A Systematic Review of Semantic Feature Analysis Therapy Studies for Aphasia

Abstract

Purpose:
The purpose of this study was to review treatment studies of semantic feature analysis (SFA) for persons with aphasia. The review documents how SFA is used, appraises the quality of the included studies, and evaluates the efficacy of SFA.

Method:
The following electronic databases were systematically searched (last search February 2017): Academic Search Complete, CINAHL Plus, E-journals, Health Policy Reference Centre, MEDLINE, PsycARTICLES, PsycINFO, and SocINDEX. The quality of the included studies was rated. Clinical efficacy was determined by calculating effect sizes (Cohen's d) or percent of nonoverlapping data when d could not be calculated.

Results:
Twenty-one studies were reviewed reporting on 55 persons with aphasia. SFA was used in 6 different types of studies: confrontation naming of nouns, confrontation naming of verbs, connected speech/discourse, group, multilingual, and studies where SFA was compared with other approaches. The quality of included studies was high (Single Case Experimental Design Scale average [range] = 9.55 [8.0-11]). Naming of trained items improved for 45 participants (81.82%). Effect sizes indicated that there was a small treatment effect.

Conclusions:
SFA leads to positive outcomes despite the variability of treatment procedures, dosage, duration, and variations to the traditional SFA protocol. Further research is warranted to examine the efficacy of SFA and generalization effects in larger controlled studies.

Full text

Journal of Speech, Language, and Hearing Research
1
Final Author version
A Systematic Review of Semantic Feature Analysis Therapy Studies for
Aphasia
E. A. Efstratiadou 1,2, I. Papathanasiou 2,3, R. Holland1, A. Archonti2, K. Hilari1
1Division of Language and Communication Science, City, University of London, UK
2Thales Aphasia Project, Department of Linguistics, School of Philosophy, University of
Athens, Greece
3Department of Speech and Language Therapy, TEI of Western Greece Patras, Greece
Corresponding author
Professor Katerina Hilari
Division of Language and Communication Science
School of Health Sciences
City, University of London
Northampton Square,
London
EC1V 0HB
UK
k.hilari@city.ac.uk
Tel: +44 (0) 207 040 4660

Journal of Speech, Language, and Hearing Research
2
Abstract
Purpose: The purpose of this study was to review treatment studies of semantic feature
analysis (SFA) for persons with aphasia. The review documents how SFA is used, appraises the
quality of the included studies and evaluates the efficacy of SFA.
Methods: The following electronic databases were systematically searched (last search
February 2017): Academic Search Complete; CINAHL Plus; E-journals; Health Policy
Reference Centre; MEDLINE; PsycARTICLES; PsycINFO; and SocINDEX. The quality of
the included studies was rated. Clinical efficacy was determined by calculating effect sizes
(Cohen’s d) or percent of non-overlapping data when d could not be calculated.
Results: Twenty-one studies were reviewed reporting on 55 persons with aphasia. SFA was
used in six different types of studies: confrontation naming of nouns, of verbs, connected
speech/discourse, group, multilingual and studies where SFA was compared with other
approaches. The quality of included studies was high [Single Case Experimental Design Scale
(SCEDS) average (range) =9.55 (8.0-11)]. Naming of trained items improved for 45
participants (81.82%). Effect sizes indicated there was a small treatment effect.
Conclusions: SFA leads to positive outcomes despite the variability of treatment procedures,
dosage, duration and variations to the traditional SFA protocol. Further research is warranted to
examine the efficacy of SFA and generalization effects in larger controlled studies.
Key words: Semantic feature analysis, Aphasia, Anomia, Treatment, Systematic review,
efficacy

Journal of Speech, Language, and Hearing Research
3
Introduction
A persistent and frequent symptom of aphasia is anomia, which is a difficulty or
inability to find the right word (Goodglass & Wingfield, 1997). Anomia has been
described as “the most consistent feature of aphasia” as virtually all people with aphasia
experience some degree of word finding problems (Davis, 2000, p. 6). Being unable to find
the right words impairs a person’s ability to express their wants, needs, ideas and feelings,
and participate in everyday conversations and social interactions. Reduced communicative
participation in turn affects the person’s emotional and social well-being and quality of life
(Fotiadou, Northcott, Chatzidaki, & Hilari, 2014; Hilari, Needle, & Harrison, 2012;
Northcott, Moss, Harrison, & Hilari, 2015).
Naming deficits in aphasia are very common. Naming requires processing at the level of
word meaning (semantics), which connects to the word form (phonology) (Dell, Schwartz,
Martin, Saffran, & Gagnon, 1997; Goldrick, 2006; Levelt, 1999; Levelt, Roelofs, & Meyer,
1999;). Impairment in one or both of these processing stages, or the connections between
them, can lead to difficulty in naming (Dell, Lawler, Harris, & Gordon, 2004; Levelt et al.,
1999; Schwartz & Brecher, 2000; Schwartz, Dell, Martin, Gabl, & Sobel, 2006). Therapy
for impaired naming can target semantic or phonological processing or a combination of
these. Therapy approaches have used semantic, phonological and orthographic cues
(Nickels, 2002; Wisenburn & Mahoney, 2009).
Semantic approaches aim to improve naming by restoring or strengthening semantic
representations, or by priming weak semantic representations (Maher & Raymer, 2004).
Semantic tasks described in the literature for improving naming in people with aphasia
include: spoken and written word–picture matching (Byng, 1988; Marshall et al., 1990);
generating semantic features of the object to be named - semantic feature analysis (Boyle,
2004; Boyle & Coelho, 1995; Coelho, McHugh, & Boyle, 2000; Lowell, Beeson, &
Holland, 1995); semantic feature verification (Kiran & Thompson, 2003); generating or
matching synonyms (Hough, 1993); contextual priming (Martin, Fink, & Laine, 2004;
Renvall, Laine, & Martin, 2007); and making judgments about functions, semantic
features, or relatedness of objects (Drew & Thompson, 1999; Nickels & Best, 1996a,
1996b).

