
on written form

2 Communicates message completely

Appropriate to question/item
Judged to be effective in real ordering situation

175

 

c I .uG0EuH0Q0. Hmmm.
N I.AH20 0E0: u00Hum 00>ww
N I u00Hum hum H0n55= m0>ww

NH0QE55

u508uummo ham
mcfio5aonfl 0005000
GGAHHHQ 550% m0
u00Hum 050 509855

was we was: .m

 

c I 0050Q00H 0~Qflowmm0ucwflb
H I Aflco «00000 H0» Nuwo m0>ww
N I.%Huo0hhoo 0u0um onm.hufl0 m0>ww

w0000HQ .0005000
mcflaafln 550%

m0 0umum paw NDNU
0£u we Dong .b

 

Q I sumgu mkwmfifi “snow H:
H I uwowo Newmmflz
N I.A~uo0hhoo 0000 QHN m0>wm

mmmwueoa manafian
“50% m0 0000
new was ma page .m

 

c I ma
H I NGNHH0Q0 05 Dan 0505 m0>w0
N I uo0hhoo NQNNH0Q0 hum 0802

wmcflaa0mm
0£u one 0E05 umufim
550% me Dong .m

 

o I ma
H I NGNHH0QW on non 050: 00>HO
N I uo0hhoo NGNHH0QW bum 0Emz

m0000HQ .mGflHH0Qm
030 050 0805 umMH
550% we u0£3 .v

 

o I mm0hoo0 m0>ww

HI

0o00 00km 0: non b00MH00 H0QE5: 050am
N I 0o00 00km D50 H0QE55

020£Q «mwmsuoc HON» Q5 0o08.H0 ~05u0<

m0ooo 0050
0£u Suez mcaunmuw
H0DE55 050£Q 0&0:

H50N we D033 .m

 

c I 305¥ p.moo H
H I 50£050> NODUHS :52

N I NvamNVNH

mmoamumo 0:0

No #002 030 50 0805
0Q» 0>0Q0 H0QE55
0:» ma owns .m

 

o I m2
H I £00m
N I M0DH0 50 000~Q.ob 0Mw~ D.H

m50h QHO£
H >02 .mcflaamo
How 50% x5039 .H

 

 

OHOOW

 

GEAR
Donnddfl

 

uOHQBdXH ounonnom unannflm

 

EouH Hung

 

 

176

 

Q I 2M0:
m I 08 0888

05H58%50 80 ..050H8 005 0.0050. ..05.

. AUGDOEM UUGHHOUCflV
0U mmEOU HMUOU
HGUHO H50? . PH

 

o I 00505008 02

m%0H050mU
050 050 800H 0805

 

 

N I 0H 0MHH 05H58%50 80 .00H0_05H5005. 850% 08 00:3 .0H
0 I .80H5m08. 80% .05HQ. 0%00 m5H

H I 0NH0 00000800 005 055 0000H08 0%00. 0050 0xHH 50% 0H503
N I 0NH0 00000800 0%00 08H0 0053 .mH

m0xHH 0H503

50% #5850 50% 00

c I .05HQ. 80% .0080? 0%00 80.30 0053 . A800H

H I .05HQ. 80% .50080. 0800. 00088000 050 0.0050

N I 80H00 0008800 0%00

I %8800 8.H .vH

 

o I 00505008 02
N I «800H 0008800» 0 0.0050 05.
80 ..o5. 050%00 %5 805H80x0 00008800

.A800H 00088005Hv
0 0.0050 .MO .MH

 

o I .08H50. 0%00
H I.%H50 %0H05050 0%00
N I %0H05050 050 805855 E00H 0%00

880805020 000 0:0
805855 8008 008Hm
850% 0H 00:3 .NH

 

o I 005mm008 02

HI

00088005H 0H 805855 055 .0008800. 0%00
N I 80880 00008800 80 .0580

0H 0050 050 .0008800. 80 .00%r 0%00

A000nn50 00 £005
0000 50H008me0
050 805855 0800
080080 0005000 .HH

 

o I 0805855 00088005H 00>HU

H

I 0000 50H008HQX0 05 055 805855 00>H0
N I %H000H5800

0000 503080.90 050 800855 00>H0

~0000H5 .0000
508008H580 000 050
800855 0800 0H0080

850% 08 0053 .OH

 

 

 

 

o I .0800 0H0080. 0x80

H I 00805 00H>

80 .00Hb. No 000005H .0800800008. 0%00
N I %H0008800 0050500m

 

~008> 80
0800800008 .m5H05
0Q 50% HHHS 0800

uflnmuo 0082 .m

 

 

177

 

 

 

 

N I swag:

o I 00505008 02
3.80.. .. 30% 85050..

