This article was downloaded by:[B-on consortium][B-on consortium]

On: 3 May 2007Access Details: [subscription number 758066635]Publisher: Informa HealthcareInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Augmentative and AlternativeCommunicationPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713692248

AAC aphasiology: partnership for future researchLynn E. Fox ab; Melanie Fried-Oken ca Communication Disorders and Sciences Program, University of Oregon, Eugene,Oregon, USA.b Speech Pathology Section, Portland Veterans Affairs Medical Center, Portland,Oregon, USA.c Department of Neurology, Oregon Health Science University, Oregon, USA.

To cite this Article: Lynn E. Fox and Melanie Fried-Oken , 'AAC aphasiology:partnership for future research', Augmentative and Alternative Communication, 12:4,257 - 271To link to this article: DOI: 10.1080/07434619612331277718

URL: http://dx.doi.org/10.1080/07434619612331277718

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction,re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expresslyforbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will becomplete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should beindependently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with orarising out of the use of this material.

© Taylor and Francis 2007

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

Treatment for severe aphasia has always been aproblematic issue for aphasiologists. Most of thehighly structured language and process stimulationapproaches used to treat aphasia are ultimatelyunsuccessful in providing language systems that meetthe daily needs of severely impaired individuals(Rosenbek, LaPointe, & Wertz, 1989; Shewan, 1986).Even authors of the earliest comprehensive treatmentapproaches suggested that severely aphasic individ-uals should be treated differently than those with mod-erate to mild impairments (Schuell, Jenkins, &Jimenez-Pabon, 1964; Wepman, 1972). According toSchuell et al. (1964), “the best results we have seenwere all those where a new routine, adapted to theabilities, interests, and limitations of the patient, andto the needs and interests of the rest of the family aswell, was thoughtfully planned and systematicallyadhered to” (p. 379).

Over the past 30 years, clinicians and researchersin aphasiology have sought to explain severe aphasiaand to develop effective treatment paradigms for thedifferent classes of aphasia. Early treatments oftenattempted to stimulate the damaged language func-tions of the brain (Schuell, 1974; Schuell et al. 1964).Later, approaches sought to use intact symbolic andprosodic capacities as compensations for impairedlinguistic abilities (Helm-Estabrooks & Albert, 1991;Meyers, 1980; Rao & Horner, 1978; Skelly, Schinsky,Smith, & Fust, 1974; Sparks & Holland, 1976). Simul-

taneously, treatments have addressed the functionalcommunication needs of persons with aphasia, withan emphasis on development of compensatory toolsand skills (Aten, 1986; Aten, Caligiuri, & Holland,1982; Collins, 1986; Davis & Wilcox, 1985; Holland,1980, 1982; Morgan & Helm-Estabrooks, 1987). Hol-land (1980) has suggested that functional communi-cation treatment differs from other forms of aphasiatreatment. “Functional” is defined as “getting mes-sages across in a variety of ways ranging from fully-formed grammatical sentences to appropriate ges-tures, rather than being limited to the use ofgrammatically correct utterances” (p. 50). According tothis definition, training in the use of augmentative andalternative communication systems (AAC) may be acomponent of functional communication treatment.

In order to gain a better understanding of where thefield currently stands on issues related to AAC andaphasia, this paper will review three areas of theoryand practice: (1) functional communication approachesfor severe aphasia, (2) standard assessment practicesfor AAC interventions, and (3) recent research in thearea of AAC treatment outcomes with severely apha-sic adults. This body of research will be critically exam-ined to evaluate the methodologies used, the compre-hensiveness of AAC assessment practices, and thevalidity of the conclusions reached by the researchers.Ultimately, the goal of this review will be to draw on the-ory and practice in the field of AAC to suggest methods

0743-4618/96/1204-0257 $3.00/0; Volume 12, December 1996AAC Augmentative and Alternative CommunicationCopyright © 1996 by ISAAC

FORUM

AAC Aphasiology: Partnership for Future Research

Lynn E. Fox and Melanie Fried-OkenCommunication Disorders and Sciences Program, University of Oregon, Eugene, Oregon, USA and Speech Pathology Section, PortlandVeterans Affairs Medical Center, Portland, Oregon, USA (L.E.F.) and Department of Neurology, Oregon Health Science University, Oregon,USA (M.F.)

This paper explores issues related to augmentative and alternative communication (AAC) inter-ventions for aphasia. Recent literature on functional communication treatment is reviewed with agoal of clarifying how aphasiologists see AAC fitting into aphasia treatment. AAC assessment mod-els are proposed as a method for achieving the objectives of functional communication treatment.Recent AAC and aphasia treatment efficacy studies are critically examined for their use of acceptedAAC assessment models. Recommendations address the need to incorporate AAC assessmentwhen addressing questions of effectiveness, efficiency, and generalization in AAC aphasiologyresearch.

KEY WORDS: augmentative and alternative communication (AAC), aphasia, assessment, inter-vention, research

257

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

for improving the social validity of research investigat-ing AAC and aphasia interventions.

FUNCTIONAL APPROACHES TO APHASIA TREATMENT

A number of aphasiologists have explored issuesrelated to functional treatment and have addressedthe need to create working communication systemsfor individuals with severe aphasia. Davis and Wilcox(1985) introduced one of the early treatment ap-proaches designed to improve pragmatic aspects ofcommunication in aphasia. Their approach to treat-ment, Promoting Aphasics’ Communicative Effective-ness (PACE), encouraged the use of all availablecommunication modalities for conveying new infor-mation in meaningful exchanges.

In PACE treatment, clients are given free choice ofcommunication channels or modalities, and cliniciansprovide natural feedback based on communicativeadequacy rather than on the accuracy of symbolsused to convey the message. Davis and Wilcoxincluded two case studies using PACE treatment forsevere aphasia in their discussion of treatment imple-mentation. One of these provided an early example ofpartner training. This case study illustrated the use ofPACE treatment in training the communication partnerof an adult with severe Wernicke’s aphasia to recog-nize and accept the aphasic person’s use of multiplecommunication modalities.

Collins (1986) took a similar total communicationapproach in his discussion of treatment for globalaphasia. He recommended that clinicians treatingindividuals with global aphasia emphasize a variety ofindirect and direct approaches, using multiple com-munication modalities, to improve functional commu-nication. He encouraged clinicians to focus onenhancing residual skills and persuading family mem-bers that communication is more important than “lin-guistic elegance” (p. 99).

Emphasis on functional communication methodsand reliance on context to improve language behav-ior was further explored by Aten (1986, 1994). Unlikeothers who suggested that “functional” approachesshould be implemented after the conclusion of tradi-tional language treatment (e.g., Wertz, 1984), Aten(1986) proposed that functional communication treat-ment (FCT) be employed concurrently with languagestimulation and continue when language stimulation isno longer effective.

Aten’s FCT emphasizes incorporation of daily con-tacts and activities into treatment for informationexchange. He favors treatment that takes place in avariety of natural settings when possible, and thatinvolves interaction between individuals and withingroups. He suggests that treatment should be basedon the results of a variety of assessment measuresincluding formal language tests, functional communi-cation measures, discourse analysis, and observa-

tion of performance in natural settings. Beyond thesebasics of FCT, Aten (1994) considers communicationneeds of the aphasic adult, as well as the severity ofdeficit, when planning treatment.

The importance of a functional or “ecological”approach to aphasia treatment was illustrated by Sim-mons (1989), who described the use of a simulatedenvironment to provide a transitional setting for fos-tering carry-over of clinically learned communicationskills to more naturalistic settings. In a similar vein,Lyon (1989) described a treatment approach using vol-unteers who became communication partners for apha-sic adults within and outside the clinic. His anecdotalevidence pointed to increased communicative confi-dence and improved communicative performance forthe aphasic clients being treated in this manner.

More recently, Lyon (1992) has asserted that thereis little evidence linking restored communication abil-ity in structured or protected settings to improved useof communication and participation in daily life. Hesuggests that aphasic adults’ inability to communi-cate in natural settings may be as related to handicap(the psychosocial dysfunction of aphasia) as it is todisability (impaired use of communication modes).Lyon proposes treating both the disability and thehandicap of aphasia. When aphasic adults activelybecome agents of change in their own lives, Lyonbelieves that the psychosocial burden is reduced forthem and their families.

His approach to treatment involves use of commu-nity volunteers as liaisons for participation and re-entry into an active life in the community. Volunteersand aphasic clients participate in two phases of treat-ment. During the first phase, a volunteer partnerassumes the role that the clinician would typically takein PACE-like activities. In phase two, the aphasic per-son begins to plan and execute activities in the com-munity with his or her volunteer companion.