Journal of Speech, Language, and Hearing Research
4
Phonological approaches aim to strengthen representations at the level of the word form
(Maher & Raymer, 2004), or strengthen the connections from the semantic system to the
word form (Laine & Martin, 2006). Naming impairment due to deficits in post-
semantic/phonological processing may be the result of impaired access to the phonological
output lexicon, or in the lexical representations themselves (Laine & Martin, 2006).
Phonological tasks include those that provide information about the phonology of the
target (repetition, phonemic cues). Therapy tasks that have been shown to improve naming
in people with aphasia include the use of cueing hierarchies and repetition (Raymer,
Thompson, Jacobs, & Le Grand, 1993); reading aloud (Eales & Pring, 1998; Howard,
1994; Nickels & Best, 1996b), syllable judgments, initial phoneme discrimination, and
rhyme judgment (Franklin, Buerk, & Howard, 2002; Robson, Marshall, Pring, & Chiat,
1998). Repetition is the most common phonological task, used in the majority of
treatments (Nickels & Best, 1996; Nickels, 2002). A subset of phonological approaches
has used orthographic cues, such as providing the first letter of the target word
(Greenwood, Grassly, Hickin, & Best, 2010; Hickin, Best, Herbert, Howard, & Osborne,
2002; Leonard et al., 2004; Lorenz & Nickels, 2007). Lorenz & Nickels (2007) provide a
detailed account of the mechanisms underlying the effectiveness of orthographic cues.
Traditionally, semantic and phonological tasks were thought to have different effects on
word retrieval (Mitchum, Haendiges, & Berndt, 1995; Nickels & Best, 1996a, 1996b).
Early research reported phonological tasks improved naming for a very short time, up to
10-15 minutes, whereas semantic tasks improved naming for up to 24 hours (Howard et al,
1985). However, more recent studies have shown that phonological cues can produce
durable effects (Best & Nickels, 2000). Howard (2000) suggests that the difference
between semantic and phonological tasks may well be overstated. As Howard (1994) and
Nickels (2002a) indicated, most treatments comprise tasks that involve semantic,
phonological, and sometimes orthographic tasks, despite the fact that researchers and
clinicians typically characterize their treatments as either semantic or phonological. In the
majority of the studies using semantic tasks, the form of the word is provided, as a spoken
or written word, and/or repetition is required (suggesting phonological processing), and in
phonological tasks, the picture is usually present (suggesting semantic processing). This is

Journal of Speech, Language, and Hearing Research
5
also the case in Semantic Features Analysis (SFA) (Boyle & Coelho, 1995; Coelho et al.,
2000; Conley & Coelho, 2003).
Ylvisaker and Szekeres (1985) were the first to introduce SFA as an organized method for
facilitating semantic network activation. Ylvisaker, Szekeres and their colleagues provided
general descriptions of SFA treatment, which emphasized the importance of using the
structured procedure consistently. The approach was further developed and tested by
Massaro and Tompkins (1994). They published the first data from a multiple-baseline,
single-subject study of two individuals who had sustained traumatic brain injury.
Theoretically, SFA is based on the concept of spreading activation within the semantic
system (Collins & Loftus, 1975). Specifically, it was proposed that the level of semantic
processing is conceptualized as a network of semantic representations and links to other
related representations. Semantic representations with many shared properties were
thought to link more closely together than representations that had minimal or no shared
properties. The presentation of features that are strongly related to a target results in a
spreading of activation that converges onto the target concept, which thus receives a higher
level of activation than other similar concepts (Boyle, 2004; Boyle & Coelho, 1995;
Coelho et al., 2000; Conley & Coelho, 2003; Haarbauer-Krupa, Moser, Smith, Sullivan, &
Szekeres, 1985a & 1985b; Lowell, Beeson, & Holland, 1995; Massaro & Tompkins,
1994). For example, Boyle (2010) uses the example of “apple”. Its semantic features
include <fruit >, <has a core>, <has skin>, <has seeds>, <grows on trees>, and <used for
cider>. The information provided by its features differs, with some features providing more
distinctive information (distinctive features) than others (common features). The feature
<used for cider> distinguishes it from other fruits, like orange, whereas <has skin> does
not distinguish it because most fruits have skin. The target concept then activates the
phonological information associated with it, resulting in the production of the target word.
SFA, therefore, relies upon re–learning, or applying a strategy you have learnt, which
encourages activation between strongly associated features that in turn drives naming of a
target picture or semantic concept (Hashimoto & Frome, 2011).
The SFA treatment protocol involves employing a “feature analysis chart” that includes the
following semantic features for object naming: group, action, use, location, properties, and
associations (Boyle, 2010) and for action naming: subject, purpose of action, part of body
or tool used to carry out the action, description, usual location and associated objects or

Journal of Speech, Language, and Hearing Research
6
actions (Wambaugh & Ferguson, 2007) (see Figure 1). During SFA treatment, individuals
with word retrieval difficulties are shown a picture to name and they are encouraged to
generate the semantic features of the target word by completing the feature analysis chart.
The completion of the feature analysis chart is achieved by systematic cueing techniques,
like asking questions or using sentence completion. For example, for ‘rabbit’, ‘It is
an…’(‘animal’), whilst pointing to the picture, ‘It has ….’ (‘long ears / fluffy tail’), ‘What
does it do?’ (‘It hops’). The clinician guides the person with aphasia to complete the chart
and gradually cueing is faded so that the person with aphasia becomes increasingly
independent in generating features. It is argued that generation of such semantic features
works as a compensatory strategy to enhance activation of the target word via the
processing of shared features, which enables the individual to find the target word.
Persistent and systematic practice in producing semantic features in this way enables
individuals to achieve more organized word retrieval without the deliberate use of
compensatory strategies (Boyle, 2010).
[figure 1 about here]
Two reviews have been previously conducted on SFA treatment. Boyle’s (2010) report
was the first and examined the efficacy of SFA. The review comprised seven studies where
SFA was used for confrontation naming of nouns. Results were reported for 17 participants
with aphasia, 16 of whom improved their ability to name pictured nouns. These
participants had a variety of classic fluent and non-fluent aphasia syndromes. The review
concluded that SFA treatments improve naming of treated items for most participants,
regardless of whether they require participants to generate the features themselves or
whether participants analyze features that have been generated by others (Boyle, 2010).
Maddy, Capilouto and McComas (2014) conducted a systematic review on the same area,
but excluded studies that involved verification rather than generation of features (Edmonds
& Kiran, 2006; Kiran & Roberts, 2010). The review comprised 11 studies with 24
participants with aphasia. Seventeen of them had non-fluent aphasia and seven participants
had fluent aphasia. Cohen’s d was calculated and the majority of participants showed a
small effect size. The percent of non-overlapping data was also calculated and a large
treatment effect was present for the majority of participants. The review concluded that
SFA is an effective intervention for improving confrontational naming of items trained in
therapy; however, limited generalization to untrained items and connected speech was