 

.mcHHHmo

800 50% #5039
.00QQH00 0H 80080
050 5053 00HHHQ

0Q HHHS 0800 080080
850% 050 0%00 0H 5H
0>H880 HHHB 000x005
850% .m5HH0505

050 m5HQQH£0 05HQ
.A055080 0008800v
0H0000 80080 850%

I %8800 8.H .mH

 

 

178

APPENDIX K

LIST OF THE 54 BASELINE STIMULI FOR THE CN TREATMENT,

DERIVED FROM RESPONSES ON THE CATALOG ORDERING TASK

CQLQRS
BLUE, BLACK, GREY, PINK, PURPLE, RED
NUMBER NAMES
ZERO, ONE, TWO, THREE, FOUR, FIVE, SIX, SEVEN, EIGHT, NINE
LETTER NAMES
ALL 26 LETTERS A THROUGH Z
L THI NAME
COAT, SHIRT, PANTS, SOCKS, SKIRT, DRESS
EDIE

PEACH, PINEAPPLE, ORANGE, APPLE, PEAR, BANANA

179

APPENDIX L

 

 

 

DATA RECORDING FORM FOR CN TREATMENT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subject ID Date
Session #
Training Cue I
Probe (no cues)

1 Target Rgponse Ac curacx
2.
3.
4.
5.
6.
7.
8.
9.
10.

Total

 

180

 

Training (with cues and feedback)

1 - First trial; 2 - Second trial; 3 I Third trial

 

Target RB sponse AC curacy Cue Repet it ion

 

Set 3

 

1.

 

2.

 

 

 

 

 

 

 

 

 

10.

 

 

 

 

 

Total

 

181

 

APPENDIX M

LIST OF THE TARGETED ITEMS FOR PRODUCTION DURING FUNCTIONAL

TRAINING

CQLQR NAMES

BLUE, BLACK, RED

NUMBER NAMES
ZERO, ONE, TWO, THREE, FOUR, FIVE, SIX, SEVEN, EIGHT, NINE

LETTER NAMES
LETTERS REQUIRED TO SPELL FIRST AND LAST NAME
THI AME

COAT, SHIRT, PANTS

EEBSQNAL INFQRMATIQN
NAME, STREET ADDRESS, CITY AND STATE, ZIP CODE, PHONE NUMBER

182

APPENDIX N

DATA RECORDING FORM FOR FUNCTIONAL TREATMENT

183

 

momma—o .300 00008580
05 050 00055: 008
0020 5.0% mm 0055 .3

 

new;
00 080008008 .wEms on so»
=05 0000 000.6 00:3 .o

 

80008:: 00085050
H50 @5022: $2000
wEEb 050% 00 080m

050 0008:: 20 mm 855 .m

 

Somme—o £02000
waif 50% Co 880
05 00 20 a an? s

 

3.02000 wizB 50%
00 0000 EN 05 8 08:5 .0

 

NEE—.050 05 0:0
058: 00.5.0 050» 8 8:? .m

 

woman—o £5:on 20 0:0
058: 090. 000% 8 00:3 .0

 

«“0000 00.8 20
£05 wfitfim 000558 28:5
050: 050% m0 8:? .m

 

sumo—800 20 00 0.000
20 :0 080: 05 3000

sass: 2: s as; .8

 

8.00% Eon _ .32
05:00 000 00% 0505. A

 

 

SnoEEcUflfionv

850.08% Leona—5

 

 

Evan hvflmsgm—

 

 

0025020 00000080 .50 050.0 $000002 Sea

184

 