Another treatment approach, which focuses onaphasic clients becoming agents of their own change,is recreation-focused group treatment (Fox, 1990).Severely aphasic nonspeaking adults are encouragedto create their own communication opportunitiesthrough the selection of activities in a group milieu.Communication needs are assessed for each activity,and group members are prepared for communicativeindependence in the natural environment. Pre- andpostmeasures over a 6-month period of group treat-ment showed no change in language ability asreflected in the Porch Index of Communicative Ability(PICA) (Porch, 1967), but functional communicationimproved as measured by the Communicative Effec-tiveness Index (CETI) (Lomas, Pickard, Bester,Elbard, Finlayson, & Zoghaib, 1989) and by a prag-matic assessment of group interaction (Fitzpatrick,Nicholas, & DiNapoli, 1988).

Currently, the work of Kagan and Gailey (1993) isdrawing much attention among those interested inpromoting improved functional communication for

258 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

aphasic adults. Their work with the Aphasia Centre-North York focuses primarily on providing opportuni-ties for aphasic adults to participate in conversationswith other aphasic adults and with trained volunteersin group settings. The volunteer-run groups at thecentre provide a number of different models for learn-ing and using communicative skills in a conversationalsetting. A core of 100 to 125 aphasic adults participatein structured groups run by volunteers. The goal ofthese groups is to practice techniques that facilitateconversation. Volunteers come to these sessions withplanned activities in mind. Professional staff of thecentre provide a packet of materials and props thatwill facilitate communication for each group (informa-tion about members, photos of staff, maps, clocks,etc.). Less formal opportunities are offered at the cen-tre for the participants to practice their conversationalskills, including coffee-time chats and volunteer-runrecreational activities, lectures, and outings.

The volunteer-run treatment at the North York Cen-tre remains fairly traditional, with volunteers selectingtopics and leading activities. Unlike treatments devel-oped by Lyon and Fox, the North York approach doesnot address client selection of topics and activities.Communicative autonomy appears to be offered onlyduring participation in less formal groups. Additionally,Kagan and Gailey do not address the need for partici-pants to assume responsibility for having necessarycommunication materials available, and spouses orfriends are not currently being trained to act as con-versational partners in the community. In the NorthYork program, spouses are encouraged to observe thevarious group activities and may implement tools andstrategies that they find useful outside of the centre.

Clearly, many aphasiologists are beginning toaddress the need for a “nontraditional” approach tothe treatment of severe aphasia, and much work hasbeen done describing the types of treatments that arelikely to be of greatest benefit to severely impairedindividuals. Treatment has moved beyond the deficitreduction approaches of early traditional therapies tonewer treatments that address the communicationneeds and lifestyles of aphasic individuals. Such treat-ments require comprehensive assessment of individ-ual, family, and environmental factors that impactcommunicative effectiveness (Aten, 1994; Lubinski,1994; Parr, 1996). Aten (1994) states that “theessence of functional appraisal is to let the patientguide you. The hierarchy of losses and needs andresidual strengths that are revealed become the data-base for intervention” (p. 296). Unfortunately, aphasi-ologists have not yet defined assessment models thatwill guide functional treatments.

AAC ASSESSMENT MODELS

In addition to standardized language tests and thelimited functional communication assessment toolsavailable to aphasiologists, the field of AAC offers

models for the assessment of communication needs,capabilities, and environmental factors that must beconsidered in planning functional communicationtreatment (Beukelman & Garrett, 1988; Beukelman &Mirenda, 1988, 1992; Beukelman, Yorkston, & Dow-den, 1985; Dowden, Beukelman, & Lossing, 1986;Hux, Beukelman, & Garrett, 1994; Yorkston & Karlan,1986). Over the past decade, three models of AACassessment have evolved: (1) candidacy or processmodels, (2) the communication needs model, and (3)the participation model.

These models address issues such as candidacyfor AAC use, system selection based on an individ-ual’s communication needs and capabilities, andassessment of environmental barriers and opportuni-ties. The common thread of all of the models currentlyin use is the concept that assessment must be anongoing process integrated with treatment. Criticaldecision-making points act as triggers for reassess-ment as needs change, when new unmet needsemerge, or if change is likely to occur in the nearfuture (Yorkston & Karlan, 1986). This idea of inte-grating assessment with treatment is not new to thefield of aphasiology. What is new, and potentially veryuseful to aphasiologists, is the concept that commu-nication needs and patterns of interaction must beconstantly monitored. Changes in any of these areasshould trigger reassessment.

Candidacy and Process Models

Beukelman and Mirenda (1992) described earlypractitioners in the field of AAC who worked primarilywith models of assessment that attempted to deter-mine if individuals were candidates for AAC interven-tion. These have been grouped as candidacy models.They seek to ascertain whether AAC intervention isrequired and whether individuals considered for inter-vention meet the criteria of strong cognitive and lin-guistic abilities in the presence of chronic expressivecommunication disorders. Individuals who do notmeet these criteria are required to perform “perpetualreadiness” activities that are designed to teach skillsassumed to be prerequisites for communication.

Candidacy models have lost favor in recent years forall populations of communicatively impaired individu-als, including those with the most severe intellectualdisabilities (Mirenda, 1993). Mirenda says that “breath-ing is the only prerequisite that is relevant to commu-nication” (p. 4). She asserts that AAC techniques anddevices are available now that are simple to learn andoffer flexible symbol options for those who categorizethe world differently than the majority culture. It hasbeen suggested recently that there are only two AACcandidacy criteria for adults with aphasia: (a) “thosepersons who do not regain sufficient natural speech forcommunication of basic needs, and (b) those personswho rely on natural speech to meet many of their com-munication needs but find it inadequate or inefficient in

AAC Aphasiology 259

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

certain instances” (Hux et al., 1994, p. 341). This perspec-tive suggests that there must be very few severelyaphasic adults for whom some level of AAC assess-ment and intervention is not appropriate.

In contrast to answering questions only related tocandidacy, the process model has three objectives:(1) to determine whether a discrepancy existsbetween skills possessed by normal communicatorsand the individual being assessed, (2) to explore pat-terns of communication being used, and (3) to deter-mine the benefit of remediation for a dysfunctionalspeech system (Beukelman & Garrett, 1988). Theseobjectives remain viable for assessing communica-tion disabilities in persons with severe aphasia. How-ever, they provide only a fragment of the informationcurrently considered necessary for a complete AACassessment.

The standardized norm-referenced tests most fre-quently used to assess aphasic adults, the BostonDiagnostic Aphasia Examination (BDAE) (Goodglass& Kaplan, 1983), Western Aphasia Battery (WAB)(Kertesz, 1982), and the PICA (Porch, 1967), areexamples of instruments that address skill discrepan-cies between the individual being assessed and nor-mal communicators. Although the process modelseeks to determine the benefit of remediation and toidentify patterns of communication being used, sub-sequent models more comprehensively address theassessment of communication needs and preservedcommunication capabilities.

Communication Needs Model

The communication needs model documents unmetcommunication needs, identifies communicationmethods that are effectively meeting some needs, andprovides guidance for planning AAC interventionusing residual communication abilities (Beukelman &Mirenda, 1992). It represents an important stepbeyond the process model. Rather than emphasizingthe identification of communication deficits, thisapproach attempts to detect unique communicativeskills in the individual, and to enhance those skills inareas where communication needs are unmet(Beukelman et al., 1985; Dowden et al., 1986; York-ston & Karlan, 1986). All methods available for col-lecting information about residual skills are consid-ered. These may include use of norm-referenced orcriterion-referenced standard measures, nonstandardmeasures, and observation of functional communica-tion in natural settings. Branching decisions that relyupon assessment of maximum communicative perfor-mance, in areas required for specific AAC interven-tions, may be considered in lieu of comprehensiveassessment in selected cases. Two instruments cur-rently within the aphasiologist’s inventory of assess-ment tools may address communication capabilities forthis model. The Boston Assessment of Severe Apha-sia (BASA) (Helm-Estabrooks, Ramsberger, Morgan, &

Nicholas, 1989) offers important contextual informa-tion about an aphasic adult’s residual communicativeabilities. The Communicative Abilities in Daily Living(CADL) (Holland, 1980) may also be useful for deter-mining how an aphasic individual employs alternativemodalities when communicating in natural settings.

In addition to these aphasia tests, a guide forassessment of communication needs is included inCommunication Augmentation: A Casebook of Clini-cal Management (Beukelman et al., 1985). Anotheruseful tool that was designed to measure functionalcommunication outcomes, but that may also provideclues to the communication needs of aphasic adults,is the CETI (Lomas et al., 1989).

Beyond the use of those tools mentioned above, thecommunication needs model guides observation andassessment of communication needs in the naturalenvironment. Hux et al. (1994) suggest that a firststep to such an assessment entails consideration ofthe typical communication patterns of an individual’snondisabled peers. For example, retelling stories,engaging in conversations that emphasize friendsrather than family members, and increased use ofgames as a focus for interaction all represent uniquecommunication patterns of elderly adults (Stuart,1991). These communication patterns may serve asa basis for the communication needs assessment ofelderly aphasic individuals.

