Aphasiology, 2015

Vol. 29, No. 7, 781798, http://dx.doi.org/10.1080/02687038.2014.989953

Effects of impairment-based individual and socially oriented group

therapies on verb production in aphasia
Elizabeth Louise Hoovera*, David Caplanb, Gloria Watersa and Andrew Budsonc
a
Speech-Language and Hearing Sciences, Boston University, Boston, MA 02215, USA;
Neuropsychology Laboratory, Massachusetts General Hospital, Boston, MA 02114, USA;
c
Cognitive & Behavioral Neurology, VA Boston Healthcare System, Jamaica Plain, MA 02130, USA
b

(Received 28 April 2014; accepted 16 November 2014)

Purpose: Two philosophies of intervention exist in aphasia rehabilitation: impairmentbased approaches and socially oriented approaches. Both approaches have been shown
to improve communication in persons with aphasia, but no studies have directly
compared the effects of each approach or a combined approach on a targeted linguistic
skill. This article explores the effects of individual and group therapies used both in
isolation and in combination on verb production in aphasia.
Methods: Twelve individuals with chronic aphasia were trained on transitive verbs
under three conditionsindividual, group and combinedover a 6-week interval.
Treatment was counterbalanced across subject and training groups. A delayed-treatment, within-participant design was used. Verb probe data were collected at 10 points
throughout the study. Language measures were taken at two intervals pre- and two
intervals post-treatment. Functional, narrative and quality-of-life measures were taken
once pre- and once post-treatment.
Results: Significant change was observed on linguistic, functional communication and
quality-of-life measures. There was no significant effect of treatment condition.
Conclusions: The results provide evidence of linguistic and psychosocial change in
individuals with chronic aphasia following this treatment. Results failed to find that
one treatment condition was superior to others.
Keywords: aphasia; treatment; verb production; socially oriented treatment; impairmentbased treatment

Language interventions are frequently classified along a continuum of naturalness (Paul &
Cascella, 2007). At one end of the continuum are impairment-based approaches that aim
to remediate a particular area of language. A cognitive neuropsychological model of
aphasia deficits and intervention is often used in this approach. This approach begins
with a linguistic evaluation to identify the linguistic process(es) disrupted in an individual.
Therapeutic tasks are subsequently developed to remediate or compensate for that
damaged process. Successful treatments often utilise models of cognitivelinguistic processing and have been shown to improve language performance in individuals with
aphasia (Thompson & Shapiro, 2005; Wertz et al., 1981; Whitworth, Webster, &
Howard, 2005).
Impairment-based interventions are typically delivered in an individual setting.
Sessions are didactic and frequently follow a request-response-evaluation sequence in
which the therapist requests performance (e.g., asks client to name a picture), the client
*Corresponding author. Email: ehoover@bu.edu
2014 Taylor & Francis

782

E.L. Hoover et al.

responds and the therapist evaluates the clients response (Simmons-Mackie, Elman,
Holland, & Damico, 2007). By controlling the structure and pace of the session, it is
possible to focus on and increase the volume of practice items (Damico, SimmonsMackie, Oelschlaeger, & Tetnowski, 2000; Horton, 2004).
A contrasting approach, found at the other end of the continuum of naturalness, is a
participation-based or socially oriented approach. Foremost in socially oriented
approaches is The Life Participation Approach to Aphasia (LPAA Project Group, 2008).
LPAA encourages reengagement in life throughout the rehabilitation process, strives to
empower the individual with aphasia and reduce the consequences of aphasia in the
individuals quality of life. Group treatment is an example of a socially oriented or
participation-based approach. In this treatment, individuals engage in functional language
tasks, such as group-oriented conversation. Because conversation is not scripted, the
participant has a greater choice in how he or she responds to a conversational prompt
that in turn fosters a more natural and equal role in the communicative exchange. Studies
support the use of conversation group treatments to improve language performance in
individuals with aphasia (Elman & Bernstein-Ellis, 1999a; Wertz et al., 1981).
Considerable evidence exists in the literature to support both these types of interventions (Martin, Thompson, & Worrall, 2008). However, research to date has not compared
the effects of these types of interventions or investigated the benefits of an integrated
application of both impairment-based individual and socially oriented group interventions.
The purpose of this article is to determine the effects of treatment type (impairmentbased, socially oriented or combined) on verb production in individuals with chronic
aphasia. A specific treatment motivated by each philosophy described earlier was used.
Later, we review the evidence for each treatment.

Impairment-based individual treatment of verb production

Verb network strengthening treatment (VNeST, Edmonds, Nadeau, & Kiran, 2009) is an
impairment-based treatment that focuses on the retrieval of verbs within a sentence. This
intervention, based on theories of cognitive psychology and generative semantics (Plaut,
1996), identifies the verb as the predicate core of all simple sentences. Edmonds argues
that thematic roles are features of the verb itself and that one verb/event can have a
multitude of related thematic roles, potentially representing a large network of schemas/
semantic relationships. VNeST aims to improve the lexical retrieval of content words in
the sentence context by promoting systematic retrieval of verbs and their thematic roles.
Three articles have detailed the effects of VNeST on individuals with aphasia
(Edmonds & Babb, 2011; Edmonds et al., 2009; Edmonds, Ojeda, Mammino, & Wu,
2013). In total, 17 participants with chronic profiles of aphasia participated in the studies.
All were trained on 10 transitive verbs using pairs of agents/and patient, e.g., chef/sugar,
carpenter/lumber, with the verb measure. Edmonds and colleagues trained verbs with
more specific semantic uses versus general verbs, e.g., fly or drive versus go, to
increase the specificity of the communicative act. Treatment consisted of five steps during
which the participant was asked to generate agents given a verb, followed by patients,
which corresponded to the agentverb pairings. Choices with foils were presented if the
participant struggled to generate these nouns. Next, participants were asked wh-questions
about the agentpatient pair to expand upon the sentence, e.g., Where does the baker
measure the sugar? Finally, 12 sentences were read to the client containing the target verb
combining all agent/patient options, and the participant was asked to make a semantic