05:00 00: 50% 000:5.
00050:... 3 00000 0.:

00:3 00:3 00 :05 0000
00000 0:0.» 0:0 0:00 o:

E 03.50 :05 $00—00: 50%
05:05: 000 widow? 8::
600080 00503 00000
000.8 000% - H.500 8.: .m:

 

600080 0005000: 00
00800 00000 000.8 050% .5

 

0:40:05: 05 050
Eu: :8: 000% m: 00:? .0:

 

0:: 00:0 00:—
00.: 0:53 000 00:? .m:

 

000:—

0303 :0.» 0:50.: 50% 00
8.8 02.5 .85: .8803
05 0.00.0 - %tom 8.: .3

 

.058
000.5000: 0 0.00.: .MO .2

 

0.305050 05 0:0 00:55:
5.0: 000: So: 0: 00:? .N:

 

 

 

 

900.300 00 0.000
0000 00000098 000 00055.:

0000 :0000 000.50% .2

 

 

185

BIBLIOGRAPHY

186

Bibliography

Ackerman, P. L. (1988). Determinants of individual
differences during skill acquisition: Cognitive

abilities and information processing. qurnal pf

Experimental Psychology: General, 117, 288-318.

Ackerman, P. L. (1990). A correlational analysis of skill

specificity: Learning, abilities, and individual

differences. qurnal pf Experimental Psyghplpgy:

Learning, Mempry, and Cpgpitigu, l5, 883—901.
Ackerman, P. L. (1992). Predicting individual differences

in complex skill acquisition: Dynamics of ability

determinants. Journal of Applied Psygholpgy, 11, 598—
614.

American Speech-Language-Hearing Association. (1985).

Guidelines for identification audiometry. ASHA, 27,

 

49-52.

Aten, J. L. (1994). Functional communication treatment. In

Chapey, R. (Ed.), Lapguage Interyenuipn Strategies in
Adult Aphasia. 3rd Ed. Baltimore, MD: Williams &

Wilkins.

187

Aten, J. L. (1986). Functional communication treatment. In
Chapey, R. (Ed.), Lan a In n ion t ' s in
Adult Aphesia. 2nd Ed. Baltimore, MD: Williams &

Wilkins.

Aten, J. L., Caligiuri, M. P. & Holland, A. L. (1982). The
efficacy of functional communication therapy for

chronic aphasic patients. rn f h an

Hearing Disorders, 41, 93-96.

Avent, J. R., Edwards, D. J., Franco, C. R., Lucero, C. J.

(1995). A verbal and non-verbal treatment comparison
study in aphasia. Aphasielegy, 2, 295-303.

Basso, A., Lecours, A. R., Moraschini, S. & Vanier, M.
(1985). Anatomo—clinical correlations of the aphasias

as defined through computerized tomography: On

exceptions. W. 15, 201-229.
Battig, W. F., & Montague, W. E. (1969). Category norms for

verbal items in 56 categories: A replication and

extension of the Connecticut category norms. Journal

pf Experimental Psyehelegy, 8Q, 15—46.
Beeson, P. M. (1993). Memory impairment and executive

control in individuals with stroke—induced aphasia.

Brain and Languege, 35, 253—275.

188

Bernstein, S. E., & Carr, T. H. (1996). Dual—route
theories of pronouncing printed words: What can be
learned from concurrent task performance? qurpal of
Experimental Payehology: Learning, Memory, and
Cognition, 22, 86-116.

Bloom, L. & Lahey, M. (1978). Language Develepment and

Language Disorders. New York: Wiley & Sons.

 

Brink, T. L., Yesavage, J. A., Lum, 0., Heersema, P. H.,

Adey, M., & Rose, T. L. (1982). Screening tests for
geriatric depression. li ' r ' , 1, 37—
43

Brookshire, R. (1983). Subject description and generality

of results in experiments with aphasic adults. Journal

pf Speeeh and Hearing Diserders, 48, 342-346.
Byng, S. (1995). What is aphasia therapy? In Code, C. &

Muller, D. (Eds.), The Treatment pf Aphaeia; From
Theory te Practice. San Diego: Singular Publishing.