Patterns of communication will differ among indi-viduals depending on the age group to which theybelong, the restrictiveness of their living setting, andtheir level of language impairment. Regardless ofthese differences, Light (1988) has identified foursocial purposes of communication that are shared byall disabled and nondisabled persons. These four pur-poses include (1) communication of wants and needs,(2) communication for transfer of information, (3) com-munication for social closeness, and (4) communica-tion for social etiquette. Although the relative impor-tance of these communication purposes will vary, allindividuals will communicate to achieve each ofLight’s four purposes. A perceptive clinician willexplore each of these areas when completing a com-munication needs assessment in the natural environ-ment and will also attempt to determine which com-munication needs are being met through the use ofnatural methods, such as gesture or automaticphrases. Hux et al. (1994) caution that messagesalready being conveyed with natural communicationmethods are not likely to be accessed by an aphasicperson if they are included in an AAC system.

Participation Model

In addition to communication needs, the latestassessment model considers an individual’s partici-pation patterns when implementing AAC interven-tions. The participation model takes into considerationaccess and opportunity barriers that prevent treat-

260 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

ment candidates from communicating in a mannersimilar to their peers (Beukelman & Mirenda, 1992).The underlying causes of these barriers are identifiedand interventions are directed at managing deficien-cies in opportunity, natural ability, and environmentaladaptation for AAC system introduction. Simultane-ously, environments most conducive to communica-tion are exploited. This model incorporates the con-cept of ongoing assessment. As interventions createadditional communication opportunities, reassess-ment takes place, and interventions are created ormodified accordingly (Beukelman & Mirenda, 1988,1992).

Barriers to opportunity may include limitations inavailable professional support, limitations in financialsupport, and the level of family and friends’ commu-nication skills and support (Hux et al., 1994). The par-ticipation model recognizes that the speech-languagepathologist’s knowledge regarding AAC options andtheir implementation will affect the outcome of anintervention. The choice of AAC systems may also belimited by the financial resources of an aphasic indi-vidual and the willingness of a third-party payer topurchase necessary equipment. Family members’ orfriends’ knowledge and willingness to provide appro-priate assistance will also influence the use and use-fulness of AAC systems. Guidelines are available forassessing partner attitudes toward the aphasic adult’scommunication loss, and checklists have been devel-oped for identifying interactional skills important foraphasic adults’ successful communication (Garrett &Beukelman, 1992; Hux et al., 1994).

Natural ability is examined in the participationmodel, with a difference in emphasis from the earlierprocess model. Natural ability is considered for itspotential to create barriers and influence opportunitiesfor communication. The participation model examinescontextual awareness, language comprehension,recognition of communicative desire, message recog-nition, and the formulation and revision necessary foruse of AAC systems. Each of these abilities isassessed during communication interactions betweenpersons with aphasia and their communication part-ners. AAC support is implemented at each level wherebreakdowns occur. For example, language compre-hension deficits may be augmented through writing,drawing, or picture/symbol choices presented by theconversational partner (Hux et al., 1994).

Beyond the Participation Model

The next step in the application of AAC assessmentand treatment models with aphasic adults might becalled the communicative independence model.Although the participation model encourages the cre-ation of opportunities for aphasic clients, it does notaddress the adult’s need to learn in the context of hisown experience and interests. The participation modelhelps identify opportunity and access barriers and to

view AAC as a means for overcoming these environ-mental constraints (Mirenda, 1993). However, it doesnot identify motivating factors that are likely to facili-tate learning and to promote use of AAC systems innonclinical environments. The next step will requirespeech-language pathologists to employ treatmentmethods that provide incentives for learning. It may benecessary for speech-language pathologists torelease some control of clinician-directed assessmentand treatment tasks to which they have becomeaccustomed. If this is the case, and if clients are ableto become active agents in their own communicationrecovery, they may demonstrate improved motivationto learn and may more efficiently generalize use ofnewly learned skills to other environments (Fox, 1990;Fox & Fried-Oken, 1996; Lyon, 1992).

Developing methods for assessing preferences andlevels of motivation in severely aphasic adults willrequire creativity and ingenuity from speech-languagepathologists. It will require development of nonverbalmethods for determining preferred communication topicareas, for assessing motivation for AAC system use,and for assuring investment of aphasic individuals andtheir conversational partners in treatment outcomes.

Tools and techniques that augment an aphasic per-son’s auditory comprehension and verbal expressionwill need to be developed in order to determine com-munication preferences. A method for obtaining inputregarding preferred communication topics might becreated by examining communication patterns ofnondisabled peers (Stuart, 1991). Pictorial systemsmay be created so that aphasic adults can select top-ics from those typically of interest to their peers.Selection of topics might be as simple as choosingfrom an array, or may require more structured assis-tance from the clinician. Paired comparisons of pic-tured topics may be useful, or it may be possible torank-order preferred topics using methods such asthe Q-sort (Stephenson, 1953). Written choice com-munication would be used to clarify specific issuesand interests within a topic area.

Motivation to learn and willingness to accept non-verbal communication methods will need to be con-sidered as interventions are begun, before systemsare selected for aphasic adults. Tools, such as thosedeveloped to predict a child’s receptivity to differentcommunication options, will need to be modified foruse with adults (Culp & Carlisle, 1988). In contrast tothe participation model, which addresses barriers thatmay interfere with use of AAC systems, the commu-nicative independence model will address internal andexternal factors that motivate aphasic individuals tolearn and to use communication systems in naturalenvironments.

As new treatment approaches for severe aphasiacontinue to be introduced, it will be critical for investi-gators to use comprehensive assessment methodsbefore and during treatment. In order to empiricallyevaluate the benefit of AAC interventions, it will be

AAC Aphasiology 261

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

necessary for investigators to prescribe interventionsthat are appropriate for subjects in terms of commu-nication needs and capabilities, address social andenvironmental barriers to communication, and con-sider each individual’s interests and preferences.Assessment of each of these areas will need to beaddressed repeatedly over time so that as needs, abil-ities, and requirements change, treatment plans willbe adjusted (Fried-Oken, 1992). Without appropriateassessment, treatment findings and generalizationimplications will be meaningless and misleading.

AAC AND APHASIA RESEARCH

Studies investigating outcomes of AAC interven-tions with aphasic persons must incorporate acceptedAAC assessment practices in their design to bemethodologically sound. It should be clear from thediscussion above that, without comprehensive AACassessment, treatment is unlikely to be successful,and generalization of newly learned skills to otherenvironments cannot be expected. Recent outcomestudies in AAC and aphasia will be critically examinedto evaluate the methodologies used, the comprehen-siveness of AAC assessment practices, and the valid-ity of the conclusions reached by the researchers.

Outcome studies in AAC and aphasia fall into fourbroad classifications: comprehensive case studies,carefully controlled single-case experimental studies,group studies, and descriptive or comparative studiesthat do not fit into the other classifications. An exam-ination of some of these recent studies, with anemphasis on methodology and attention to assess-ment methods used, may suggest appropriate direc-tions for future research.

Case Studies

Case studies presented in both the aphasia andAAC literature illustrate the complexity of AAC inter-vention for aphasia. Cases illustrated in the literatureinvolve lengthy interventions using multiple commu-nication systems. Often, the use of these systemschanges over time as the needs and the abilities ofaphasic subjects change.

One such case study examined the use of Blissym-bols as a tool for communication with a severely apha-sic adult (Bailey, 1983). In this study, communicationneeds, capabilities, and social support were consid-ered before the intervention began. Initial process andcapability assessment included administration of theMinnesota Test for the Differential Diagnosis of Apha-sia (MTDDA) (Schuell, 1965), Raven’s Coloured Pro-gressive Matrices (RCPM) (Raven, 1938), and Koh’sBlock Design Test (Yates, 1954). The subject’s highscore on the RCPM (100% correct) was considered apositive indicator for potential success with a visuallybased system such as Blissymbols. Vocabulary wasinitially based on what the author describes as “the,

then standard, 100-word chart following the Blissym-bolics Communication Institute programme guidelines”(p. 180). Subsequently, this vocabulary was adaptedto address the subject’s communication needs.

Bailey reported the subject’s progress descriptively,through scores obtained in repeated administrationsof the MTDDA, RCPM, and Koh’s Block Design Test,and through subjective reports of system use in andoutside of the training environment. Bailey found that,after 18 to 24 months of treatment, the subject wascombining symbols to create new concepts and wasusing the symbols spontaneously at home and duringtreatment. At 24 to 30 months, the subject was usinga 200-symbol chart and was communicating usingcombinations of words and Blissymbols. Overall, therewere improvements on all MTDDA subtests, but per-formance on the RCPM and Block Design Test declinedslightly.

It is ironic that, after 33 months of apparently suc-cessful use of this system, the author concluded thatBlissymbols were not an ideal alternative communi-cation method for the subject but were an effectivetreatment strategy. Bailey based this conclusion onthe fact that, at the end of the intervention, the sub-ject’s MTDDA scores had improved; however, he hadrequested that Blissymbols be removed from his com-munication board. He wanted only words to beincluded on the board that he used for his sponta-neous communication. In fact, a more appropriateconclusion may be that Blissymbols provided a func-tional alternative communication system during 30months of intervention. Additionally, the Blissymboltraining resulted in improved reading and writing ofwords for communication, which ultimately becamethis subject’s communication method of choice.