Journal of Speech, Language, and Hearing Research
7
reported in the majority of the included studies. The present study extends the previous
reviews (Boyle, 2010; Maddy et al., 2014) in a number of ways. It includes new research.
It evaluates the methodological quality of the existing studies against standard criteria (The
Single Case Experimental Design Scale (SCEDS) critical appraisal tool (Tate et al., 2008)
and level of evidence, based on the Scottish Intercollegiate Guidelines Network
(http:sign.ac.uk/pdf/sign118.pdf, 2010). It also broadens the scope of the previous reviews
by documenting the characteristics of SFA studies (participant characteristics, type of SFA,
treatment dosage, treatment duration, total amount of treatment); and determining clinical
efficacy. In particular, the following research questions were addressed:
1) What is the methodological quality of studies evaluating the efficacy of SFA in aphasia therapy?
This will be rated against standard criteria.
2) What are the characteristics of SFA aphasia therapy studies, in terms of i) type, dosage, duration
and total amount of treatment, and ii) participant characteristics?
3) What are the results of SFA aphasia therapy studies, in terms of i) treatment outcomes, and ii)
clinical efficacy as determined by effect sizes using Cohen’s d or percent of non-overlapping
data?
Methods
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines
(Moher, Liberati, Tetzlaff, & Altman, 2009 & 2010) formed the basis of the conduct and reporting of
this systematic review. PRISMA stems from an international collaboration formed to update the
QUOROM Statement (QUality Of Reporting Of Meta-analyses). PRISMA provide an accepted,
evidence-based minimum set of items for reporting in systematic reviews, which have been updated
to address several conceptual and practical advances in the science of systematic reviews.
Search Strategy and Eligibility Criteria
A systematic search of the literature was conducted to identify studies that investigated SFA as a
primary intervention method for people with aphasia. Electronic searches of the following databases
were conducted, with the last search in November 2017, using the EBSCOHOST platform: Academic
Search Complete, CINAHL Plus with Full Text, E-Journals, MEDLINE with Full Text, PsycINFO,
ERIC and the Aphasia Treatment website of the Academy of Neurologic Communication Disorders
(http://aphasiatx.arizona.edu/).

Journal of Speech, Language, and Hearing Research
8
The search strategy comprised the following terms:
1. Semantic feature analysis
2. Semantic cues
3. 1 or 2
4. Aphasia
5. Dysphasia
6. 4 or 5
7. Naming
8. Word finding difficult*
9. 7 or 8
10. 6 or 9
11. Therap*
12. Treat,
13. Intervention
14. 11 or 12 or 13
15. 3 and 10 and 14.
After removal of duplicate studies, material resulting from the searches was screened against the
eligibility criteria. Studies were considered eligible if they were research reports and were published
in English. Studies that combined SFA with other treatment approaches were excluded, when it was
impossible to delineate specifically the effects of SFA. Where eligibility could not be assessed on the
basis of the title and abstract alone, the full text was obtained.
Study selection: Screening and data extraction
We found 357 abstracts that mentioned Semantic Feature Analysis (SFA) in their abstract and 1500
abstracts that mentioned “semantic cues”. Of these, 145 136 were relevant to aphasia / dysphasia and
130 addressed “naming” and / or “word finding difficult*”. Of these, 54 were considered for this
review as they also mentioned therapy / treatment / intervention. The full text was obtained for these
54 articles. Of these, seven were excluded as they used different therapy methods, like cueing
hierarchy approach (Linebaugh, Shisler, & Lehner, 2005), multi cue computer program (Doesborgh
et al., 2004; van Mourik, Verschaeve, Boon, Paquier, & van Harskamp, 1992), personal cueing in