Byng, S. (1994). A theory of the deficit: A prerequisite

for a theory of therapy? Clinieal Aphasielegy, g2,
265-273.

189

Byng, S. & Black, M. (1995). What makes a therapy? Some

parameters of therapeutic intervention in aphasia.

Eurppean Journal of Diserders of Communieation, 3Q,

303-316.
Byng. S., Kay, J., Edmundson, A. & Scott, C. (1990).
Aphasia tests reconsidered. Aphaeiplpgy, A. 67-92.
Calvanio, R., Levine, D., & Petrone, P. (1993). Elements

of cognitive rehabilitation after right hemisphere

stroke. Behavioral Neurology, 11, 25—57.

 

Cermak, L. S., Stiassny, D., & Uhly, B. (1984).

Reconstructive retrieval deficits in Broca’s aphasia.

Brain and Language, 21, 95-104.

Chapey, R. (1986). Cognitive intervention: Stimulation of

cognition, memory, convergent thinking, divergent
thinking and evaluative thinking. In Chapey, R. (Ed.),

n n i r t 1 in Ad A h ia.

Philadelphia: Williams & Wilkins.

Chapey, R. (1988). Aphasia therapy: Why do we say one

thing and do another? In Gerber, S. E., & Mencher, G.

T. (Eds.), Interpatienal Pexspeetiyes pp Cpmmunieation
Dieprders. Washington, DC: Gallaudet University

Press.

190

Cherney, L. R. (1995). Efficacy of oral reading in the
treatment of two patients with chronic Broca’s aphasia.

2! 57-67.

Topics in Stroke Rehabilitation,
Cohen, J., MacWhinney, B., Flatt, M., & Provost, J. (1993).
PsyScope: An interactive graphic system for designing
and controlling experiments in the psychology
laboratory using Macintosh computers. Behayior

Researoh Methods. Instruments. and Computers, 25, 257-
271.

Coltheart, M., Bates, A., & Castles, A. (1994). Cognitive

neuropsychology and rehabilitation. In Riddoch, M. J.,

& Humphreys, G. W. (Eds.). ' iv h
and Cogpitiye Rehabilitation. Hillsdale: Lawrence
Erlbaum.

Craig, H. K., Hinckley, J. J., Winkelseth, M., Carry, L.,
Walley, J., Bardach, L., Higman, B., Hilfinger, P.,
Schall, C., & Sheimo, D. (1993). Quantifying

connected speech samples of adults with chronic

aphasia. Aphasiology, 1, 155—163.
Darley, F. L. (1972). The efficacy of language

rehabilitation in aphasia. Journal of Speeoh and

Hearing Disorders, 31, 3—21.

191

Denes, G., Perazzolo, C., Piani, A., & Piccione, F. (1996).
Intensive versus regular speech therapy in global
aphasia: A controlled study. Aphasiology, 12, 385—
394.

Doyle, P. J., Oleyar, K. S., & Goldstein, H. (1991).
Facilitating functional conversational skills in

aphasia: An experimental analysis of a generalization

training procedure. Clinical Aphasiology, 229-241.

Ellis, A., Franklin, S. & Crerar, A. (1994). Cognitive
neuropsychology and the remediation of disorders of
spoken language. . In Riddoch, M. J., & Humphreys, G.
W. (Eds.). ni i ro h l i 've

Rehabilitation. Hillsdale: Lawrence Erlbaum.

Elman, R. J. & Bernstein—Ellis, E. (1999). The efficacy of

group communication treatment in adults with chronic

aphasia. Journal of Speech, Language, and Hearing

Researoh, _g, 411-419.

Ferraro, F. R., Bercier, B., & Chelminski, I. (1997).
Geriatric Depression Scale-Short Form (GDS-SF)

performance in Native American elderly adults.

Clinical Cerontologist, 18, 52—55.

192

Ferraro, F. R., Blaine, T., Flaig, S., & Bradford, S.

(1998). Familiarity norms for the Boston Naming Test
stimuli. Applied Neuropsychology, 5, 43-47.