An often-cited case study explored the use of man-ual sign by a subject with nonfluent aphasia (Coelho& Duffy, 1985). The subject of this study is describedas a retired physician who scored at the 45th per-centile overall on the PICA (Porch, 1967) and the 49thpercentile on the auditory comprehension subtests ofthe DAE (Goodglass & Kaplan, 1983). The subjectwas taught a variety of manual signs representingnouns, verbs, and adjectives, with training includingimitation, recognition, and production tasks. Theauthors report that, although the subject learned 73 of90 signs presented in training, he used a trained gestureonly once in monitored conversations and did not usethe trained signs in interactions in his home. Theauthors concluded that although aphasic subjects canacquire sign vocabularies, they do not use trainedsigns for functional communication. It must be pointedout that the authors did not conduct a needs assessmenteither in the conversational or home settings beforeselecting vocabulary. In fact, vocabulary is describedas being targeted to meet the subject’s basic needs.Although the authors suggest that all tasks weredesigned to elicit use of the acquired signs, it couldeasily be argued that the communicative purposes

262 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

involved in topical conversations and in daily interac-tions in the home environment do not frequentlyrequire communication of basic needs (Light, 1988).

Case studies found in the AAC literature often inte-grate multiple communication modalities into systemsthat function to meet specific communication needs ofaphasic adults (Beukelman et al., 1985; Garrett,Beukelman, & Low-Morrow, 1989). A case study pre-senting such a multimodality intervention demon-strated the ability of an aphasic adult to use and inte-grate a variety of communication methods at homeand work (Beukelman et al., 1985). Communicationneeds were addressed and continuously monitoredthroughout the intervention. Process and capabilityassessment included findings of the PICA (Porch,1967), subtests of the Wechsler Adult IntelligenceScale (WAIS) (Wechsler, 1955), and the Raven’s Pro-gressive Matrices Test (Raven, 1960). Results of thePICA, administered at 1 month post stroke, found thesubject with an overall score at the 22nd percentileand unable to perform any of the verbal tasks. Audi-tory comprehension scores were superior to verbalscores. The subject’s performance was within normallimits on cognitive measures, which were adminis-tered 2 }

12

} months post onset. Following the guidelinesof the participation model, the subject’s interventionevolved as the client’s needs and abilities changed.Multiple administrations of the PICA were used tomonitor language changes.

Initial intervention was designed to augment audi-tory comprehension deficits, meet the basic commu-nication needs of the rehabilitation environment, andsubsequently allow the subject to begin communicat-ing about his business. In accordance with Aten’s(1986) functional communication treatment, AACintervention was conducted simultaneously with a pro-gram emphasizing improvement in expressive andreceptive language skills. Initially, communicationbooks were used to augment and teach improvedauditory comprehension skills. At 9 months postonset, following improvements in auditory compre-hension, melodic intonation therapy was introduced(Sparks & Holland, 1976). As the subject returned towork, workplace needs were identified and communi-cation systems were changed to address new needs.The subject ultimately returned to his interior designbusiness. His home office was reorganized so thatcatalogues and design materials could be easilyaccessed and used as augmentative communicationtools, an artist was hired to assist with the communi-cation of design concepts, and a voice output com-munication aid was added to his repertoire for topicintroduction and conversational control.

Another successful AAC case study demonstratedthe ability of an aphasic adult to use a number ofcommunication modalities independently, switchingbetween a hierarchy of modalities ranked for effi-ciency of use (Garrett et al., 1989). Prior to startingthe intervention, a process and capability assess-

ment was completed with the WAB (Kertesz, 1982)and communicative competencies were assessed fol-lowing the participation model. This assessmentrevealed strengths in communicative intent, supple-mentary use of gestures, and inconsistent ability touse an alphabet card for first-letter spelling. The sub-ject used pointing, drawing, reading, and writing forcommunication. Nonstandard assessments of envi-ronmental awareness, pragmatics, reading, gesture,drawing, pointing, and first-letter spelling were con-ducted. A needs assessment indicated that the sub-ject moved around his community independently andinteracted with familiar and nonfamiliar people on adaily basis. Communication barriers were identified inthe multiple attempts required for familiar communi-cation partners to decode messages, and frequentbreakdowns that occurred with unfamiliar communi-cation partners.

Intervention in this case study involved the intro-duction of multiple components to create a functionalcommunication system, along with the use of naturalcommunication in the form of speech, writing, anddrawing. Components of the subject’s communicationsystem included the following: (1) a word dictionaryorganized by topics that could be accessed logicallyby the subject, (2) an alphabet card with first lettersused to cue himself in verbal production or to cue hislistener, (3) a new information pocket for storage ofconversational props, (4) a card containing clues toassist an unfamiliar partner in formulating questionsthat would assist the subject in word retrieval, (5)printed conversational control phrases, and (6) blankwriting paper. Once the system was established, bar-riers to effective communication were identified. Theyincluded the subject’s inability to choose the most effi-cient mode of communication, his persistence in usinginefficient communication methods, and his inability toresolve communication breakdowns. As a result, twiceweekly treatment was conducted over 7 to 8 monthsto teach switching from most to least efficient com-munication modes in conversational activities, and toteach use of natural communication or controlphrases for topic initiations and changes. Pre- andpostintervention measures with an unfamiliar conver-sational partner showed improvement in conversa-tional initiations, in total number of conversationalturns, in frequency of communication breakdowns,and in the frequency and percentage of turns in break-down repair. The subject reported satisfaction withuse of his system in the community.

The case studies reported above suggest that whena complete AAC assessment is conducted, the out-come of AAC intervention may be successful for indi-viduals with severe aphasia. In particular, treatment iseffective when intervention is directed by the aphasicperson’s communication needs and capabilities, andwhen it addresses relevant social and environmentalconsiderations. Although case studies do not demon-strate that the ultimate communication system for an

AAC Aphasiology 263

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

aphasic person will include all AAC elements usedduring recovery, all demonstrate use of AAC systemsclinically and in the natural environment when com-munication needs cannot be adequately met with theuse of natural methods. The last case study also illus-trates the ability of a well-designed AAC system tosupplement and to facilitate use of natural communi-cation methods.

Single-Case Experimental Research

One challenge of empirical AAC and aphasiaresearch is to examine the effectiveness of interven-tions that appear in case studies. This challenge can bemet only by designing studies that adhere to the stan-dards of rigor applied to all research and have docu-mented social validity. In the case of AAC and aphasiaresearch, social validity not only implies significance inthe lives of the population under study but significanceto the disciplines under study. If the accepted prac-tices of assessment for AAC intervention are not incor-porated in planning and conducting empirical research,the validity of the research is in doubt.

Single-case experimental designs offer strength intheir ability to deal with the variability of individualbehavior and their ability to generalize by replicatingthe studies’ findings (Barlow & Hersen, 1984). In single-case research, variability is controlled through the useof repeated measures, either presented sequentiallyover time or in a rapidly alternating manner. General-ization of behavior change may be measured in threeways in single-case research: (1) across subjects, (2)across behaviors, and (3) across settings. Single-caseresearch is particularly well suited to answering ques-tions about which type of treatment is appropriate for agiven client, for a specific type of activity in a given set-ting. Recently published AAC and aphasia single-caseexperimental studies will be reviewed with questionsregarding methodology (variability and generalizabilityof findings) and social validity (relevance in the lives ofclients and clinicians) in mind.

Withdrawal and Reversal Designs

Withdrawal and reversal designs are best usedwhen a treatment effect would not be anticipated aftertreatment is withdrawn. The internal validity require-ments of withdrawal designs demand that the treat-ment effect be demonstrated at three points in time(ABAB designs are therefore superior to AB or ABAdesigns). Replication across at least three subjects isa requirement for external validity in these studies.Additionally, in order to demonstrate a functionaleffect of the independent variable on the dependentvariable, stability must be documented at three pointsprior to initiation of treatment and at each phase dur-ing treatment (Barlow & Hersen, 1984).

Only one study in AAC and aphasia, Garrett’sunpublished dissertation (1993), meets all of the valid-

ity criteria outlined above. The published report basedon Garrett’s dissertation included data from only oneof her subjects, so it does not meet the external valid-ity requirement (Garrett & Beukelman, 1995). Thisstudy used an ABAB’B reversal design to examinethe effect of an augmentative communication strategy(written choice) employed by conversational partnersof severely aphasic subjects. The partner writes twoto five words or phrases that are presented to theaphasic subject as response choices during conver-sation. Dependent measures included the following:(1) proportion of exchanges per topic, (2) compre-hensibility of responses, and (3) accuracy ofresponses. Subjective ratings of the subject’s satis-faction with the interaction also were evaluated. Gar-rett found that the level of all dependent measuresimproved when either thematic or nonthematic writtenchoice was used. The aphasic subject’s satisfactionratings were idiosyncratic and unrelated to the otherdependent variable measures.

