EFFECTS OF TOPICAL VOCA MESSAGES ON CONVERSATIONS WITH SEVERELY APHASIC COMMUNICATORS

Kathryn L. Garrett, Ph.D., CCC-SLP     Duquesne University, Pittsburgh, PA

Kelly M. Sestric, MS, CFY     Mt. Lebanon School District, Pittsburgh, PA

A  poster presentation at the ASHA 2000 Convention, Washington D.C.

Abstract

In this single subject experiment, two communicators with severe aphasia conversed with partners about a personal event and a news event in baseline, VOCA topical messages, and VOCA + instruction conditions.  Dependent variables included exchanges per topic, participant initiations, successful exchanges, modality, and elapsed topic time.  Qualitative ratings were also obtained.

 

Review of the Literature/Statement of the Problem

Because of their extensive communication deficits, many people with severe aphasia are nonspeaking or have extremely limited expressive communication skills.  They have difficulty sharing adequate amounts of specific information to sustain in-depth topical conversations; therefore the communication partner often carries the burden of an interaction (Packard & Hinkley, 1997).  A variety of alternative communication strategies have been proposed to assist persons with severe aphasia to communicate more effectively, including a written choice conversational strategy (Garrett & Beukelman, 1995); pre-stored information in communication notebooks or voice output devices (Garrett & Beukelman, 1992); and use of naturalistic, nonverbal communication modalities such as writing, gesturing, and drawing (Doyle & DeRuyter, 1995; Glosser, Weiner, & Kaplan, 1986; Lyon, 1995).   Although having external, printed information available may assist some nonspeaking persons with aphasia to communicate, it is not clear whether individuals with very severe aphasia can initiate consecutive messages in a conversational exchange.  Garrett, Heldman, and Dunning (1994) described a case study in which a gentleman with severe aphasia successfully learned to use a voice output communication aid (VOCA) to communicate consecutive messages in a practiced conversation about farming.  In this particular case, the VOCA appeared to simplify the message initiation process.  However, no controlled data are presently available to validate this intervention.

 

Specifically, the present study investigated:  1) the impact of pre-stored topical messages in a voice output communication aid (VOCA) on the conversational behaviors of persons with severe aphasia; and 2) the ability of persons with aphasia to learn to purposefully use these pre-stored messages in conversations given specific instructional intervention.  3) the impact of  topic types on interactions (current event vs. personal event).  A modified single subject (ABAC) sequential design was utilized.

 

Experimental Conditions/Procedures:    

a) Baseline (condition A):  Participants were instructed to discuss a pre-selected personal event or current event in detail for as long as possible with their partner in a dyadic exchange.  No context or messages were provided.

b) VOCA-Only Condition (condition B):  Prior to data collection, the primary investigator stored 2-to-4 messages for each current event or recent personal event (total of 6-to-8 messages) on the VOCA.  Messages included: 1) a starter question – “Have you heard about (topic)?; 2) opinions represented with a 3-point graphic scale (good—so-so—bad); 3) “Yes” and “No” if the participant could not communicate these with head nods or verbally; and 4) 3-4 messages about the current event or personal event.  Messages were reviewed with the participant prior to data collection.

c) Instructional Phase (condition C ):  Following review of topical VOCA messages, the primary investigator then provided the following verbal and written instructions:

1.  You can use the machine to tell them something interesting! [demo]

2.  You can ask them a question to get things going! (demo: Have you heard about..........?):

3.  Tell them your opinion with the rating scales on the machine.

4.  Take turns - first you go, then she goes.

5.  Here’s an example of a conversation (shows them a large print written script)

 

The primary investigator and the participant then role-played a conversation about the current event topic using a printed script as a reference if necessary.  Speech-language pathology students served as a conversational partners in the dyadic condition.  They were trained and given written instructions prior to the initiation of the study to ensure that their interactional behaviors conformed to the procedural requirements of this investigation.  An emphasis was placed on teaching the experimental partners to interact in an open-ended, naturalistic manner with the participants.  Target behaviors for the experimental partners included:  pausing, asking open-ended questions, attempting to confirm participants’ turns, limiting the number of yes/no questions, and changing topics after 2 consecutive attempts to repair communication breakdowns. Topics were changed for each experimental session.  Messages for recent personal events were obtained by having the participant’s spouse or significant other complete an information form.  Current events were selected from a daily newspaper.  Topics were introduced in alternate order across all sessions and conditions to prevent an order effect.