Journal of Speech, Language, and Hearing Research
9
natural settings (Olsen, Freed, & Marshall, 2012), phonological components analysis (PCA)
(Leonard, Rochon, & Laird, 2008), orthographic cueing (Leonard, Rochon, & Laird, 2004) and a
different semantic approach which compared a phonological and orthographic approach (Lorenz &
Ziegler, 2009). One study was excluded as it evaluated SFA in participants with primary progressive
aphasia and Alzheimer’s Disease (Hung et al., 2017). In this review, only studies that used SFA
treatment with semantic feature generation have been included. Reports on treatment studies
involving semantic feature review or verification have been excluded. Thus 15 articles were
excluded, as they reported on a different semantic features approach, such as semantic feature
verification rather than generation, or combined SFA with other treatment approaches in the same
therapy protocol, such as response elaboration training (RET), communication based therapy,
semantic priming, semantic judgment tasks, auditory concept feature and gesturing treatment
(Antonucci, 2014a; Boo & Rose, 2011; Cameron, Wambaugh, Wright, & Nessler, 2006; Carragher,
Conroy, Sage, & Wilkinson, 2012; Conley & Coelho, 2003; Edmonds & Kiran, 2006; Hashimoto,
2016; Kintz, Wright, & Fergadiotis, 2016; Kiran & Roberts, 2010; Knoph, Simonsen & Lind, 2017;
Law, Wong, Sung, & Hon, 2006; Lowell, Beeson, & Holland, 1995; Raymer, Rodriguez, & Rothi,
2007; Wallace & Kimelman, 2013; Wambaugh, Mauszycki, Cameron, Wright, & Nessler, 2013).
Moreover, one study was excluded as it evaluated treatment integrity of elaborated SFA
(Kladouchou, Papathanasiou, Efstratiadou, Christaki & Hilari, 2017). An additional seven studies
were excluded, as they were not research reports (Antonucci, 2014b; Bose & Buchman, 2007; Boyle,
2010; Durand & Ansaldo, 2014; Kiran & Bassetto, 2008; Maddy et al., 2014; van Hees, Mcmahon,
Angwin, De Zubicaray, & Copland, 2014a). Lastly, two studies were excluded because they were
not relevant to naming, instead one was treating oral reading (Kiran & Viswanathan, 2008), and the
other comprehension SFA (Munro & Siyambalapitiy, 2016). The remaining 21 articles were included
in the review. The selection process of the articles is illustrated in figure 2.
The 21 studies covered six main areas: confrontation naming of nouns studies, confrontation naming
of verbs studies, studies covering both nouns and verbs, connected speech – discourse studies of
which two were group studies, multilingual study, and studies where SFA was compared with other
approaches, like Phonological Components Analysis (PCA) (Hashimoto, 2012; Neumann, 2017;
Sadeghi, Baharloei, Zadeh, & Ghasisin, 2017; van Hees, Angwin, McMahon, & Copland, 2013).
[figure 2 about here]

Journal of Speech, Language, and Hearing Research
10
Critical Appraisal and Methodological Quality
We appraised the methodological quality of included studies and assigned levels of
evidence as an indication of risk of bias. Two aphasia-specialist speech-language
pathologists critically evaluated the included studies for their methodological quality. All
studies were single case studies (N=21). The Single Case Experimental Design Scale
(SCEDS) critical appraisal tool (Tate et al., 2008) was used to examine the quality of the
studies. SCEDS is an 11-point scale evaluating the methodological quality of experimental
single case studies. A perfectly designed and executed study would receive a summative
score of 11 across eleven different criteria. A score of 1, per criterion, is given if the study
adequately addresses the specified quality item and a score of 0 is given if the item is
poorly addressed or not addressed at all. The eleven specified quality items are: (i) clinical
history, (ii) target behaviors, (iii) design, (iv) baseline, (v) sampling behavior during
treatment, (vi) raw data record, (vii) inter-rater reliability, (viii) independence of assessors,
(ix) statistical analysis, (x) replication and (xi) generalization. All included studies were
evaluated with SCEDS by two raters. When disagreements between raters were present, an
average score was calculated. The first author randomly selected six studies (29%) and re-
calculated SCEDS scores to determine intra-rater reliability. Intra-rater reliability was
ICC=1.0 (100% agreement). To reduce bias and ensure ratings were not dependent upon
one another, re-scoring was completed two weeks after the initial scoring.
Level of evidence was also assigned to each of the studies. Level of evidence refers to the
hierarchy of study designs based on the ability of the design to protect against bias. While
there is no one universally accepted hierarchy, randomized controlled trials (RCTs) are
considered to be the design least susceptible to bias, and various hierarchies follow from
there through observational studies and non – experimental designs. We followed the
Scottish Intercollegiate Guidelines Network (2010) hierarchy, where RCTs, systematic
reviews of RCTs and meta-analyses are considered level 1 evidence; case control and
cohort studies and systematic reviews of these are considered level 2 evidence; non-
analytic studies, such as case reports and case series are considered level 3 evidence; and
expert opinion is considered level 4 evidence. Full information on this classification
system is available on http://www.sign.ac.uk/pdf/sign118.pdf.

Journal of Speech, Language, and Hearing Research
11
Phase of treatment was also considered for each study, using the coding of Robey and
Schultz (1998 & 2004), which is a five – phase model: Phase 1 studies are pre – efficacy
studies, where the goal is to determine if there is evidence to suggest that the treatment has
therapeutic value. Phase 2 are pre- efficacy studies, where the goal is to develop,
standardize, validate, and optimize procedures to explain why a therapy works and who are
the ideal candidates. Phase 3 are efficacy studies, where treatment is tested for efficacy
under ideal conditions. Phase 4 are effectiveness studies, where treatment is tested for
effectiveness under ordinary conditions of use. Lastly, phase 5 are effectiveness studies
exploring efficiency, cost-benefit, and patient reported outcomes such as satisfaction and
quality of life.
Treatment outcomes and clinical efficacy
As well as describing the treatment outcomes of included studies, the clinical efficacy of
SFA was determined by calculating effect sizes. Effect sizes could be calculated only in
those studies that reported sufficient data. To calculate, it was necessary to determine the
individual values for the pre- treatment and post-treatment phases for each set of trained
items. Cohen’s d statistic was used to calculate effect size as described by Busk and Serlin
(1992). The magnitude of change in performance was determined according to the
benchmarks for lexical retrieval studies described by Beeson and Robey (2006). The
benchmarks were 4.0, 7.0, and 10.1 for small, medium, and large effect sizes respectively.
Where Cohen’s d could not be calculated, the percent of non-overlapping data (PND) was
calculated. PND is the most widely used method of calculating effect size in single case
experimental designs (Gast, 2010; Schlosser, Lee, & Wendt, 2008). PND is the percentage
of phase B data points (the treatment phase) that do not overlap with phase A data points
(baseline or no treatment). To determine the magnitude of effect, benchmarks put forth by
Scruggs et al. (1987) were used. PND scores higher than 90% were considered to
demonstrate a highly effective treatment, PND of 70–90% were interpreted as a moderate
treatment outcome and PND scores of 50–70% were considered a questionable effect.
PND scores less than 50% were interpreted as an ineffective intervention since
performance during intervention had not affected behavior beyond baseline performance.