Subject selection for the study involved administra-tion of a number of criterion-referenced measuresdesigned to identify individuals who possessed thenecessary communication abilities required for theintervention (e.g., word matching and pointing to writ-ten words to respond to questions in a conversationalcontext). The author also screened for communicationneed by determining that the subjects were not ableto use natural communication methods to respond toquestions in conversation. Personally relevant con-versational topics were selected from a questionnairecompleted by a person familiar with the aphasic sub-jects’ interests and experiences. Each of these pre-measures represents an adaptation of standard AACassessment practices. Together, they lend strength tothe validity of the study’s findings. Although externalvalidity of the unpublished dissertation is not in doubt,unfortunately, the published paper’s presentation ofonly one subject leaves the consumers of AAC andaphasia research in doubt about the generalizability ofthe study’s conclusions.

Multiple-Baseline Design

Studies conducted with multiple-baseline designsare used when treatment effects would be anticipatedafter treatment is withdrawn. A multiple-baselinedesign across behaviors will control for the effects ofa treatment variable as treatment is sequentially initi-ated for different behaviors. A multiple-baselinedesign across subjects examines changes in a singlebehavior across multiple subjects who are, presum-ably, exposed to identical environmental conditions. Amultiple-baseline design across environments exam-ines changes in the behavior of a subject who isexposed to different environmental conditions (Barlow& Hersen, 1984).

A study conducted by Bellaire, Georges, andThompson (1991) used a multiple-baseline design

264 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

across behaviors to investigate the use of communi-cation boards by two subjects with severe Broca’saphasia. This study sought to achieve three objec-tives: (1) to determine whether aphasic individualscould be trained to use communication board vocab-ulary items, (2) to determine whether training wouldresult in generalization to other items or other set-tings, and (3) to determine whether role playing ortraining communication board use in the natural set-ting would improve use of the boards in a coffee-hoursetting. WAB (Kertesz, 1982) scores serve as processand capability assessment. The subjects receivedaphasia quotients of 11.5 and 11.0 on a 100-pointscale. Both subjects scored 0 on all subtests, with theexception of the auditory comprehension subtest, onwhich they scored 5.75 and 4.1, respectively. Scoresof 20.7 and 14.3 were reported for the subjects’ corti-cal quotients. The cortical quotient represents anaggregated score for reading and writing subtests,Raven’s Progressive Matrices (Raven, 1960), and ablock design test. Although no formal communicationneeds assessment was conducted, the subjects weretrained to use communication boards with itemsintended for use during the coffee hour at their nurs-ing home. Generalization of communication board useto a natural setting was measured during these coffee-hour periods.

The authors sequentially trained their subjects toidentify pictures representing vocabulary thataddressed three different communication purposes:(1) items that might be requested during coffee hour,(2) personal or historical information about the sub-jects, and (3) social responses (e.g., “hi” or “thankyou”). With five pictures devoted to each of the threecommunication purposes, the dependent measurewas the ability of each subject to point to the 15 tar-get responses on the communication board.

They found that both subjects were able to usecommunication boards for making requests and forpersonal responses but not for social responses dur-ing training sessions. No generalization to the coffee-hour setting was noted until training in that environ-ment was instituted. Following training in theenvironment, both subjects used their communicationboards to make requests and to communicate per-sonal information during coffee hour, but not to pro-duce social responses, such as “hi” and “thank you.”As the subjects did not use the communication boardfor social responses, the authors suggested that com-munication boards may only be appropriate for com-munication of “specific content” items used for makingrequests or conveying specific information. Theyspeculated that social responses may not have beenused because they could not be depicted on theboards in a nonabstract manner or, as Hux et al.(1994) have suggested, because they were able toconvey these messages using head nods and othernaturally occurring gestures.

The Bellaire et al. (1991) study confirms that a well-designed single-case experimental study may sub-stantiate some of the findings suggested by compre-hensive case studies. The authors have shown that,when individuals with severe aphasia are taught touse communication boards designed to meet theirneeds in a natural setting, learning occurs. They havealso demonstrated the importance of training severelyaphasic people to use communication systems in thenatural setting in order for generalization to take place.

Alternating Treatment Design

An alternating treatment design compares two treat-ments in a single subject. Important procedural ruleshelp to control for the potential confounding effects ofthis design. One rule calls for counterbalancing treat-ments so that an order effect does not occur. Carry-over (always possible when learning is occurring) maybe reduced by separating treatments with a time inter-val, counterbalancing, and slowing alterations in treat-ments (Barlow & Hersen, 1984).

A recent study used an alternating treatment designto explore the benefits of verbal versus nonverbalaphasia treatment on three subjects’ ability todescribe picture stimuli (Avent, Edwards, Franco,Lucero, & Pekowsky, 1995). The verbal treatmentconsisted of a program to improve spontaneous lan-guage production. The nonverbal treatment was aPACE-like program that emphasized gesture, writing,and drawing. Counterbalancing morning and after-noon sessions controlled for order effect. An attemptwas made to control for potential carry-over effectsthrough the use of different stimuli in the two treatmentconditions. Avent et al. did not report assessment pro-cedures or findings for their clients, other than to noteaphasia type and aphasia quotients for each subject.They did not report which communication modalitieswere dominant for each of their subjects prior to initi-ation of treatment, nor did they report language sub-test differences or findings of cognitive testing. There-fore, although the study is methodologically strong,the absence of a comprehensive AAC assessmentcasts doubt on the validity of its conclusions.

As each of the three subjects showed uniqueresponse patterns to the two treatments, the authorsconcluded that optimal treatment programs may beidiosyncratic for individuals with chronic aphasia.They recommended that the stability of baseline datapoints be used as an indicator of treatment outcome,as they interpreted variability of a communicationmodality’s baseline condition to predict greaterimprovement during and following treatment. It is pos-sible, however, that other factors related to the lan-guage or cognitive skills of their subjects explain theapparently idiosyncratic performance. For example, ifbaseline performance is considered an indicator ofnatural communication abilities, one might conclude

AAC Aphasiology 265

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

that the subjects responded best to the treatment thattargeted the communication modality each favorednaturally. Unfortunately, without comprehensiveassessment data for each subject, the validity of anyconclusion is in doubt.

Although not referred to as an alternating treatmentdesign, another recently published study fits thedesign description (Steele, Kleczewska, Carlson, &Weinrich, 1992). This study compares a single aphasicperson’s ability to comprehend instructions giventhrough three different language modalities. Theresearchers counterbalanced treatments within eachsession. However, they did not report attempts to con-trol for carry-over effects. Although comprehensiveassessment information is not provided, it is apparentthat the subject’s communication needs and capabilitieswere considered when designing the study. The subjectscored zero in the naming, reading, and repetition sub-tests of the BDAE (Goodglass & Kaplan, 1983). Theauthors describe his fluency of speech as profoundlyimpaired and melodic line as absent. Auditory compre-hension is described as only slightly less impaired thanexpressive language. Communication need is identi-fied in the authors’ statement that “impetus for the con-trolled study came from pilot work using the C-VIC sys-tem to assist the subject in recipe preparation” (p. 189).

Using a 15-point scale adapted from PICA scoring,Steele et al. (1992) found that their subject consis-tently followed commands better when the commandswere produced by C-VIC (a computer-based systemthat uses icons for communication) versus written orspoken commands. Although no treatment, per se,was administered during the study, an alternatingtreatment design was appropriate for assessing thebenefit of different methods of communicating com-mands to the aphasic subject. The nature of the studydoes not demand a baseline period, and clear differ-ences in performance levels in the three different con-ditions lend strength to the authors’ conclusion that C-VIC is the superior input communication modality forthis type of task with this type of patient.

Experimental and Nonexperimental Group Studies

Experimental group studies offer clear advantageswhen the objective of a study is to make inferencesabout causes and effects. According to Keppel andZedeck (1989), three elements are critical to experi-mental within-subjects group designs: (1) subjectsmust be randomly selected, (2) treatments are manip-ulated by the experimenter, and (3) the experimenterhas control over the conduct of the experiment. Therehave been no experimental group studies conductedto examine AAC interventions with aphasic peoplethat meet all of these criteria.

In the past 5 years of aphasia and AAC literature,there has been one group study of multimodality train-ing with severely aphasic adults (Purdy, Duffy, &

Coelho, 1994). This study used a within-subjectsdesign to determine whether 15 aphasic adults coulduse trained symbols in structured communicationtasks, which modalities they would use, and whetherthey would spontaneously switch between modalitieswhen necessary. The study does not meet the crite-ria of an experimental design, as there is no indicationthat subjects were randomly selected.