Flanagan, J. L, Jackson, S. T. (1997). Test-retest
reliability of three aphasia tests: Performance of
non-brain-damaged older adults. Journal of

Communioation Disorders, JC, 33-43.

Frattali, C. (1993). Perspectives on functional

assessment: Its use for policymaking. Disability and

R ili i n, ii, 1-9.

Frattali, C. M., Thompson, C. M., Holland, A. L., Wohl, C.
B. & Ferketic, M. (1995). The FACS of Life: ASHA
FACS - A functional outcome measure for adults. ASHA,

.311 40-46.

Gallagher, T. M., & Guilford, A. M. (1977). Wh— questions:

Responses by aphasic patients. Cortex, l3, 44-54.

Girelli, L., Delazer, M., Semenza, C., & Denes, G. (1996).
The representation of arithmetical facts: Evidence

from two rehabilitation studies. Cortex, 32, 49-66.

Goodglass, H. (1990). Commentary: Cognitive psychology

and clinical aphasiology. Aphasiology, A. 93—96.

193

Goodglass, H., Hyde, M. R., & Blumstein, S. (1969).
Frequency, picturability and availability of nouns in

aphasia. Cortex, 5, 104-119.

Goodglass, H. & Kaplan, E. (1983). Boston Diagnostio

Aphasia Examination. Philadelphia: Lea & Febiger.

Goodglass, H. & Menn, L. (1985). Is agrammatism a unitary
phenomenon? In Kean, M. L., (Ed.), Agrammatism. New

York: Academic Press.

Grasby, P. M., Frith, C. D., Firston, K. J., Bench, C.,
Frackowiak, R. S. J., & Dolan, R. J. (1993).
Functional mapping of brain areas implicated in

auditory verbal memory function. Brain, 115, 1-20.

 

Grant, D. A., & Berg, E. A. (1993). i n in r or in

Test. Los Angeles, CA: Western Psychological

Services.
Grossman, M. & Carey, S. (1987). Selective word learning
deficits in aphasia. Braiu and Language, 5;, 306-324.
Habib, M., & Demonet, J. (1996). Cognitive neuroanatomy of

language: The contribution of neuroimaging.

Aphasiology, 10, 217-234.

194

Hammill, D. D., Pearson, N. A., & Voress, J. K. (1993).
D e ment Te 0 Visual Pe e ion. 2nd edition.

Austin, TX: Pro-Ed.

Hartman, J. & Landau, W. M. (1987). Comparison of formal
language therapy with supportive counselling for

aphasia due to acute vascular accident. Arohiyes of

Neurology, 24, 646-649.

 

Helm-Estabrooks, N. (1981). Helm Elioited Language Ptogxam
f r tax imul i n. Austin, TX: Pro-Ed.

Helm-Estabrooks, N., & Ramsberger, G. (1986). Treatment of

agrammatism in long-term Broca’s aphasia. British

Joutual of Disotdets of Communioation, 21, 39-45.

 

Hillis, A. E. (1989). Efficacy and generalization of

treatment for aphasic naming errors. Arohiyes of

Physioal Medioine and Rehabilitation, 10, 632-636.

Hillis, A. E. (1994). Contributions from cognitive

analyses. In Chapey, R. (Ed.), Language Intetyention

ie f A l with A hasia. Baltimore:

Williams & Wilkins.

195

Holland, A. L. (1991). Pragmatic aspects of intervention

in aphasia. Journal of Neurolinguistic , 5, 197-211.

Holland, A. L. (1975). The effectiveness of treatment in

aphasia. In R. H. Brookshire (Ed.), Clinioal

Aphasiology Conference Proceedings. Minneapolis, MN:

BRK Publishers.

Holland, A. (1980). Communicative Abilities in Daily

Living. Baltimore: University Park Press.
Holland, A. (1982). Observing functional communication of
aphasic adults. Journal of Speeoh and Hearing

Disorders, 47, 50-56.

 

Holland, A., Frattali, C., & Fromm, D. (1999).
Communioation Aotiyities of Daily Liying. Second

Edition. Austin, TX: Pro-Ed.