 

Dependent Variables:  Quantitative measures derived from session transcriptions included: number of conversational turns, number of information exchanges, number and proportion of successful exchanges, number and proportion of participant-initiated exchanges, modality, and elapsed time per topic.  Additionally, all participants rated their own communicative competence, comfort, enjoyment, and partner’s competence after each session.  Data were initially plotted by individual sessions in accordance with a single subject analysis technique.  However, inherent variability (perhaps because of the uniqueness of each topic) caused any potential effect to be obscured.  To compensate for variability, quantitative data on turns, exchanges, initiations, etc. were averaged by condition/topic and graphed.  Results are summarized here:

 

CURRENT EVENT TOPIC DATA

Participant 1 (T.K.)

Time (secs)

# Exch.

# Ss turns

Turns/exch

% SS inits

%success

Ave A

181

7.6

25.6

6.37

.3

.14

SD

100

4.16

16.9

1.47

.18

.15

Ave B

204

11

24.5

4.61

.39

.3

SD

87.57

3.91

8.58

.46

.09

.18

Ave C

268

12

32.5

5.6

.4

.08

SD

70

2.82

12.02

1.27

.02

.02

Participant (A.A.)

Time (secs)

# Exch

#SS Turn

Turns/ex

%SS inits

%success

Ave A

162

8.5

22.8

6.04

.14

.59

SD

.69

1.73

6.18

1.19

.14

.19

Ave B

166.2

13.8

32

5

.25

.84

SD

25.5

3.56

10

1.3

.08

.07

Ave C

136

13.8

24.3

3.7

.28

.68

SD

33.3

2.98

5.18

.91

.21

.23

 

PERSONAL EVENT TOPIC DATA

Participant 1 (T.K.)

Time (secs)

# Exch.

# Ss turns

Turns/exch

% SS inits

%success

Ave A

173

6.4

19.4

6

.24

.53

SD

48

2.5

5.9

1.8

.11

.09

Ave B

285

12.67

31.3

4.79

.44

.78

SD

157

5.5

20

1.21

.14

.07

Ave C

277

14

34.5

4.89

.44

.82

SD

36.7

2.82

7.77

.08

.08

.09

Participant (A.A.)

Time (secs)

# Exch

#SS Turn

Turns/ex

%SS inits

%success

Ave A

122

8.5

20

5.4

.08

.30

SD

52

3.41

8.2

1.5

.09

.28

Ave B

157.8

13.8

32

5.2

.11

.1

SD

76.4

5.26

15

.99

.08

.11

Ave C

162.7

18.75

34.75

3.83

.18

.22

SD

70.6

3.3

10.21

1.04

.11

.16

 

These descriptive data revealed potential Condition effects:  Length of interaction for Participant 1; total communicative turns increased for both participants for the Personal Event topic.  % Initiations increased for both topics from baseline to VOCA conditions across both participants.  Participant 1 demonstrated incremental increased in successfulness of exchanges in  the VOCA condition.  However, increases were very slight for Participant 2.  Participant Effect:  Observations that Participant 1 was a more tenacious, resourceful communicator were supported empirically – there was more of a tx effect with regard to # of exchanges, initiations, and % successfulness for T.K. than AA.  Topic Effect:  Length of interaction was greater for personal event; however, initiations increased for the current event (AA.), as did successfulness of exchanges in the VOCA condition. Additional statistical analyses will be performed to determine if statistically significant differences existed between conditions.