Although they did not clearly follow AAC assess-ment practices, the investigators made an effort toconsider communication needs and capabilities oftheir subjects in the study design. Process assess-ment consisted of administration of the PICA (Porch,1967). Criteria for admission to the study required aPICA overall percentile of 25 or greater based on find-ings from Coelho and Duffy (1985). Communicationneeds were considered, as trained symbols wouldlater appear as targets in conversational probes.Vocabulary used in the probes was drawn from theCADL (Holland, 1980) and was “judged to be repre-sentative of everyday communicative activities” (p.347). Subjects were required to be living at home withcommunication partners, as it was thought that theywould be more likely to have opportunities for com-munication and would be more likely to succeed atfunctional communication tasks.

All 20 selected target symbols were trained in eachcommunication modality. However, training taskswere slightly different for each communication modal-ity. Communication board training consisted of theexaminer first pointing to each symbol and identifyingit. The subject was asked to identify the symbol whenits name and a sentence containing its name weregiven. If the subject was unable to identify the symbol,the examiner gave verbal prompts, demonstrated acorrect pointing response, and then asked the subjectto point to the symbol again, giving its name in isola-tion and in a sentence. Gestural training began withthe examiner showing the subject a picture of the tar-get and demonstrating the target gesture while a ver-bal cue for the gesture was provided. Then the sub-ject was asked to imitate the gesture. Physicalprompts were provided if necessary. Verbal re-sponses were taught by the examiner first showing apicture, saying the target word, and asking the subjectto imitate it. If necessary, phonemic, semantic, andvisual placement cues were provided. Training ineach modality ended when subjects reached 80%accuracy on all symbols. Baseline probes continuedin previously trained modalities as training continuedin new modalities. Eighty percent accuracy in at leasttwo of the trained modalities was required for partici-pation in the final measures of the study.

In the testing sessions, the subjects participated intwo communication tasks: one task involved answer-ing questions that required production of the targetsymbols and the second task required the subject tocommunicate the target symbols they identified in 15multisymbol pictures. For both tasks, six dependent

266 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

variables were measured: (1) number of symbols pro-duced spontaneously, (2) number of symbols pro-duced following a cue, (3) method of communicationused, (4) successful attempts to switch modalities, (5)unsuccessful attempts, and (6) the ratio of switchingattempts to opportunities.

Results showed that, following training, subjectsused the verbal modality more frequently in both thesimulated conversation and picture description tasks(p < .01). This was despite the fact that all 15 subjectshad reached 80% criterion for successful use of com-munication board and gestural modalities, and only 4reached criterion during training on the verbal task.The authors also found that, despite training in multi-ple modalities, the subjects switched modalities only37% and 41% of the time, respectively, when oppor-tunities arose in the two tasks. They speculated thatpreference for verbal responses was due to the moreautomatic nature of those responses, and that switch-ing was rare because aphasic subjects do not havethe ability to recognize their failure or plan alternatestrategies. The authors suggested that research isstill needed to examine cognitive variables that influ-ence aphasic adults’ ability to use alternate commu-nication strategies and to assess the benefit of train-ing in modality switching.

This study provides a clear example of researchthat offers little new information relevant to AAC inter-vention. Although the study documents the internalbarrier that switching communication modalities pre-sents for many aphasic individuals, it does notaddress factors that might influence learning ofmodality switching. The barrier of resistance to switch-ing modalities had earlier been identified in the Gar-rett et al. (1989) case study. This case study offeredno evidence that the aphasic subject could switchmodalities without training. In fact, the study docu-mented that successful switching required 7 to 8months of treatment using role playing.

An interesting follow-up question, with more clinicalrelevance, might be to ask what influence differenttreatments have on teaching modality switching. Isrole playing the most effective and economicalmethod for teaching this behavior or are other treat-ments more effective? Bellaire et al. (1991) foundenvironmental training critical to carry-over of com-munication board use to the natural environment. Astudy with greater social validity might compare theeffects of different levels of environmental training onteaching modality switching. Is it necessary to train abehavior in a specific environment, or will the behav-ior be generalized if training takes place in a simulatedenvironment or in a group treatment environment?

Descriptive and Comparative Studies

In addition to the studies described above, whichincorporate accepted research designs, a few studieshave been published that fit into no clear design cat-

egory but are significant to the field in some manner.Most of these studies attempt to link language or lin-guistic mapping theory with AAC interventions inaphasia (Funnell & Allport, 1989; Goodenough-Trepagnier, 1995; Salvatore & Nelson, 1995; Wein-rich, McCall, Weber, Thomas, & Thornburg, 1995).One represented a follow-up to the Coelho and Duffycase study of sign acquisition (Coelho & Duffy, 1990).

Studies that explore the relationship between AACsystem learning and underlying language processesoffer potentially significant information to AAC andaphasia clinicians. By exploring the relationshipbetween symbol use and underlying language com-petence, clinically relevant information is likely toemerge. It is important to remember, however, thatthe methodological weaknesses of these studies donot allow for generalization of the authors’ conclu-sions to all candidates for AAC intervention. Theymay, however, provide clues to assist in process andparticipation assessment for AAC interventions inaphasia. For example, Goodenough-Trepagnier’s(1995) study suggests that C-VIC performance mayidentify preserved linguistic capabilities that are notaccessible through other communication modalities,and that deficits not likely to respond to treatment maybe identified as well.

Unfortunately, some descriptive studies have notalways been interpreted as providing guidelines forexploring internal opportunities or barriers to commu-nication. Coelho and Duffy’s (1985) case study of signacquisition in an aphasic adult is often cited in sub-sequent research as justification for excluding apha-sic individuals with low PICA scores from AAC inter-vention. Use of such weak evidence for determiningoverall candidacy of an aphasic individual for AACintervention represents a contradiction of acceptedAAC assessment practices.

DIRECTIONS FOR FUTURE RESEARCH

Many of the case studies, single-case experimental,descriptive, and comparative studies discussed abovefall short of proving the efficacy (benefit derived underideal conditions) and the effectiveness (benefitderived under typical conditions) of multimodality orAAC treatments for severe aphasia (Blockberger,1993). Case studies provide examples of treatmentpractices that appear to have merit, but only empiri-cal research allows us to prove the usefulness ofthose practices. The absence of comprehensive AACassessment in much of the empirical research andover-reliance on general candidacy questions seri-ously hamper our ability to make clear statementsabout aphasic adults’ ability to use specific communi-cation strategies and tools.

Although the field of AAC has moved well beyondsimple candidacy models, some studies examiningoutcomes of multimodality or AAC intervention stillseek to determine whether aphasic adults are candi-

AAC Aphasiology 267

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

dates for system use (Coelho and Duffy, 1985, 1990).This emphasis on candidacy results in research thatfrequently asks the wrong questions and often resultsin disappointing answers. As Bailey (1983) found inher attempt to study the outcome of a Blissymbolssystem intervention, asking only questions pertainingto candidacy makes it easy to miss more clinically rel-evant information. Had Bailey sought to examine how,not whether, a Blissymbol system benefited an apha-sic adult, the conclusions would have had more ther-apeutic significance.

Less emphasis on questions of overall candidacyand the addition of needs, capabilities, and environ-mental constraints assessment will not weaken futurestudies. Research that incorporates these critical AACassessment practices will lend strength to the ratio-nale for a treatment approach and will make it pos-sible to draw conclusions regarding both the efficacyand the effectiveness of that treatment.

This review of recent research illustrates that AACintervention is a complex endeavor, dependent onassessment practices that consider multiple individ-ual, environmental, and social factors. The difficulty ofdoing good AAC research with severely aphasicadults is further complicated by the nature of aphasiaitself. Just as there is not only one AAC interventionto investigate, there is not one manifestation of severeaphasia. When research subjects’ test scores andother evidence of communication strengths and weak-nesses are considered, it is easy to see the complex-ity of the population under study. The wise investiga-tor of the future will frame research questions withthese considerations in mind.

In the summary of their position paper on aphasiaresearch, Thompson and Kearns (1991) noted thatmost of the applied aphasia studies that they reviewedfocused on evaluating effectiveness and few investi-gated treatment efficiency. They also observed thatresearch examining generalization across settingsand persons was uncommon and that social validationwas often absent from the studies. The same conclu-sions may be drawn regarding the current body ofAAC and aphasia research.

The literature itself provides some clues as to whereour energy should be spent in AAC and aphasiaresearch in the years ahead. It is clear from thisreview that the most clinically relevant information iscontained in case studies and in single-case experi-mental research. As Purdy et al. (1994) have illus-trated, questions of overall candidacy are not clini-cally relevant. Important information related toindividual aphasic subject’s communication needsand capabilities, and to the environmental factorsinfluencing their outcomes, are lost when subjects aregrouped together. This may explain why only onesuch study was found in the recent literature. It alsoexplains why single-case experimental designs aremore appropriate for many of the questions currentlybeing asked about treatment outcomes. These

designs offer the researcher the opportunity to selectsubjects who possess the communication capabilitiesnecessary for a specific intervention. The use of oneto four subjects in a study allows an investigator thefreedom to design a question that considers the com-munication needs and participation patterns of indi-vidual subjects as well as the demands of the exper-imental task. This review provides evidence that,using these methods, it is possible to conduct goodsingle-case experimental research replicated acrossmultiple subjects that is socially valid (Garrett, 1993).