Holland, A. L., Swindell, C. S., & Fromm, D. (1983). A
model treatment approach for the acutely aphasic
patient. In R. H. Brookshire (Ed.), Clinioal

Ap asiology. Minneapolis, MN: BRK Publishers.

Hollingshead, A. B. (1975). Four factor index of social

status. Working paper.

196

Howard, D. (1986). Beyond randomized controlled trials:
The case for effective case studies of the effects of

treatment in aphasia. British Journal of Disorders of

Communication, _l, 89-102.

Howard, D., Patterson, K., Franklin, S., Orchard—Lisle, V.,
& Morton, J. (1985). Treatment of word retrieval

deficits in aphasia. Brain, 108, 817-829.

Jordan, L. (1998). ‘Diversity in Aphasiology’: A social

science perspective. Aphasiology, 12, 474-480.

 

Kaplan, E., Goodglass, H., & Weintraub, S. (1983). Boston

Naming Test. Philadelphia: Lea & Febiger.

Kapur, S., Craik, F. I. M., Tulving, E., Wilson, A. A.,
Houle, S., & Brown, G. N. (1994). Neuroanatomical
correlates of encoding in episodic memory: Levels of

processing. r in f h ati n l e of

Scienoe, 9i, 2008—2011.

Katz, R. C. & Wertz, R. T. (1997). The efficacy of

computer-provided reading treatment for chronic aphasic

adults. Journal of Speech, Language, and Hearing

Research, 40, 493-507.

197

Kay, J., Lesser, R., & Coltheart, M. (1992).
i i ic A men of Lan a e ro in in

Aphasia. Psychology Press.

Kearns, K. (1997). Broca’s aphasia. In LaPointe, L.
(Ed.), Aphasia aud Related Neutogenio Language
Disorders. "d Edition. New York: Thieme.

 

Kertesz, A. (1982). Western Aphasia Battery. New York:

Grune & Stratton.

Kertesz, A. (1988). Cognitive function in severe aphasia.

In Weiskrantz, L. (Ed.). Thought Without Language.

Oxford: Clarendon Press.

Kertesz, A. (1990). What should be the core of aphasia

tests? (The authors promise but fail to deliver).

Aphasiology, 4, 97-102.

Knopman, D. S., Selnes, O. A., Niccum, N., & Rubens, A. B.
(1984). Recovery of naming in aphasia: Relationship
to fluency, comprehension and CT findings. Neurology,

14, 1461-1470.

Laiacona, M., Allamano, N., & Capitani, E. (1996).
Performance consistency in picture naming: A study of

the rehabilitation effect on two aphasic patients.

198

Journal of Clinical and Experimental Neuxopsyohology,
1L8! 923-933.

Lamport-Hughes, N. (1995). Learning potential and other
predictors of cognitive rehabilitation. Ibe Jouxpal of
Coguitiye Rehabilitation, July/August, 16-21.

Le Dorze, G., & Pitts, C. (1995). A case study evaluation

of the effects of different techniques for the

treatment of anomia. Neuropsyohologioal

Le Dorze, G., Boulay, N, Gaudreau, J., & Brassard, C.
(1994). The contrasting effects of a semantic versus a
formal-semantic technique for the facilitation of

naming in a case of anomia. Aphasiology, 8, 127-141.

Lesser, R., & Algar, L. (1995). Towards combining the

cognitive neuropsychological and the pragmatic in

aphasia therapy. Neuropsyohologioal Rehabilitation, 5,
67-92.
Li, E. C., Williams, S. E., & Volpe, A. D. (1995). The

effects of topic and listener familiarity on discourse
variables in procedural and narrative discourse tasks.

Journal of Communication Disoxders, 28, 39—55.

 

199

Lomas, J., Pickard, L., Bester, S., Elbard, H., Finlayson,
A., & Zoghaib, C. (1989). The Communicative
Effectiveness Index: Development and psychometric
evaluation of a functional communication measure for

adults. Journal of Speech and Hearing Disorders, 54,
113-124.

Lomas, J., Pickard, L., & Mohide, A. (1987). Patient
versus clinician item generation for quality-of-life

measures: The case of language-disabled adults.

Medical Care, 25, 764-768.