Aphasiology

783

judgment about the accuracy of the sentence. Treatment was provided twice a week for
1.5- to 2-hour periods for between 6 and 10 weeks.
Together, the results of these studies showed replicated improvement and generalisation of confrontation naming skills of nouns and verbs, sentence production and
discourse. Generalisation to sentence production for sentences containing trained and
untrained semantically related verbs was tested weekly and immediately after treatment
for all participants. The probe was also administered at a 3-month interval for the 11
participants in the last two studies. Overall, results showed that the group improved on
sentences with trained and semantically related untrained verbs following treatment. No
change was observed in the control task. Stable performance was observed for 7 of 11
participants at 3 months post-treatment, indicating maintenance of improvement for both
trained and untrained probe.
No change was observed in sentence production as measured by the Northwestern
Assessment of Verbs and Sentences (Cho-Reyes & Thompson, 2012). Significant changes
of pre- to post-treatment scores on the Western Aphasia Battery-Revised (WAB-R;
Kertesz, 2006) and the communicative effectiveness index (Lomas et al., 1989) were
also observed, indicating improvements both on formal measures of language performance and in functional communication as reported by the caregiver. Maintenance data
were not available for these formal measures of generalisation. Treatment effects in
connected speech beyond the sentence level was not a focus of the study, so it is unknown
whether these learned items resulted in generalisation in discourse production or across
communicative partners.
While sample sizes were small in each of the VNeST studies described earlier, these
data together suggest that VNeST is an effective treatment to promote production of verbs
and nouns in individuals with aphasia. Further study is needed to explore the generalisability of this treatment to other language of greater complexity (sentence/discourse) and to
other communicative environments.
Socially oriented group treatment
Conversation group treatment is one socially oriented approach intended to improve social
communication, which has been shown to improve discrete language skills in people with
aphasia as measured by total scores on formal language tests (Elman & Bernstein-Ellis,
1999a).
Wertz and colleagues (1981) completed the first large-scale, randomised control trial
to compare the efficacy of conversation group treatment of aphasia to individual aphasia
treatment. Five Veterans Administration Medical Centers participated in the study. Sixtyseven participants with aphasia, 4 weeks post-onset of a single stroke on entrance to the
study, were randomly assigned either individual or group treatment for a 44-week treatment programme. Individual treatment consisted of 4 hours of stimulus-response treatments in all communicative modalities plus 4 hours of machine-assisted treatment and
speech-language drill. Group treatment consisted of 4 hours of clinician-led group
discussion treatment, supplemented by 4 hours of recreational activities. Results showed
that groups in both cohorts made statistically significant gains pre- to post-treatment on
the overall communicative ability on the Porch index of communicative ability (PICA;
Porch, 1967), word fluency measure (Borkowski, Benton, & Spreen, 1967) and a conversation rating scale. The individual treatment group demonstrated significantly better
performance on the PICA written language tasks; no other significant differences were
observed between individual and group treatment groups on any measure at 33 weeks and

784

E.L. Hoover et al.

48 weeks of treatment. Caution must be observed in interpreting these results, given that
participants were still in the acute stage of recovery (enrolled at 4 weeks post-onset) and
therefore still within the window of spontaneous recovery.
Elman and Bernstein-Ellis (1999a) enrolled 24 persons, greater than 6-months postonset of aphasia, into two treatment groups (immediate and deferred) to determine the
effect of speech-language group treatment compared to socialisation group treatment. The
treatment condition consisted of 5 hours of language group treatment each week over a 4month period. The deferred treatment group engaged in support, performance or movement groups to control for the effects of social contact. Groups were conversational in
format and encouraged communicative initiation and overall effectiveness. Results
showed statistically significant improvement of scores on language-dependent measures
(SPICA and WAB) only after treatment. These results extend the findings of Wertz et al.
(1981) due to the inclusion of a no-treatment period and demonstrate that conversational
group treatment is effective in improving language function as measured by a standardised
battery in chronic stages of aphasia.
Elman and Bernstein-Ellis (1999b) also analysed the data taken from qualitative
interviews with participants at the Aphasia Center of California in the study described
above (Elman & Bernstein-Ellis, 1999a). Participants reported several psychosocial and
linguistic benefits. Caregivers also noted comparable psychosocial gains such as improved
confidence, improved social skills, improved independence and increased motivation.
Van der Gaag et al. (2005) evaluated the impact of attending an aphasia centre by
quantifying participants perceptions of their communication and quality of life. Over
20 weeks, 28 participants with aphasia received an average of 1.7 hours a week of various
socially oriented communication group therapies plus a total of 8 hours of counselling.
Results showed statistically significant changes were on one of the quality-of-life measures (EuroQoL (EQ-5D; EuroQol group, 1990) and on a communication rating scale.
Without adequate control data, these measures are subject to rater bias and should be
interpreted with caution.
Vickers (2010) measured social networks and ratings of social participation in 28
persons with chronic aphasia who attended a bi-weekly, socially oriented aphasia group
treatment programme compared to 12 persons who had received no group treatment. All
participants responded to interviews to complete The Social Networks Communication
Inventory (Blackstone and Hunt-Berg, 2003), The Survey of Communicative and Social
Participation (Vickers & Threats, 2007) and The Friendship Scale (Hawthorne, 2006).
Between-group comparisons revealed that participants who attended the aphasia group
treatment programme reported spending more time on average with friends and acquaintances and reported reduced levels of perceived isolation than those who did not attend the
programme. No significant differences between groups were noted on any of the communication variables between the two groups. Significant methodological issues exist in
this study in the way the sample was obtained (non-random convenience sampling with
unequal group sizes) and potential bias during interviews(authors of one measure conducted the interviews and developed the group treatment programme). However, these
data offer some support for the claims that individuals with aphasia are at risk for isolation
and that group treatment positively impacts the quality of life.
In summary, studies support the use of both impairment-oriented and socially oriented
approaches in remediation of aphasia, and both lead to treatments that strive to improve
function for the individual with aphasia. However, these approaches are rarely provided
concurrently, or in an integrated manner. Often group treatment is considered as an
adjunct to individual therapy (Elman & Bernstein-Ellis, 1999a). If group treatment is