Journal of Speech, Language, and Hearing Research
12
Results
Study selection
Twenty-one studies were included in this systematic review. The studies cover six
different research areas. Nine studies investigated SFA with confrontation naming of
nouns (Boyle, 2004; Boyle & Coelho 1995; Coelho, McHugh & Boyle, 2000; Davis &
Stanton, 2005; DeLong, Nessler, Wright, & Wambaugh, 2015; Hashimoto & Frome,
2011; Massaro & Tompkins, 1994; Mehta & Isaki, 2016; Rider, Wright, Marshall, &
Page, 2008). Two studies examined SFA with confrontation naming of verbs
(Wambaugh & Ferguson, 2007; Wambaugh, Mauszycki, &Wright, 2014) and a further
two tested SFA with confrontation naming of nouns and verbs (Kristensson, Behrns, &
Saldert, 2015; Marcotte & Ansaldo, 2010). Kristensson’s study additionally explored
everyday conversation and functional communication outcomes. Connected speech –
discourse - was examined in one study (Peach & Reuter, 2010), group SFA was
evaluated in two studies (Antonucci, 2009; Falconer & Antonucci, 2012), and
multilingual SFA was tested in one study (Knoph, Lind, & Simonsen, 2015). Finally,
four studies compared SFA with other approaches, like Phonological Components
Analysis (PCA) (Hashimoto, 2012; Neumann, 2017; Sadeghi, Baharloei, Zadeh, &
Ghasisin, 2017; van Hees et al., 2013). Before presenting the characteristics and details
of the above studies their methodological quality will be considered.
Critical Appraisal and Methodological Quality
Across the 21 studies, scores on the SCEDS ranged from 8.0 to 11 with an average score of
9.55 out of 11 (Table 1). After SCEDS scoring, level of evidence was assigned for the
studies. All studies were determined to be well – designed non – experimental / non –
analytic studies and assigned a level III rating, except of Marcotte and Ansaldo (2010),
which was classified as an observational controlled study.
[table 1 about here]
Phase of treatment was obtained for all studies. Chronologically earlier studies, from 1994
until 2007 and Hashimoto’s and Frome’s study (2011), were Phase 1 studies (see Table 1),

Journal of Speech, Language, and Hearing Research
13
i.e., pre–efficacy studies (n=11), where the goal was to determine if there was evidence to
suggest that the treatment had therapeutic value. All other studies, except for Rider et al.,
(2008) were Phase 2 pre-efficacy studies (n=9), where the goal was to develop,
standardize, validate, and optimize procedures to explain why SFA worked and who were
the ideal candidates. Rider and colleagues’ study (2008) was a Phase 3 efficacy study,
where treatment was tested for efficacy under ideal conditions. The prevalence of high
SCEDS scores suggests the included studies were of good/adequate methodological
quality, despite being pre-efficacy studies.
Characteristics of studies:
Type and duration of treatment
Study and participant characteristics are shown in Tables 2 and 3. Table 2 details the
number of participants, type of SFA treatment, dosage and duration of treatment and total
amount of treatment expressed in minutes. A total of 55 participants have been treated in
the included studies. Total amount of treatment ranged from 315 minutes (Sadeghi et al.,
2017) to 1500 minutes (Boyle, 2004) [mean (SD) = 1019.69 (337.17)].
SFA of nouns
Nine studies, with a total of 18 monolingual individuals, tested SFA in confrontation
naming tasks of single nouns (Boyle, 2004; Boyle & Coelho 1995; Coelho et al., 2000;
Davis & Stanton, 2005; DeLong et al., 2015; Hashimoto & Frome, 2011; Massaro &
Tompkins, 1994; Mehta & Isaki, 2016; Rider et al., 2008). Treatment duration ranged
from five to 12 weeks and treatment was delivered in two to three 60 minute sessions per
week, with a total amount of treatment of 12 to 24 hours [mean (SD)= 18 (4.38)].
SFA of verbs
Two studies, with five monolingual participants, applied SFA in confrontation naming
tasks that targeted single verbs (Wambaugh & Ferguson, 2007; Wambaugh et al., 2014).
The treatment duration was four weeks and treatment was delivered in three 45 - 60
minutes’ sessions per week.
SFA of nouns and verbs
Two SFA studies combined confrontation naming tasks of single nouns and verbs
(Kristensson et al., 2015; Marcotte & Ansaldo, 2010). In Marcotte and Ansaldo’s (2010)
study the treatment duration for the individual was three weeks and he had three 60
minutes’ sessions per week resulting in nine hours of therapy in total. In Kristensson and

Journal of Speech, Language, and Hearing Research
14
colleagues’ (2015) study the three participants received 20 hours of treatment delivered in
20 sessions lasting 60 minutes each for a period of five to six weeks.
Discourse SFA
Discourse SFA was evaluated in three studies, one using an individual approach (Peach &
Reuter, 2010) and two using a group approach (Antonucci, 2009; Falconer & Antonucci,
2012). Individual discourse SFA was evaluated with two participants, one monolingual
and one bilingual (Peach & Reuter, 2010). Treatment was delivered in 50 minutes’
sessions and lasted ten weeks, with a total amount of treatment of 11-12 hours. Group
approach SFA was tested in two studies (Antonucci, 2009; Falconer & Antonucci, 2012),
with seven monolingual participants, for seven weeks, with a small difference on the
amount of hours in each study. In Antonucci (2009) each session ranged from 60 to 90
minutes and in Falconer and Antonucci (2012) from 90 - 120 minutes, resulting in a total
amount of treatment of 1050 - 1470 minutes [mean (SD) = 1260 (296.98)].
Multilingual SFA
Multilingual SFA was tested in one study (Knoph, Lind, & Simonsen, 2015), with one
quadrilingual participant, for two and a half weeks, each session ranged from 45 to 55
minutes, resulting in a total amount of 1320 minutes.
Comparing SFA to PCA
Four studies compared SFA with PCA (Hashimoto & Frome, 2012; Neumann, 2017;
Sadeghi et al., 2017; van Hees et al., 2013) in a total of 18 participants. In the Hashimoto
& Frome (2012) study, two participants were seen twice weekly and had two 45-60
minute sessions on each of these two days for 15 to 25 weeks. In the van Hees et al (2013)
study, eight participants received three 45-90 minute sessions per week for four weeks. In
the Neumann (2017) study, four participants were seen two to three times per week for a
two-hour session, the duration of treatment varied from two to six and a half weeks. In
Sadeghi et al (2017), four participants received seven 45 minute sessions for two weeks.
Participant characteristics
Table 3 presents the demographic characteristics of the 55 participants from the 21
reviewed studies. Considerable heterogeneity was found across the participants in terms
of age and time post onset. Age ranged from 24 to 80 years, with a mean (SD) age of
55.39 (12.66). Time post onset ranged from 4 to 384 months, with a mean (SD) of 59.75