Lowell, S., Beeson, P. M. & Holland, A. (1995). The

efficacy of a semantic cueing procedure on naming

performance of adults with aphasia. e ' n urn 1
of Speeob-Language Pathology, 4, 109-114.

Marr, D. (1982). i i n- A cm i n l 'n ' ' n
.10 ‘ 11.11:. f9-‘ n -.o a o .0 0 ‘ in o i a1

information. San Francisco: W.H. Freeman.

Methe, S., Huber, W. & Paradis, M. (1993). Inventory and
classification of rehabilitation methods. In Paradis,
M. (Ed.), F n i n f A ia R ha il't ion. New

York: Pergamon Press.

200

Milner, B. (1963). Effects of different brain lesions on

card sorting. Archives of Neurology, 2, 90-100.

Mills, R. H., Knox, A. W., Juola, J. F. & Salmon, S. J.
(1979). Cognitive loci of impairments in picture

naming by aphasic subjects. ou n l of c an

Hearing Research, 22, 73—87.

Mitchum , C. C. & Berndt, R. S. (1995). The cognitive

neuropsychological approach to treatment of language

disorders. Neuropsychological Rehabilitation, 5, 1-16.

Murray, L. L., & Holland, A. L. (1995). The language
recovery of acutely aphasic patients receiving

different therapy regimens. Aphasiology, 9, 387-405.

Murray, L. L., Holland, A. L., & Beeson, P. M (1997).
Accuracy monitoring and task demand evaluation in
aphasia. Aphasiology, ll, 401-414.

Murray, L. L., Holland, A. L., & Beeson, P. M. (1998).
Spoken language of individuals with mild fluent aphasia
under focused and divided-attention conditions.

urn e ch Lan n a in R rc , 4_,

213—227.

201

Nicholas, L. E. & Brookshire, R. H. (1993). A system for
quantifying the informativeness and efficiency of the

connected speech of adults with aphasia. Journal of

 

Speeoh and Hearing Research, 35, 338-350.
Nickels, L., & Best, W. (1996). Therapy for naming

disorders (Part I): Principles, puzzles and progress.

Aphasiology, lfl, 21-47.

Ostergaard, A. L., & Meudell, P. R. (1984). Immediate
memory span, recognition memory for subspan series of
words, and serial position effects in recognition
memory for supraspan series of verbal and nonverbal
items in Broca's and Wernicke's aphasia. Brain and

Language, 22, 1-13.

 

Parmalee, P. A. & Katz, I. R. (1990). The Geriatric

Depression Scale. Journal of the American Geriatrics
0 'et , 38, 1379.

Petharam, B. & Parr, S. (1998). Diversity in aphasiology:
Crisis or increasing competence? Apbasiology, l2, 435-
487.

Porch, B. (1967). r n ic i Abil't .

Palo Alto, CA: Consulting Psychologists Press.

202

Pring, T. R. (1986). Evaluating the effect of speech
therapy for aphasics: Developing the single case
methodology. British Journal of Disordexs of
Communication, 21, 103-115.

 

Prins, R. S., Schoonen, R., & Vermeulen, J. (1989).
Efficacy of two different types of speech therapy for

aphasic patients. Applied Psyoholinguistios, lg, 85-
123.

Raven, J. C., Court, J. H., & Raven, J. (1979). Manual for
.2 -1’- Poo -_ 5 - . ,'ce and . to 5-0 _,—_-_.

London: Lewis & Co.

Raymer, A. M., Thompson, C. K., Jacobs, B., & Le Grand, H.
R. (1993). Phonological treatment of naming deficits

in aphasia: Model-based generalization analysis.

Aphasiology. 1. 27—53.

Risse, G. L., Rubens, A. B., & Jordan, L. S. (1984).
Disturbances of long term memory in aphasic patients.

Bxain, 1Q7, 605-617.

Robey, R. R. (1994). The efficacy of treatment for aphasic
persons: A meta-analysis. Brain and Language, 41,
585-608.