Aphasiology

785

provided concurrently with individual therapy in a subacute or outpatient setting, it is

frequently administered in a separate facility and/or by different clinicians with little
collaboration towards the participants goals. It would seem that providing theoretically
sound impairment-based treatment and explicitly training target items in a naturalistic
functional environment would provide both individualised attention and psychosocial
benefits in one cohesive focused treatment approach. Further, it would allow for increased
practice of trained items across the linguistic hierarchy (word, sentence, discourse levels)
and to new communication partners. Logic therefore suggests that the combination of both
approaches may be more beneficial than one approach alone. This study was designed to
investigate the relative benefits of these approaches on the remediation of a specific
linguistic skill in aphasia and to determine whether improvements of targeted skills are
greater in an integrated context compared to those trained in individual or group contexts
alone.
This study addressed the following questions:
(1) Does a treatment approach which combined an individual, impairment-based
treatment (VNeST) and a socially oriented group treatment result in increased
verb production compared to only individual or group treatment?
(2) Does such a treatment approach result in a linguistic change for individuals with
aphasia?
(3) Does such a treatment approach result in increased verb production in connected
speech?
(4) Does such a treatment approach result in functional communication and qualityof-life improvements for the individual with aphasia?

Methods
This study employed a within-subject, delayed-treatment, experimental design, using
counterbalanced, repeated measures. Verb treatment was provided for participants in
twice-weekly sessions of impairment-based individual and socially oriented group treatment. Each subject received 2.25 hours of individual and 2.25 hours of group therapies
across two treatment days, for 6 weeks. Different verbs were treated in individual
treatment, group treatment, and both individual and group treatments.
Subject selection
Twelve participants with more than 6 months post-onset of a single, language-hemisphere
dominant cerebrovascular accident participated in this study. Power was calculated using
the G*Power 3.1 software program (Buchner, Erdfelder, Faul, & Lang, 2009). For post
hoc Wilcoxon signed-rank test analyses, an effect size of 0.8 (Robey, 1994) can be shown
with a total sample size of 12.
Participants were native English speakers who were diagnosed with aphasia by a
licensed speech-language pathologist (SLP). As noted on previous neuropsychological or
speech-language reports, or during a screening prior to the enrolment in the study, all
participants demonstrated the following characteristics: a non-fluent profile of aphasia
(Brocas), difficulty with verb and sentence production and sufficient auditory comprehension skills in order to comprehend and participate in all tasks presented. Participants
were recruited from the Aphasia Resource Center at the Boston University and through

786
Table 1.

E.L. Hoover et al.

Participant demographic information.

Subject

Age

Gender/ethnicity

Education (years)

Time post-stroke (years)

1
2
3
4
5
6
7
8
9
10
11
12
Mean

65
54
59
54
56
53
56
48
70
57
65
65
58.5

Male/C
Male/AA
Female/C
Female/AA
Female/C
Male/C
Male/C
Male/C
Male/C
Female/C
Female/C
Female/C

16
11
16
12
12
12
10
16
12
20
12
18
13.9

11
20
6
1.5
6
11
12
8
2.5
12
12
3
8.75

Note: C = Caucasian, AA = African American.

referrals from other local treatment centres. Table 1 details the demographic data for the
participants.

Treatment
Three sets of obligatory or optionally transitive verbs were created. Each set contained 27
verbs that were associated with one of nine functional conversation topics, such as
dining, travel, occupation and news/current events. The verbs in each topic were matched
for the number of syllables and frequency using the Kucera and Francis (1967) frequencies. Most verbs were monosyllabic and high frequency, with a mean frequency of 124
occurrences per million (median 88) (Kucera & Francis, 1967). The verb sets are shown in
the Appendix.
The 12 recruited participants were randomly assigned to three treatment groups,
hereafter referred to as subject group 1, 2 and 3. Each subject received treatment on
each verb set, one in each training condition. The verb sets were counterbalanced across
subject groups to ensure that each verb was treated equally often in each condition.
Individual impairment-oriented treatment consisted of a modified version of
Edmonds VNeST. Each participant viewed printed verb cards from a conversation
topic one at a time and then generated three appropriate subject/agent and object/patient
pairs to create a subjectverbobject sentence. If a participant struggled to generate agent/
patient pairs, three written choices were provided as a cue. Once three sentences were
generated, the participant repeated each sentence three times before moving on to the next
verb. In keeping with Edmonds protocol, questions were also posed to expand the
utterance. For example, if a participant generated the sentence The tourist catches the
ferry, an expansion probe such as when did the tourist catch the ferry? was asked.
These lengthier sentences were then repeated three times each. Edmonds final step in the
protocol required participants to make grammaticality judgments on each of the sentences.
This step was only utilised to support corrections of incorrect sentences if participants
made grammatical errors.
Socially oriented group treatment followed a conversational format described in the
literature (Elman, 2007; Elman & Bernstein-Ellis, 1999b; Simmons-Mackie & Chapey,