Journal of Speech, Language, and Hearing Research
15
(68.70) months. Thirty-one participants were men and 24 were women. Of the
participants, 21 were described as non–fluent and 33 as fluent (one was not reported).
Aphasia was due to a stroke in 51 individuals and to traumatic brain injury in four
individuals (neuropathology for three individuals was not reported). Aphasia severity was
reported or derived from the aphasia quotient (AQ) of the WAB in 14 studies. Four
studies based aphasia severity on a different test and three did not report severity. One
participant presented with very severe aphasia, four with severe, three with moderate to
severe, 23 with moderate, three with mild to moderate, and 14 with mild aphasia. Aphasia
type was not reported for six participants. Of the remaining, 14 had Broca’s aphasia, 16
anomic, five Wernicke’s, nine conduction, one global, one mixed and three transcortical
motor aphasia.
[table 2 about here]
[table 3 about here]
Synthesis of results
Treatment outcomes
The main treatment outcomes of the reviewed studies are summarized in Table 4.
Improvement in naming of trained items was found for 45 participants (81.82%).
Maintenance of naming of the trained items was reported for 32 participants
(58.18%). Generalization effects ranged from negligible (e.g., Rider et al., 2008) to
strong (Boyle, 2004). The percentage of generalization to untrained items for all
studies was small (40%).
[table 4 about here]
In relation to aphasia type and the outcome of SFA therapy, we looked firstly at
improvement on the trained items. Twelve of the 14 (85.71%) participants with Broca’s
aphasia, 13 of the 16 anomic participants (81.25 %), four of the five (80%) individuals with
Wernicke’s aphasia, and all nine with conduction aphasia and three with transcortical motor
aphasia (100%) showed improvement on naming of trained items. Negative outcomes were

Journal of Speech, Language, and Hearing Research
16
found for the two participants with global and mixed aphasia. In terms of maintenance, the
findings were positive for eight (50%) of the anomic participants, seven (50%) participants
with Broca’s aphasia and all those with conduction and transcortical motor aphasia (100%),
whereas only two (40%) of participants with Wernicke’s aphasia, and none of the two
individuals with global or mixed aphasia showed a maintenance effect. In terms of
generalization to untreated items, it was mostly the individuals with Broca’s aphasia that
showed positive gains (57.14 %). All other aphasia type participants showed minimal gains
on generalization to untreated items. Specifically, gains were reported for 33.33 % of the
participants with conduction aphasia, 25% of Wernicke’s aphasia, 16.67 % of those with
anomic aphasia and 37.5% of the individuals with transcortical motor aphasia.
All studies assessed post - therapy gains immediately after treatment ended. The number
of assessments and the timing of follow-up assessments varied (table 5). Overall, three
studies assessed gains only once post-therapy (Knoph, Lind, & Simonsen, 2015; Marcotte
& Ansaldo, 2010; Sadeghi et al., 2017) and 18 included follow-up/maintenance
assessments. The majority of the studies (n=12) assessed maintenance up to six weeks
after the end of treatment. Five studies assessed maintenance up to 2-4.5 months after the
end of treatment (Boyle & Coelho, 1995; Coelho et al., 2000; Kristensson et al., 2015;
Mehta & Isaki, 2016; Peach & Reuter, 2010). Only one study went beyond 4.5 months
and had multiple follow-up assessments up to a year (Davis and Stanton, 2005).
[table 5 about here]
Clinical efficacy
Effect sizes for treatment outcomes were reported in eleven studies (Antonucci, 2009;
DeLong et al.,2015; Falconer & Antonucci, 2012; Hashimoto & Frome, 2011; Hashimoto,
2012; Knoph et al., 2015; Kristensson et al., 2015; Peach & Reuter, 2010; Rider et al.,2008;
van Hees et al.,2013; Wambaugh et al., 2014;). Calculation could not be performed for three
studies (Davis & Stanton, 2005; Marcotte & Ansaldo, 2010; Mehta & Isaki, 2016).
The first author of the review calculated effect sizes for two studies (n = 2) (Boyle, 2004;
Wambaugh & Ferguson; 2007), as well as average effect sizes for eight studies (n = 24)
(Antonucci, 2009; DeLong et al., 2015; Hashimoto & Frome, 2011; Kristensson et al., 2015;