The literature also provides clues as to which ques-tions need to be asked next. This review points to theneed for additional studies that examine treatmenteffectiveness questions, treatment efficiency ques-tions, and questions related to generalization of newlearning to the natural environment.

Treatment effectiveness studies should not focuson the general candidacy questions that many haveasked in the past. Instead, questions should targetspecific treatments that are most effective in facilitat-ing communication among different aphasic commu-nicator types. For this purpose, it would be helpful todesign studies targeted to one of Garrett and Beukel-man’s (1992) five aphasic communicator types (seeAppendix A). Studies such as Bellaire et al. (1991)would answer questions about treatment effective-ness for the controlled-situation communicator. Stud-ies designed to investigate switching behavior amongmultiple modalities would address treatment effec-tiveness questions for comprehensive communica-tors. Studies such as Steele et al. (1992) and Garrettand Beukelman (1995) would address the needs ofthe augmented-input communicator. As there is noresearch currently focusing on the specific needscommunicator, qualitative methodologies might bestisolate the questions that future investigators will seekto answer. A useful first step would be achieved ifqualitative studies explored areas of communicationneed and asked questions regarding participation inspecific settings. An objective of such a line of inquirywould be to identify environmental constraints andopportunities in the natural environments of individu-als with aphasia. For example, what are the commu-nication needs and the environmental barriers andopportunities for aphasic individuals who participate instroke club meetings (Fox, 1996)?

Treatment efficiency studies will be especiallyimportant in coming years as cost containment con-tinues to be a central issue in all phases of rehabilita-tion. Single-case experimental studies will need tocompare teaching methods to determine whichrequire the least time and energy to achieve compa-rable benefits. Such studies might examine direct ver-sus indirect teaching methods. For example, is it moreefficient for a clinician to teach a client to use a sys-tem or for a clinician to teach a spouse to facilitatesystem use? With which treatment method does theclient learn to use the system most effectively and

268 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

most quickly? With which method does the clientachieve the most rapid generalization of system useto the natural environment? A range of similar ques-tions could be developed for different systems andcommunicator types.

As the examples above suggest, questions regard-ing generalization will be tied to treatment effective-ness and to the efficiency of teaching methods or toolsused. However, generalization questions will requireconsideration of factors beyond the complications ofaphasic communicator types and the multitude oftreatment methods available. Generalization ques-tions will also need to consider individual and envi-ronmental influences on system use in natural set-tings. When external and internal influences oncommunication are manipulated, we may begin to seeother influences on behavior in the natural environ-ment that will be more amenable to change than theunderlying linguistic deficit of aphasia. Emphasis onthe environment will expand the focus of generaliza-tion research from considering only issues of linguis-tic disability in aphasia to encompassing the psy-chosocial handicap of aphasia as well.

ACKNOWLEDGMENTS

The authors would like to thank McKay MooreSohlberg and Marilyn Nippold for assistance with edit-ing and suggestions regarding the content of this paper.

Address reprint requests to: Lynn E. Fox, Uni-versity of Oregon, Eugene, OR 97403-5252, USA.

REFERENCES

Aten, J. L. (1986). Functional communication treatment. In R.Chapey (Ed.), Language intervention strategies in adult apha-sia (pp. 266–276 ). Baltimore: Williams & Wilkins.

Aten, J. L. (1994). Functional communication treatment. In R.Chapey (Ed.), Language intervention strategies in adult apha-sia (pp. 292–303). Baltimore: Williams & Wilkins.

Aten, J., Caligiuri, M., & Holland, A. (1982). The efficacy of func-tional communication therapy for chronic aphasic patients. Jour-nal of Speech and Hearing Disorders, 47, 93–96.

Avent, J. R., Edwards, D. J., Franco, C. R., Lucero, C. J., &Pekowsky, J. I. (1995). A verbal and non-verbal treatment com-parison study in aphasia. Aphasiology, 9, 295–303.

Bailey, S. (1983). Blissymbolics and aphasia therapy: A casestudy. In C. Code & D. Muller (Eds.), Aphasia therapy (pp.178–186). London: Edward Arnold.

Barlow, D. H., & Hersen, M. (1984). Single case experimentaldesigns: Strategies for studying behavior change. NeedhamHeights, MA: Simon and Schuster.

Bellaire, K. J., Georges, J. B., & Thompson, C. K. (1991). Estab-lishing functional communication board use for nonverbal apha-sic subjects. Clinical Aphasiology, 19, 219–227.

Beukelman, D. R., & Garrett, K. L. (1988). Augmentative and alter-native communication for adults with acquired severe commu-nication disorders. Augmentative and Alternative Communica-tion, 4, 104–121.

Beukelman, D. R., & Mirenda, P. (1988). Communication optionsfor persons who cannot speak: Assessment and evaluation. InC. A. Coston (Ed.), Proceedings of the National Planners Con-ference on Assistive Device Service Delivery (pp. 151–165).

Washington, DC: RESNA, Association for the Advancement ofRehabilitation Technology.

Beukelman, D. R. & Mirenda, P. (1992). Augmentative and alter-native communication: Management of severe communicationdisorders in children and adults. Baltimore: Paul H. Brookes.

Beukelman, D. R., Yorkston, K. M., & Dowden, P. A. (1985). Com-munication augmentation: A casebook of clinical management.Boston: College-Hill Press.

Blockberger, S. (1993). From efficacy to effectiveness in AAC:Service delivery issues for the next decade. Unpublished man-uscript, University of British Columbia, Vancouver, BritishColumbia, Canada.

Coelho, C. A., & Duffy, R. J. (1985). Communicative use of signs inaphasia: Is acquisition enough? Clinical Aphasiology, 15, 222–228.

Coelho, C. A., & Duffy, R. J. (1990). Sign acquisition in two apha-sic subjects with limb apraxia. Aphasiology, 4, 1–8.

Collins, M. (1986). Diagnosis and treatment of global aphasia. SanDiego, CA: College Hill Press.

Culp, D., & Carlisle, M. (1988). PACT: Partners in augmentativecommunication training. Tucson, AZ: Communication SkillBuilders.

Davis, G. A., & Wilcox, M. J. (1985). Adult aphasia rehabilitation:Applied pragmatics. San Diego, CA: College Hill Press.

Dowden, P., Beukelman, D. L., & Lossing, C. (1986). Serving non-speaking patients in acute care settings: Intervention outcomes.Augmentative and Alternative Communication, 2, 25–32.

Fitzpatrick, P. M., Nicholas, M., & DiNapoli, J. (1988, November).Pragmatics and communication in aphasia groups (PACAC).Paper presented at the annual convention of the AmericanSpeech Language and Hearing Association, Boston, MA.

Fox, L. E. (1990, November). Recreation focused treatment forgeneralization of language skills in aphasic patients. Paper pre-sented at the annual convention of the American Speech-Lan-guage-Hearing Association, Seattle, WA.

Fox, L. E. (1996, November). A qualitative study: Participation pat-terns of aphasic stroke club members. Paper submitted for pre-sentation at the annual convention of the American Speech-Language-Hearing Association, Seattle, WA.

Fox, L. E., & Fried-Oken, M. (1996). Interactive group treatment foraphasia: An AAC alternative. ISAAC ‘96 Proceedings: The 7thBiennial Conference of the International Society for Augmenta-tive and Alternative Communication, 390–391.

Fried-Oken, M. (1992). The AAC assessment cube for adults with severecommunication disabilities. Communication Outlook, 14, 14–18.

Funnell, E., & Allport, A. (1989). Symbolically speaking: Commu-nicating with Blissymbols in aphasia. Aphasiology, 3, 279–300.

Garrett, K. L. (1993). Changes in the conversational participationof individuals with severe aphasia given three types of partnersupport. Unpublished doctoral dissertation, University ofNebraska-Lincoln, Lincoln, Nebraska.

Garrett, K. L., & Beukelman, D. L. (1992). Augmentative communi-cation approaches for persons with severe aphasia. In K. M.Yorkston (Ed.), Augmentative communication in the medical set-ting (pp. 245–337). Tucson, AZ: Communication Skill Builders.

Garrett, K. L., & Beukelman, D. L. (1995). Changes in the interac-tion patterns of an individual with severe aphasia given threetypes of partner support. Clinical Aphasiology, 23, 237–251.

Garrett, K. L., Beukelman, D. L., & Low-Morrow, D. (1989). A com-prehensive augmentative communication system for an adultwith Broca’s aphasia. Augmentative and Alternative Communi-cation, 5, 55–61.

Goodenough-Trepagnier, C. (1995). Visual analogue communica-tion: An avenue of investigation and rehabilitation of severeaphasia. Aphasiology, 9, 321–341.