Aphasiology

787

2008; Simmons-Mackie et al., 2007). Discourse was generated using the topic headings
detailed in the verb lists (Appendix). Participants also engaged in language games,
functional scripts and discussion using the verbs on a list surrounding each conversational
topic.
Licensed SLPs or graduate students in speech-language pathology directly supervised
by an SLP provided all individual treatments. Group treatments were provided by a
licensed SLP. Orientation to the programme, followed by training with observation, was
provided to all clinicians prior to beginning the treatment to ensure treatment fidelity.
Additionally, meetings were held after each session to discuss protocol and client
performance.
To ensure that each verb set received an equal amount of treatment time across the
programme, weekly treatment was balanced for overall time and treatment order for each
participant.
Measurements
The following measurements were administered at four intervals: 1 month before treatment, immediately before the start of treatment, immediately after treatment and at 1month follow-up.
(1) Northwestern verb naming test (VNT; Thompson, 2002)
(2) Sentence production subtest of the Psycholinguistic Assessment of Language
(PAL; Caplan & Bub, 1990)
(3) The Philadelphia naming test (PNT; short form) (Walker & Schwartz, 2012)
(4) Picture description narrative task using birthday and rescue scenes (Nicholas &
Brookshire, 1993)
These measures were chosen because they are commonly used clinical tools which
provided information about verbal expression in both specific (constrained word and
sentence level).
Treatment probes were administered at the above time points as well as at the end of
the second day of individual treatment each week to track acquisition of the verbs over
time. The probes contained photographs of individuals acting out the target 81 verbs and
were organised by conversational category. Prior to the study, the probe was validated
on 12 healthy adults. Eighty per cent agreement across the group was achieved for each
picture in the probe. The categories in the verb probes were presented in random order
each week to minimise a learning effect. Responses were counted as correct if they were
verbs that were considered to be a reasonable interpretation of the action in the image and
were of equal or lower frequency than the intended target word.
Measures of discourse were collected using a picture description task. Participants
generated a story for 2 minutes in response to a standardised picture (birthday party or
rescue scene; Nicholas & Brookshire, 1993). Relevant content was calculated using the
correct information unit analysis (CIU; Nicholas & Brookshire, 1993). The total number
of words was tallied and compared to the total number of new correct information words
yielding a per cent CIU score. Given the non-fluent, agrammatic profile of this group of
participants, the number of verbs and the number of complete sentences produced in the
narrative were also calculated as a measure of generalisation to discourse.
The American Speech-Language and Hearing Association Functional Assessment of
Communication (ASHA FACS), (Frattali, Thompson, Holland, Wohl, & Ferketic, 1995)

788

E.L. Hoover et al.

and the Assessment of Living with Aphasia (ALA), (Kagan et al., 2010) were administered
once pre- and once post-treatment. The ASHA FACS was chosen to measure language in
functional (social and discourse) contexts. The ALA was chosen for its sensitivity in
measuring the impact of aphasia on an individuals quality of life (Kagan et al., 2010).
This measure poses questions using a four-point rubric on such topics as feelings about
aphasia, ability to communicate in novel and familiar environments and the impact of
aphasia on daily roles.

Inter-rater reliability
Half of all the measures administered during the baselines and all weekly probes were
scored independently by two clinicians. Queries were discussed, and agreement was
reached for all trials.

Results
Treated verbs
Accuracy in producing the treated verbs was analysed in a logistic regression with the
factors of time (week) and treatment type (individual, group, combined). Two linear
splines were introduced to model the onset of treatment (week 5) and the onset of posttreatment (week 12). There was a significant effect of time: there was a non-significant
change in performance in pre-treatment (weeks 14), 2 = 0.010, df = 1, p = .75, a
significant effect during the treatment period (weeks 512), 2 = 8.23, df = 1, p = < .01,
and a significant effect at post-treatment (week 12), 2 = 10.14, df = 1, p = < .001
(Figure 1). These data reveal that accuracy improved during the treatment period and
deteriorated post-treatment. There was no significant effect for treatment type, 2 = 3.64,
df = 2, p = .16, indicating that all treatment types performed equally well. Finally, there
was no interaction of treatment type and time: week 1 treatment, 2 = 4.57, df = 2,
p = .10, week 5 treatment, 2 = 4.09, df = 2, p = .13, week 12 treatment, 2 = 1.25,
df = 2, p = .054.

Figure 1.

Effects of treatment condition over time.

Aphasiology

789

Analysis of linguistic measures

Mean per cent correct scores for the group on all linguistic and narrative measures are
shown in Table 2.
Due to the relatively small sample size in the study, nonparametric test statistics were
used to analyse the data. Language measures were analysed using the Friedman test statistic
to determine whether a difference existed across the repeated measures (all intervals) for the
group over time. The Wilcoxon signed-rank test was subsequently calculated to determine
the interval of change. Figure 2 illustrates the data for each discrete language measure for the
group and a breakdown of performance on the PAL by sentence type.
The Friedman test calculated for each measure across all intervals showed statistically
significant effects (Table 3) for all measures: PNT, VNT, PAL sentence production
(number of correct sentences) and the verb probe. The Wilcoxon test showed stable
baselines, defined as no significant change on pre-treatment intervals (time points 12),
for all measures, indicating stable performance before the treatment programme began. It
also showed statistically significant improvements in scores for all measures from pre- to
post-treatment, demonstrating an improvement of scores after treatment occurred. Stable
scores were observed following treatment (time points 3 to 4) for all measures, indicating
that performance did not change significantly once treatment was removed.