Journal of Speech, Language, and Hearing Research
17
Neumann, 2017; Rider et al., 2008; Sadeghi et al., 2017; Wambaugh et al., 2014) (Table 6).
Average effect sizes were calculated when data were collected and reported on two or more
trials at one-time point. Further effect sizes were calculated in three studies (n = 4) (Boyle &
Coelho, 1995; Coelho et al., 2000; Massaro & Tompkins, 1994) based on substitute data from
other phases, following the recommendation of Beeson and Robey (2006). Large effect sizes
were present for eight participants (d = 10.07 – 19.23). Medium effect sizes were present for
six participants (d = 7.00 – 9.58). Small effect sizes were present for eleven participants (d =
4.14 – 6.89). For 17 participants, effect sizes were negligible and for five there was no
change. Effect size and PND could not be calculated for five participants from the studies of
Marcotte and Ansaldo (2010), Mehta and Isaki (2016) and one participant from DeLong et
al. (2015) and Antonucci (2009) studies.
PND was calculated for three studies (Boyle, 2004; Davis & Stanton, 2005; Peach & Reuter,
2010), for three participants for whom effect sizes could not be calculated. A large treatment
effect (PND > 90%) was evident for two participants and a moderate treatment effect for one
participant (PND = 85%). When examining clinical efficacy using PND, treatment was
highly effective for the majority of participants. None of the participants had PND scores
consistent with ineffective treatment.
[table 6 about here]
Discussion
The purpose of this review was to evaluate the quality of SFA therapy studies in aphasia;
detail their characteristics and synthesize their results. We reviewed 21 studies reporting on
55 persons with aphasia. Improvement in naming of trained items was found for 45
participants (81.82%). Thus, SFA improved treated items for the majority of participants.
Yet, effect size calculations indicated that there was a small or less than small treatment
effect for a substantial proportion of participants (28/45, 62.22%). Moreover, although
findings suggest that treatment was effective for improving naming of trained items, limited
generalization to untrained items and connected speech was reported (40%).

Journal of Speech, Language, and Hearing Research
18
Maintenance of the trained items post therapy was reported for 32 participants (58.18%).
Maintenance of therapy gains can be affected by factors like the timing of assessment,
treatment dosage and duration (Boyle, 2010). Timing of assessment for maintenance effects
varied (see Table 5). This variation may affect results, as when the evaluation is closer to the
end of the intervention, maintenance of gains is more likely than when maintenance is
assessed after a longer period. Looking at short-term maintenance, from the 21 studies,
short - term post – therapy gains (two weeks) were reported in only five studies (DeLong et
al., 2015; Massaro & Tompkins, 1994; Wambaugh et al., 2014; Wambaugh & Ferguson;
2007; van Hees et al., 2013). Eleven of the 20 participants (55%) in these studies showed a
maintenance effect. If we consider longer-term post – therapy gains, six studies looked at
two months or more post therapy, with 5 of 10 participants (50 %) showing maintenance of
treatment gains (Boyle & Coelho, 1995; Coelho et al., 2000; Davis & Stanton, 2005;
Kristensson et al., 2015; Mehta & Isaki, 2016; Peach & Reuter, 2010). Though the results
seem to confirm that closer to the end of therapy gains are more likely to be maintained, we
need to interpret this with caution as the number of participants assessed in the longer term
(≥ 2 months) is small.
Results of generalization to untreated items ranged from strong (e.g., Boyle, 2004) to
negligible (e.g., Rider et al., 2008; Wambaugh et al., 2014; Wambaugh & Ferguson, 2007).
Positive generalization outcomes were evident for 40% of participants. It is argued that
generalization may be related to the underlying mechanism of how SFA works. That is, if
SFA has a semantic network repair function, then untreated items that belong to the same
semantic category as trained items will indirectly benefit from treatment. Items that lie
outside of the semantic network would not be likely to benefit. However, if SFA functions as
a self-employed ‘‘semantic cueing strategy’’, as Lowell and colleagues (1995) suggested, it
would be expected that semantically related and unrelated items would improve when the
strategy is implemented successfully. In this review, it has not been possible to evaluate this
hypothesis as limited information was provided in most studies on the nature of
generalization. However, Boyle (2004) performed a post hoc analysis of categorical
membership of treated and untreated experimental stimuli and found that generalization
occurred to untreated items that were not members of the same categories as treated items.
Generalization to unrelated items suggested that SFA functioned as a mediating strategy for
naming those items.

Journal of Speech, Language, and Hearing Research
19
One study reported on a multilingual participant (Knoph et al., 2015) and found naming
improvement in the untreated languages. Similar findings have been reported in prior studies
where semantic feature verification has been used with bilingual speakers (Edmonds &
Kiran, 2006; Kiran & Roberts, 2010), with cross-linguistic transfer in some conditions for
some participants. It has been suggested that cross-linguistic transfer is difficult to achieve
(Ansaldo & Ghazi Saidi, 2014; and Faroqi - Shah et al., 2010). Yet, Knoph and colleagues
(2015) hypothesized that the semantic nature of SFA therapy would lead to cross -linguistic
transfer, and their results partly supported their hypothesis.
Although all studies focus on treating word finding difficulties in aphasia, pulling their
results together is challenging due to the expected heterogeneity of various study
components. A variety of aphasia types has been evaluated. Individuals with Broca’s,
Wernicke’s, anomic, conduction, global, and transcortical motor aphasia syndromes have
been included. Dividing participants to the broad categories of fluent and non – fluent
aphasia, people with fluent aphasia are the most represented subtype in the reviewed studies
(33/55, 60%). In terms of aphasia severity, the main body of the participants (72.73%) had
mild (n=14), mild-moderate (n=3), or moderate (n=23) aphasia. Overall, results suggested
that SFA as a treatment for word finding difficulties may be more effective for persons with
fluent and moderate or mild aphasia (Antonucci, 2009; Boyle, 2004; Coelho et al., 2000;
Hashimoto, 2012) compared to those with non – fluent and more severe aphasia (Hashimoto
& Frome, 2011; Kristensson et al., 2015). However, Boyle (2010) in a review of SFA
treatments for nouns found that participants with severe aphasia also had positive responses.
Lowell et al. (1995) suggested that aphasia severity and poor non-verbal cognitive skills
were determining factors for participants who did not show improvement post therapy.
Wambaugh and colleagues (2013) also suggested that different profiles of language,
memory, and cognition might be associated with different responses to SFA. Further
research with large numbers of participants is necessary in order to begin to unravel the
impact of different aphasic profiles and severities on the efficacy of SFA.
Another important consideration is that treatments, which are called SFA, are not always the
same in terms of their treatment protocols. Many studies changed the traditional SFA
protocol in various ways, such as modifications to the semantic feature categories (Mehta &