203

 

Robey, R. R. (1998) . A meta-analysis of clinical outcomes

in the treatment of aphasia. Journal of Speechl

 

Language. and Hearing Research, 41, 172—187.
Ross, K. B. & Wertz, R. T. (1999). Comparison of

impairment and disability measures for assessing
severity of, and improvement in, aphasia. Aphasiology,

l3, 113—124 .

Sarno, M. T. (1991) . Management of aphasia. In Bornstein,

R. A. 8. Brown, G. G. (Eds.), e r h i ral A e f

Cerebrovasoular Disease. New York: Oxford University

Press.

Sarno, M. T. & Buonaguro, A. (1983). Factors associated

with independent meal preparation in aphasic females:

A pilot study. The Qooupational Therapy Journal of

Research. 3, 23-34.

Sarno, M. T., Silverman, M., & Sands, E. S. (1970). Speech

therapy and language recovery in severe aphasia.

Journal of Speeob and Hearing Research, l3, 607-623.

Schuell, H., Jenkins, J. J. & Jimenez-Pabon, E. (1964).

Aphasia in Adults. New York: Harper & Row.

204

 

Shewan, C. M. (1986).

The history and efficacy of aphasia
treatment .

In Chapey, R. (Ed.), L

nt i
Strategies in Adult Aphasia. Baltimore, MD: Williams
& Wilkins.

Shewan, C. M. 8: Bandur, D. L. (1994).

Language-oriented
treatment: A psycholinguistic approach to aphasia. In
Chapey, R. (Ed.), In r n i n r ie in
Adult Aphasia. 3rd Ed. Baltimore, MD: Williams &
Wilkins.

Shewan, C. M., & Canter, G. J. (1971). Effects of
vocabulary, syntax, and sentence length on auditory
comprehension in aphasic patients. Cortex, 1, 209-226.

Shewan, C. M., 8: Kertesz, A. (1984).

Effects of speech and
language treatment on recovery from aphasia.

Brain and
W: 23., 272-299 .
Springer, L.,

Willmes, K

., & Haag, E. (1993). Training in

the use of wh-questions and prepositions in dialogues:

A comparison of two different approaches in aphasia
therapy- Aphasiology,

7, 251-270.

Terrell, B. Y. & Ripich, D. N. (1989). Discourse

competence as a variable in intervention.

eminars in
Speech and Language,

20, 282-297 .

205

Tikofsky, R. (1971) . Two studies of verbal learning by

adult aphasics. Cortex, 1, 106-125.

Ulatowska, H. K., Doyel, A. W., Freedman-Stern, R., Macaluso
Haynes, S. (1983). Production of procedural discourse

in aphasia. Brain and Language, 15, 315-341.

van Mourik, M., Verschaeve, M., Boon, P., Paquier, P., & van
Harskamp F. (1992) . Cognition in global aphasia:

Indicators for therapy. Aphasiology, 5, 491-499.

Weniger, D. & Sarno, M.T. (1990) . The future of aphasia

therapy: More than just new wine in old bottles?

Aphasiology, 4, 301-306.
Wertz, R. T. (1993) . Efficacy of various methods of

therapy. In Paradis, M. (Ed.) . Eoundations of Aphasia
Rehabilitation. New York: Pergamon Press.

Wertz, R. T. (1995). Efficacy. In Code, C. & Muller, D.

(Eds.). Tr mn fAh '- m h

Pragtice. _San Diego: Singular Publishing.

Wertz, R. T., Auther, L. L., & Ross, K. B. (1997). Aphasia

in African-Americans and Caucasians: Severity,

206

improvement, and rate of improvement. Aphasiology, ll,

533-542.

Wiegel-Crump, C. & Koenigsknecht, R. (1973). Tapping the
lexical store of the adult aphasic: Analysis of the
improvement made in word retrieval skills. Cortex,

411-418.

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, 0., Huang,
V., Adey, M., & Leirer, V. O. (1983). Development and
validation of a geriatric depression screening scale:

A preliminary report. Journal of Psyohiatrio Research,
17, 37-49.

_—

Yorkston, K. & Beukelman, D. (1980). An analysis of

connected speech samples of aphasic and normal

speakers. Journal of Speeoh and Hearing Disorders, 45,
27—36.

207