PAL analysis by sentence type

Results from the sentence production subtest from the PAL were analysed at the sentence
level (Table 3). The PAL sentence production test elicits five types of sentences using a
constrained picture description task. Specifically, the examinee is asked to generate a
sentence about a picture, given a verb and target items to include. The sentence-initial
word is cued, which requires the use of a particular grammatical structure. A total of 25
sentences were targeted, five each of the following grammatical structures: passive, active,
dative, dative-passive, and relative clause. A significant treatment effect was seen only for
the dative sentence type (Figure 2). A non-significant degradation of performance at the
month follow-up was also observed.
Table 2.

Descriptive statistics for specific linguistic and narrative measures.

Pre-treatment 1

Pre-treatment 2

Post-treatment

Follow-up

Measure

Mean

SD

Mean

SD

Mean

SD

Mean

SD

PNT
VNT
PAL total score
PAL sentence accuracy
Probe score
ALA
ASHA FACS
Language samples
Per cent CIUs
Number of verbs
Number of complete
sentences

66.8
53
43
13.5
30.2
60.1
5.31*

28.7
23.8
20.9
14
15
15.7
1.1

68.2
54.7
49.8
15.5
31.9

28.4
22.2
17.1
11.3
16.7

83
80.9
60.88
24
64.8
68
5.95*

25
20.9
16.6
14.6
20.9
15.8
0.8

81.6
77.7
61
16.5
60.1

28.7
26.3
15.6
10.8
25.2

42.2
2.8
1.0

35.4
1.8
2.4

50
4.1
1

24.6
1.2
2.9

56
4.7
2.3

26.6
2.3
2.8

56
4.4
2.4

22.8
2.5
3.7

Note: *Mean raw score from seven points possible.

790

Figure 2.

E.L. Hoover et al.

Group performance on discrete linguistic measures.

Analysis of narrative samples

The Friedman test statistic was used to determine whether a change in measures of
narrative production occurred over time for the group. If a statistically significant change
was observed, the Wilcoxon signed-rank statistic was calculated to determine the interval
of change. Table 4 shows the results of the analysis for the narrative measures described
earlier over time.
No significant change was noted in per cent CIU as a result of treatment. A statistically stable baseline before treatment was observed for the mean per cent CIU in the
narrative. An improvement was seen when comparing the mean percentages before and

Aphasiology
Table 3.

791

Results of group means on linguistic measures.

Measure

Interval

Test statistic

Significance

All intervals (4)

Pre-treatment
Pre- to post-treatment
Post-treatment
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment

2
Z
Z
Z
2
Z
Z
Z
2
Z
Z
Z
2
Z
Z
Z

=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=

25.486
0.679
2.866
0.119
30.273
.1.483
3.062
1.069
10.421
0.406
2.375
1.612
31.119
1.616
3.063
.1.584

<.001*
.497
.004*
.905
<.001*
.138
.002*
.285
.015*
.684
.018*
.107
<.001*
.106
.002*
.113

PAL: Analysis by sentence type

Active
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
Passive
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
Dative
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
Dative passive
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
Relative clause
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment

2
Z
Z
Z
2
Z
Z
Z
2
Z
Z
Z
2
Z
Z
Z
2
Z
Z
Z

=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=

2.565
1.134
1.190
0.973
0.636
0.000
1.089
0.816
9.455
1.633
2.251
1.807
3.000
0.000
1.000
1.000
3.000
0.000
1.000
1.000

.464
.257
.234
.330
.888
1.00
.276
.414
.028*
.102
.024*
.071
.392
1.00
.317
.317
.392
1.000
.317
.317

PNT

VNT

PAL sentence production

(number of sentences)

Verb probe

Note: *Statistical significance at or below p = .05.

after treatment although not at a level of significance (Tables 2 and 4). Performance
remained stable for 4 weeks once treatment was withdrawn.
No significant change was observed in the average number of verbs produced over
time. The average number of verbs produced improved for the group between pretreatment baselines, indicating an unstable performance prior to treatment.
A stable baseline was observed for the number of complete sentences produced before
treatment commenced. A significant change was noted for the number of complete
sentences produced following treatment. Figure 3 illustrates the trend in performance
over the four data points for both the number of verbs and the number of complete
sentences produced.

792
Table 4.

E.L. Hoover et al.

Group narrative analyses over time.

Measure
Per cent CIU

Number of verbs produced

Number of complete
sentences produced

Interval

Test Statistic

Significance

All intervals (4)

Pre-treatment
Pre- to post-treatment
Post-treatment
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment
All intervals (4)
Pre-treatment
Pre- to post-treatment
Post-treatment

2 = 9.113
Z = 0.445
Z = 1.690
Z = 0.204
2 = 8.912
Z = 1.911
Z = 0.806
Z = 0.476
2 = 16.986
Z = 1.236
Z = 2.558
Z = 1.000

.028*
.657
.09
.838
.03*
.056
.420
.634
<.001*
.216
.011*
.317

Note: *Statistical significance at or below p = .05.

Figure 3.

Mean number of verbs and complete sentences produced in a narrative.

Analysis of functional communication and quality of life measures

The Wilcoxon signed-rank test statistic was calculated to measure improvements in
functional communication and quality-of-life communication measures as a result of
treatment. Significant improvements were seen on the ALA when comparing pre- to
post-treatment scores, Z = 1.964, p = .05. A significant effect of treatment was also
seen on the ASHA FACS, Z = 2.136, p = .03.
Discussion
Effect of training environment
The primary focus of this study was to understand the effect of the training environment
on verb production. Analysis of the verb probe data revealed that there was no significant
difference between performance on verbs in each treatment condition: all verbs trained
improved regardless of whether they were trained exclusively in individual or group

Aphasiology

793

conditions or in the combined condition. These data do not support the hypothesis that
training verbs in the combined environment would lead to better performance than
training verbs exclusively in one condition.
There were several features of this study that may have affected this result. The group
treatment allowed for dynamic and open dialogues. As a result, topic shifts in conversation were observed frequently, and many verbs that were assigned to individual treatment
were produced in the group sessions. Also verbs, in general, are less specific and offer
more flexibility in contextual use than nouns. For example, the verb choose can be used
in any number of different conversational topics. The flexibility in both the group format
and nature of the stimuli created an unforeseen opportunity for participants to use all the
verbs in the group environment. Therefore, many individual treatment only verbs were
heard and or/practiced in the group treatment sessions. A between-subject design is
needed to address this issue.