Goodglass, H., & Kaplan, E. (1983). Boston diagnostic examina-tion for aphasia. Philadelphia: Lea & Febiger.

Helm-Estabrooks, N., & Albert, M. L. (1991) Manual of aphasiatherapy. Austin, TX: Pro-ed.

AAC Aphasiology 269

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

Helm-Estabrooks, N., Ramsberger, G., Morgan, A. R., & Nicholas,M. (1989). Boston assessment of severe aphasia. Chicago:Riverside Press.

Holland, A. L. (1980). Communicative abilities in daily living. Balti-more: University Park Press.

Holland, A. L. (1982). Observing functional communication of apha-sic adults. Journal of Speech and Hearing Disorders, 47, 50–56.

Hux, K., Beukelman, D. R., & Garrett, K. L. (1994). In R. Chapey(Ed.), Language intervention strategies in adult aphasia (pp.338–357). Baltimore: Williams & Wilkins.

Kagan, A., & Gailey, G. (1993). Functional is not enough: Trainingconversation partners for aphasic adults. In A. L. Holland & M.Forbes (Eds.), Aphasia treatment: World perspectives (pp.199–225). San Diego, CA: Singular.

Keppel, G., & Zedeck, S. (1989). Data analysis for researchdesigns: Analysis of variance and multiple regression/correlationapproaches. New York: W. H. Freeman.

Kertesz, A. (1982). Western aphasia battery. New York: Grune &Stratton.

Light, J. (1988). Interaction involving individuals using augmentativeand alternative communication systems: State of the art and futuredirections. Augmentative and Alternative Communication, 4, 66–82.

Lomas, J., Pickard, L., Bester, S., Elbard, H., Finlayson, A., &Zoghaib, C. (1989). The communicative effectiveness index:Development and psychometric evaluation of a functional com-munication measure for adult aphasia. Journal of Speech andHearing Disorders, 54, 113–124.

Lubinski, R. (1994). Environmental systems approach to adultaphasia. In R. Chapey (Ed.), Language intervention strategies inadult aphasia (pp. 270–291). Baltimore: Williams & Wilkins.

Lyon, J. G. (1989). Communicative partners: Their value in reestab-lishing communication with aphasic adults. In T. E. Prescott(Ed.), Clinical aphasiology (pp. 11–18). Austin, TX: Pro-Ed.

Lyon, J. G. (1992). Communication use and participation in life foradults with aphasia in natural settings: The scope of the problem.American Journal of Speech-Language Pathology, 1, 7–14.

Meyers, P. S. (1980). Visual imagery in aphasia treatment. In R. H.Brookshire, (Ed.), Clinical aphasiology (pp. 68–77). Minneapo-lis, MN: BRK Publishers.

Mirenda, P. (1993). AAC: Bonding the uncertain mosaic. Augmen-tative and Alternative Communication, 9, 3–9.

Morgan, A., & Helm-Estabrooks, N. (1987). Back to the drawingboard. A treatment program for nonverbal aphasic patients. In R.H. Brookshire (Ed.), Clinical aphasiology (pp. 64–71). Min-neapolis, MN: BRK Publishers.

Parr, S. (1996). Everyday literacy in aphasia: Radical approaches tofunctional assessment and therapy. Aphasiology, 10, 469–479.

Porch, B. E. (1967). Porch index of communicative ability. PaloAlto, CA: Consulting Psychologists Press.

Purdy, M. H., Duffy, R. J., & Coelho, C. A. (1994). An investigationof the communicative use of trained symbols following multi-modality training. In M. Lemme (Ed.), Clinical aphasiology (pp.345–357). Austin, TX: Pro-Ed.

Rao, P., & Horner, J. (1978). Gesture as a deblocking modality ina severe aphasic patient. In R. H. Brookshire (Ed.), Clinicalaphasiology (pp. 180–187). Minneapolis, MN: BRK Publishers.

Raven, J. C. (1938). Coloured progressive matrices. London: H. K.Lewis.

Raven, J. C. (1960). Guide to the standard progressive matrices.London: H. K. Lewis.

Rosenbek, J. C., LaPointe, L. L., & Wertz, R. T. (1989). Aphasia:A clinical approach. Austin, TX: Pro-Ed.

Salvatore, A. P., & Nelson, T. R. (1995). Training novel languagesystems in severely aphasic individuals: How novel is it? Clini-cal Aphasiology, 23, 267–277.

Schuell, H. (1965). The Minnesota test for the differential diagno-sis of aphasia. Minneapolis, MN: University of Minnesota Press.

Schuell, H. (1974). The treatment of aphasia. In L. F. Sies (Ed.),Aphasia theory and therapy: Selected lectures and papers ofHildred Schuell (pp. 137–152). Baltimore: University Park Press.

Schuell, H., Jenkins, J. J., & Jimenez-Pabon, E. (1964). Aphasiain adults: Diagnosis, prognosis and treatment. New York: Hoe-ber Medical Division, Harper & Row Publishers.

Shewan, C. M. (1986). The history and efficacy of aphasia treat-ment. In R. Chapey (Ed.), Language intervention strategies inadult aphasia (pp. 28–43). Baltimore: Williams & Wilkins.

Simmons, N. N. (1989). A trip down easy street. In T. E. Prescott(Ed.), Clinical aphasiology (pp. 19–29). Austin, TX: Pro-Ed.

Skelly, M., Schinsky, L., Smith, R., & Fust, R. (1974). AmericanIndian sign (Amerind) as a facilitator of verbalization in the oralapraxic. Journal of Speech and Hearing Disorders, 39,445–456.

Sparks, R., & Holland, A. L. (1976). Method: Melodic intonationtherapy for aphasia. Journal of Speech and Hearing Disorders,41, 287–297.

Steele, R. D., Kleczewska, M. K., Carlson, G. S., & Weinrich, M.(1992). Computers in the rehabilitation of chronic, severe apha-sia: C-VIC 2.0 cross-modal studies. Aphasiology, 6, 185–194.

Stephenson, W. (1953). The study of behavior. Chicago: Univer-sity of Chicago Press.

Stuart, S. (1991). Topic and vocabulary use patterns of elderlymen and women of two age cohorts. Unpublished doctoral dis-sertation, University of Nebraska-Lincoln, Lincoln, Nebraska.

Thompson, C. K., & Kearns, K. P. (1991). Analytical and technicaldirections in applied aphasia analysis: The midas touch. Clini-cal Aphasiology, 19, 41–54.

Wechsler, D. (1955). Wechsler adult intelligence scale manual.New York: Psychological Corporation.

Weinrich, M., McCall, D., Weber, C., Thomas, K., & Thornburg, L.(1995). Training on an iconic communication system for severeaphasia can improve natural language production. Aphasiol-ogy, 9, 343–364.

Wepman, J. (1972). Aphasia therapy: A new look. Journal ofSpeech and Hearing Disorders, 37, 203 – 214.

Wertz, R. T. (1984). Language disorders in adults: State of the clin-ical art. In A. Holland (Ed.), Language disorders in adults (pp.1–77). San Diego, CA: College-Hill Press.

Yates, A. J. (1954). The validity of some psychological tests ofbrain damage. Psychological Bulletin, 51, 359–379.

Yorkston, K., & Karlan, G. (1986). Assessment procedures. In S.W. Blackstone (Ed.), Augmentative communication: An intro-duction (pp. 163–195). Rockville, MD: ASHA.

270 Fox and Fried-Oken

Dow

nloa

ded

By: [

B-on

con

sorti

um] A

t: 15

:20

3 M

ay 2

007

AAC Aphasiology 271

APPENDIX

Assessment of Capabilities and Communicator Type

Type Communication Skill

Basic-Choice Communicator Points to clothing items given choice during morning dressing routine

Points to photos in catalog to answer “favorite outfit” questionLooks up when greetedTakes objects, returns them

Controlled-Situation Communicator Attends to printPoints to photos or picture symbols of needs (n = 2) to answer

questionsCan confirm or select topics of interestCan point to or look at written choices to answer conversational

questionsAware of daily routine (e.g., gets glasses before therapy)

Comprehensive Communicator Speaks some wordsWrites some words or word fragmentsCan communicate by drawing schematics, maps, objectsCan locate items by category (structured task OK)Can communicate a specific word by pointing to first letterGesturesPantomimesRecognizes own errorsRecognizes communication breakdownsDemonstrates some pragmatic competence in discourseKnows which communication modality to use and whenWants to communicate in more than one setting with more than

one partnerInitiates questions and comments

Specific-Need Communicator Has indicated need to perform specific communication task more efficiently:

• talking on the phone• writing letters• saying prayers• saying names of family members• signing name• making purchases• making lists• making memos• communicating destination on public transportation system• calling for assistance

Demonstrates most skills from other communicator types

Augmented-Input Communicator Attends to printAttends to gesturesWritten key words appear to enhance comprehensionPartner gestures appear to enhance comprehensionSignals lack of understanding/breakdowns

From Garrett & Beukelman (1992).