Specific language measures

Performance was measured on standardised tests of noun (PNT) and verb naming (VNT), a
sentence production test (PAL) and a verb generalisation probe consisting of 81 photographs
of the trained verbs. Results showed stable scores (no significant change in performance) for
all the measures on the pre-treatment baselines, a statistically significant change pre- to posttreatment and stable scores (no significant change in performance) on the post-treatment
baselines. These results indicate that the treatment was associated with a positive change in
performance on generalised measures of noun naming, verb naming and the number of
accurate sentences produced during a constrained sentence production test. These data also
reveal significant improvements in the use of trained verbs in a different communicative
environment, i.e., participants generalised the use of the trained verbs in response to a written
stimulus to the use of the trained verb in response to a pictured stimulus.

Narrative analyses
Data taken regarding the use of verbs in connected speech yielded mixed results. A
picture description task (Nicholas & Brookshire, 1993) was used to elicit a 2- to 3minute narrative. No statistically significant change was observed for the group in
terms of mean per cent CIUs as a result of the treatment. The number of verbs
produced in each of the narratives was also not significantly different as an effect of
the treatment. A stronger change was seen in verb count between the pre-treatment
baselines compared to the pre- to post-treatment scores (Z = 1.1911, p = .06; Z =
0.806, p = .42, respectively), which suggests that performance on this measure was not
stable. The number of complete sentences produced by the group in the narratives did
demonstrate a significant effect of treatment. Stable scores on pre-treatment measures,
significant change on pre-to post-treatment scores and stable scores on post-treatment
measures were observed for the group. The lack of change in per cent CIUs for the
group indicates that the improved noun and verb retrieval seen on the PNT and VNT
did not generalise further to connected speech. These findings suggest that while the
percentage of the accurate and relevant words produced in the narrative did not change
significantly, participants as a group showed improved use of grammatical rules as a
result of the treatment.

794

E.L. Hoover et al.

Functional and quality-of-life outcomes

Functional and quality-of-life measures were taken at two points during the study: before
and after the treatment. The group demonstrated a significant change pre- to post-treatment on the communicative independence scores on the ASHA FACS. The ALA was
administered to the individuals with aphasia at week 1 (pre-treatment) and at week 12
(post-treatment) to quantify the participants perceptions of aphasia and the impact of
aphasia on their quality of life. The overall score for each participant was used to
determine changes in self-perceptions of communication and the impact of aphasia on
their daily lives. Data revealed a significant change in these scores pre- to post-treatment.
These results indicate that participants identified a lesser impact of aphasia on their quality
of life and that other individuals noticed a change in the participants functional communication as a result of this treatment.

Recovery and time post-onset

Many studies have indicated that the greatest amount of language change post-stroke
occurs in the first 34 months (Culton, 1969; Kertesz & McCabe, 1977; Lendrem &
Lincoln, 1985) and that participants made less progress in therapy after the first year
(Moss & Nicholas, 2006). However, several more recent studies demonstrate that language change can occur following speech-language treatment in participants who were as
much as 5 years post-onset (Fitzpatrick, 1999; Robey, 1998). The participants in this study
were on average 8.75 years post-onset of their stroke (range 1.520, median 9.5), yet the
group demonstrated significant linguistic and psychosocial changes as a result of this
treatment approach. This study lends support to the idea that individuals with longstanding chronic aphasia have the potential for linguistic and psychosocial improvements
following treatment. This programme was also delivered on a typical outpatient delivery
schedule (~2 hours of intervention, twice per week for 6 weeks) and therefore may be
easily applied in current health care settings.

Conclusions
The research question in this study is whether a treatment approach which combined an
individual, impairment-based treatment (VNeST) and a socially oriented group treatment
would result in increased verb production compared to only individual or group treatment
and how the last two treatments would compare. The results showed that verbs trained in
the combined treatment did not improve more than verbs trained in only individual or
group treatments. Rather, verbs trained in all environments improved equally well when
measured on a probe of picture-confrontation naming. These findings lend support to the
earlier work of Edmonds and colleagues (2009, 2011, 2013) that VNeST results in
improvements of verb production of individuals with aphasia. This study also supports
the findings of Elman and Bernstein-Ellis (1999a) that conversational group treatment for
aphasia results in improved language performance. Finally, this outcome extends the
results of Wertz et al. (1981) that group treatment is equally as effective as individual
treatment for verb production in the chronic stages of aphasia and the conclusion that the
cost advantages of group treatment warrant its consideration as part of any treatment plan.
A second research question was whether this treatment approach resulted in linguistic
change for individuals with aphasia. Data for the group revealed strong treatment effects
on measures of lexical retrieval for untrained items in both verb and noun picture naming

Aphasiology

795

tasks. Significant treatment effects were also seen for sentence production on a constrained sentence production task.
A third research question investigated whether a treatment approach that combined
individual plus focused group intervention would result in increased verb production in
connected speech. Data showed that the number of verbs produced in each of the
narratives was not significantly different as an effect of the treatment. Additionally, the
per cent CIU produced over time did not show significant change as an effect of
treatment. These findings together suggest that the changes in lexical retrieval observed
in single-word tasks did not extend to connected speech. A significant effect of treatment
was seen in the number of complete sentences produced by the group in the picture
description narratives. The findings suggest that the treatment led to an effect in grammatical encoding rather than lexical usage in discourse.
A final research question in this study was whether this treatment approach resulted in
functional communication and quality-of-life improvements for the individual with aphasia. Family members observed significant functional linguistic change on the part of the
participants as measured by the ASHA FACS. Data taken from the ALA revealed that
participants perceived a reduced impact from aphasia on their quality of life as a result of
this treatment.
There are several limitations to this study, which should advise caution when interpreting the results. The sample size is small, and despite an attempt to enrol a homogenous participant group, significant variability in performance existed. Additionally, the
within-subject design did not sufficiently control the training environment for each
participant. However, despite these issues, the results of this study demonstrate significant
linguistic, functional and psychosocial changes as a result of an intervention programme
that combined an impairment-based treatment in a collaborative manner with a socially
oriented treatment. This study also demonstrates that individuals with longstanding
chronic profiles of aphasia can benefit from treatment, which is delivered in a nonintensive, cost-effective manner. The results did not demonstrate superior outcomes for
verbs trained in a combined manner in comparison to those trained exclusively in an
individual or group environment, which may be because this study did not control the
environment sufficiently. To better understand the specific benefits of a combined model
of intervention, a randomised controlled study should be undertaken. In such a phase 2
study, the treatment environments would be contained and a larger cohort of participants
should be recruited.

References
Blackstone, S., & Hunt Berg, M. (2003). Social networks: A communication inventory for individuals with complex communication needs and their communication partners. Monterey, CA:
Augmentative Communication.
Borkowski, J. G., Benton, A. L., & Spreen, O. (1967). Word fluency and brain damage.
Neuropsychologia, 5, 135140. doi:10.1016/0028-3932(67)90015-2
Buchner, A., Erdfelder, E., Faul, F., & Lang, A. (2009). G*Power (Version 3.1.2) [Computer
program]. Retrieved from http://www.psycho.unduesseldorf.de/aap/projects/gpower/
Caplan, D., & Bub, D. (1990). Psycholinguistic assessment of aphasia. Mini-seminar presented at
the annual convention of the American Speech-Language-Hearing Association, Seattle, WA.
Cho-Reyes, S., & Thompson, C. K. (2012). Verb and sentence production in aphasia: Northwestern
assessment of verbs and sentences (NAVS). Aphasiology, 23, 13831397.
Culton, G. L. (1969). Spontaneous recovery from aphasia. Journal of Speech Language and
Hearing Research, 12, 825832. doi:10.1044/jshr.1204.825

796

E.L. Hoover et al.

Damico, J., Simmons-Mackie, N., Oelschlaeger, M., & Tetnowski, J. (2000). An investigation of
therapeutic control in aphasia therapy. Poster presented at the Clinical Aphasiology
Conference, Waikaloa, Hawaii.
Edmonds, L. A., & Babb, M. (2011). Effect of verb network strengthening treatment in moderate-tosevere aphasia. American Journal of Speech-Language Pathology, 20, 131145. doi:10.1044/
1058-0360(2011/10-0036)
Edmonds, L. A., Nadeau, S., & Kiran, S. (2009). Effect of verb network strengthening treatment
(VNeST) on lexical retrieval of content words in sentences in persons with aphasia.
Aphasiology, 23, 402424. doi:10.1080/02687030802291339
Edmonds, L. A., Ojeda, J., Mammino, K., & Wu, S. (2013). The effects of verb network strengthening treatment on a group of persons with aphasia: Replication and extension of previous
findings. Seminar presentation Clinical Aphasiology Conference, Tucson, AZ.
Elman, R., & Bernstein-Ellis, E. (1999b). Psychosocial aspects of group communication treatment
preliminary findings. Seminars in Speech and Language, 20, 6572. doi:10.1055/s-20081064009
Elman, R. J. (Ed). (2007). Group treatment of neurogenic communication disorders: The expert
clinicians approach. (2nd ed.). San Diego, CA: Plural Publishing.
Elman, R. J., & Bernstein-Ellis, E. (1999a). The efficacy of group communication treatment in
adults with chronic aphasia. Journal of Speech, Language and Hearing Research, 42, 411419.
doi:10.1044/jslhr.4202.411
EuroQol group. (1990). EuroQoL-a new facility for the measurement of health- related quality of
life. Health Policy, 16, 199208. doi:10.1016/0168-8510(90)90421-9
Fitzpatrick, P. M. (1999). Long-term recovery of naming and word-finding in narrative discourse in
aphasia (Unpublished dissertation). Emerson College Library, Boston, MA.
Frattali, D. M., Thompson, C. K., Holland, A. L., Wohl, C. B., & Ferketic, M. M. (1995). The
American speech-language-hearing association functional assessment of communication skills
for adults. Rockville, MD: American Speech-Language-Hearing Association.
Hawthorne, G. (2006). Measuring social isolation in older adults: Development and initial validation
of the friendship scale. Social Indicators Research, 77, 521548. doi:10.1007/s11205-0057746-y
Horton, S. (2004). Critical reflection in speech and language therapy: Research and practice.
International Journal of Language and Communication Disorders, 39, 486490.
Kagan, A., Simmons-Mackie, N., Rowland, A., Huijbregts, M., Shumway, E., McEwen, S., &
Dickey, L. (2010). Assessment for living with aphasia. Toronto, ON: Aphasia Institute.
Kertesz, A. (2006). The western aphasia battery, revised. San Antonio, TX: PsychCorp.
Kertesz, A., & McCabe, P. (1977). Recovery patterns and prognosis in aphasia. Brain, 100, 118.
doi:10.1093/brain/100.1.1
Kucera, H., & Francis, W. N. (1967). Computational analysis of present-day American English.
Providence, RI: Brown University Press.
Lendrem, W., & Lincoln, N. (1985). Spontaneous recovery of language in patients with aphasia
between 4 and 34 weeks after stroke. Journal of Neurology, Neurosurgery & Psychiatry, 48,
743748. doi:10.1136/jnnp.48.8.743
Lomas, J., Pickard, L., Bester, S., Elbard, H., Finlayson, A., & Zoghaib, C. (1989). The communicative effectiveness index: Development and psychometric evaluation of a functional measure
for adult aphasia. Journal of Speech and Hearing Disorders, 54, 113124. doi:10.1044/
jshd.5401.113
LPAA Project Group. (Chapey, R., Duchan, J., Elman, R., Garcia, L, Kagan, A., Lyon J., &
Simmons-Mackie, N.) (2008). Life participation approach to aphasia: A statement of values
for the future. In R. Chapey (Ed.), Language intervention strategies in aphasia and related
neurogenic communication disorders (5th ed.). Baltimore, MD: Lippincott, Williams & Wilkins.
Martin, M., Thompson, C. K., & Worrall, L. (Eds.). (2008). Aphasia rehabilitation: The impairment
and its consequences. San Diego, CA: Plural Publishing Inc.
Moss, A., & Nicholas, M. (2006). Language rehabilitation in chronic aphasia and time post onset: A
review of single-subject data. Stroke, 37, 30433051.
Nicholas, L. E., & Brookshire, R. H. (1993). A system for quantifying the informativeness and
efficiency of the connected speech of adults with aphasia. Journal of Speech-Language and
Hearing Research, 36, 338350. doi:10.1044/jshr.3602.338

Aphasiology

797

Paul, R., & Cascella, P. W. (2007). Introduction to clinical methods in communication disorders.
Baltimore, MD: Brookes Publishing.
Plaut, D. C. (1996). Relearning after damage in connectionist networks: Toward a theory of
rehabilitation. Brain and Language, 52, 2582. doi:10.1006/brln.1996.0004
Porch, B. E. (1967). Porch index of communicative ability. Palo Alto, CA: Consulting Psychologists
Press.
Robey, R. (1998). A meta-analysis of clinical outcomes in the treatment of aphasia. Journal of
Speech Language and Hearing Research, 41, 172187. doi:10.1044/jslhr.4101.172
Robey, R. R. (1994). The efficacy of treatment for aphasic persons: A meta-analysis. Brain and
Language, 47, 582608. doi:10.1006/brln.1994.1060
Simmons-Mackie, N. (2008). Social approaches to aphasia intervention. In R. Chapey (Ed.),
Language intervention strategies in aphasia and related neurogenic communication disorders
(5th ed.). Baltimore, MD: Lippincott Williams & Wilkins.
Simmons-Mackie, N., Elman, R., Holland, A., & Damico, J. (2007). Management of discourse in
group therapy for aphasia. Topics in Language Disorders, 27, 523. doi:10.1097/00011363200701000-00003
Thompson, C. K. (2002). The Northwestern verb production battery. (Unpublished).
Thompson, C. K., & Shapiro, L. P. (2005). Treating agrammatic aphasia within a linguistic framework: Treatment of underlying forms. Aphasiology, 19, 10211036. doi:10.1080/
02687030544000227
Van Der Gaag, A., Smith, L., Davis, S., Moss, B., Cornelius, V., Laing, S., & Mowles, C. (2005).
Therapy and support services for people with long term stroke and aphasia and their relatives: A
six month follow up study. Clinical Rehabilitation, 19, 372380. doi:10.1191/
0269215505cr785oa
Vickers, C. (2010). Social networks after the onset of aphasia: The impact of aphasia group
attendance. Aphasiology, 24, 902913. doi:10.1080/02687030903438532
Vickers, C., & Threats, T. (2007). Measuring increased life participation associated with attending
an aphasia group. Poster presented at the annual convention of the American Speech-Language
and Hearing Association, Boston, MA.
Walker, G. M., & Schwartz, M. F. (2012). Short-form Philadelphia naming test: Rationale and
empirical evaluation. American Journal of Speech-Language Pathology, 21, S140153.
doi:10.1044/1058-0360(2012/11-0089)
Wertz, R. T., Collins, M. J., Weiss, D. G., Kurtzke, J. F., Friden, T., Brookshire, R. H.,
Resurreccion, E. (1981). Veterans administration cooperative study on aphasia; a comparison
of individual and group treatment. Journal of Speech Language and Hearing Research, 24,
580594. doi:10.1044/jshr.2404.580
Whitworth, A., Webster, J., & Howard, D. (2005). A cognitive neuropsychological approach to
Assessment and Intervention in Aphasia. New York, NY: Psychology Press.

798

E.L. Hoover et al.

Appendix.

Verb categories: Example for subject group 1.

Individual treatment verbs

Dining
Wipe
Choose
Drink
Eat
Order
Pay
leave
Taste
Share
Group treatment verbs
Travel
Miss
Search
Carry
Board
Change
Visit
View
Buy
Remember
Combined treatment verbs
Occupation
Bake
Sell
Count
Save
Teach
Repair
Write
Own
Sort

Hobbies
Cook
Collect
Read
Paint
Draw
Find
Walk
Build
Dance

Sports
Finish
Play
Start
Hit
Pass
Race
Catch
Run
Score

Communication
Meet
Call
Hear
Fight
Show
Marry
Kiss
Answer
Tell

News/events
Vote
Charge
Protest
Lost
Judge
Stop
Free
Arrest
Attack

Household
Break
Clean
Hold
Move
Open
Polish
Slice
Wash
Sweep

Health care
Pack
Prescribe
Push
Check
Test
Weigh
Take
Rest
Scan

Copyright of Aphasiology is the property of Psychology Press (UK) and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.