Journal of Speech, Language, and Hearing Research
20
Isaki, 2016; Wambaugh et al., 2014; Wambaugh & Ferguson, 2007), eliciting fewer features
(Hashimoto & Frome, 2011; Mehta & Isaki, 2016), writing the features in addition to or
instead of saying them (Hashimoto & Frome, 2011), following different treatment stages
(Davis & Stanton, 2005), and adding new factors, such as independent homework (Falconer
& Antonucci, 2012). This variability again makes it difficult to determine which aspects of
SFA were most effective.
Different treatment outcomes could also be due to different treatment durations, dosages and
total amount of treatment. Therefore, another limiting factor is the lack of a standardized
dosage and treatment duration across studies. Some studies, like Hashimoto and Frome
(2011) reported longer treatment sessions over a shorter duration. Across the studies
reviewed, duration of treatment varied from two weeks to twelve weeks [mean (SD) = 5.92
(2.56)]. Treatment sessions per week also varied from two to four sessions [mean (SD) =
2.64 (0.59)], and duration of sessions varied from 45 minutes to 90-120 minutes [mean (SD)
= 63.28 (18.42)]. The most common duration per session was one hour (identified in eight
different studies). It may be that total amount of treatment may relate to treatment outcomes.
The findings of this review partly support this finding. There were eight studies with low
amount of treatment, i.e. 315-720 minutes (Davis & Stanton, 2005; Marcotte & Ansaldo,
2010; Mehta & Isaki, 2016; Peach & Reuter, 2010; Rider et al., 2008; Sadeghi et al., 2017;
Wambaugh et al., 2007 & 2014). Eighteen of the 19 participants in these studies made gains
in naming post-therapy, nine of the 19 maintained these gains and seven generalized to
untreated items. In the six studies with high overall treatment amount (1260-1470 minutes),
11 of 11 participants made gains post-therapy, and 9 of 10 maintained these gains and
generalized to untreated items (Boyle, 2004; Coelho et al., 2000; Falconer & Antonucci,
2012; Hashimoto & Frome, 2011; Hashimoto, 2012).
Despite the complicating factors of variability of treatment procedures, dosage, duration and
changes to the traditional SFA protocol, this systematic review of SFA studies suggests that
SFA is an effective intervention that can elicit positive therapy outcomes. Synthesizing the
findings of 21 single case and case series studies suggests that SFA is effective in improving
treated items and has a small effect on generalization to untrained items. In summary, the
evidence-base for SFA as a therapeutic intervention is growing, but further research with
larger numbers of participants is warranted to examine differential gains across aphasia types

Journal of Speech, Language, and Hearing Research
21
and explore generalization to untreated items and longer term maintenance with greater
confidence.

Journal of Speech, Language, and Hearing Research
22
Acknowledgments
This research was supported by a doctoral studentship to the first author by City, University
of London School of Health Sciences.
References
References marked with an asterisk (*) indicate studies included in the review.
*Antonucci, S.M. (2009). Use of semantic feature analysis in group aphasia treatment. Aphasiology, 23 (7–8),854–866.
Antonucci, S. M. (2014a). How do persons with aphasia use semantic feature information during lexical retrieval: Evidence
from verbal-description and naming-to-definition of living and non - living concepts. Journal of the
International Neuropsychological Society,20 (S1).
Antonucci, S. M. (2014b). What matters in semantic feature processing for persons with stroke-aphasia: Evidence from an
auditory concept–feature verification task. Aphasiology,28 (7), 823–839.
Aphasia Treatment Evidence Tables. (2015). Academy of Neurologic Communication Disorders in profile 2015. Retrieved
from http://aphasiatx.arizona.edu/.
Beeson, P., & Robey, R. (2006). Evaluating single-subject treatment research: lessons learned from the aphasia literature.
Neuropsychological Review, 16, 161–169.
Best, W. & Nickels, L. (2000). From theory to therapy in aphasia: Where are we now and where to next?.
Neuropsychological Rehabilitation, 10(3), 231-247
Boo, M., & Rose, M. L. (2011). The efficacy of repetition, semantic, and gesture treatments for
verb retrieval and use in Broca’s aphasia. Aphasiology, 25(2), 154–175.
Bose, A. and Buchanan, L. (2007). A cognitive and psycholinguistic investigation of neologisms. Aphasiology, 21(6-8),
726-738.
*Boyle, M., & Coelho, C. (1995). Application of semantic feature analysis as a treatment for aphasic dysnomia. American
Journal of Speech- Language Pathology, 4, 94-98.
*Boyle, M. (2004). Semantic feature analysis treatment for anomia in two fluent aphasia syndromes. American Journal of
Speech- Language Pathology, 13, 236–249.
Boyle, M. (2010). Semantic feature analysis treatment for aphasic word retrieval impairments: what’s in a name? Top
Stroke Rehabilitation, 17, 411–422.
Busk, P.L., & Serlin, R.C. (1992). Meta-analysis for single-case research. Single-case research design and analysis. In:
Kratochwill TR, Levin JR, editors. Hillsdale, NJ: Lawrence Erlbaum, p. 187–212.
Byng, S. (1988). Sentence processing deficits: Theory and therapy. Cognitive Neuropsychology, 5(6), 629-676.
Cameron, R., Wambaugh, J., Wright, S., & Nessler, C. (2006). Effects of a combined semantic/phonologic cueing treatment
on word retrieval in discourse. Aphasiology,20(2-4), 269-285.
Carragher, M., Conroy, P., Sage, K., & Wilkinson, R. (2012). Can impairment-focused therapy change the everyday
conversations of people with aphasia? A review of the literature and future directions. Aphasiology, 26(7),
895–916.
Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82,
407–428.
*Coelho, C., McHugh, R., & Boyle, M. (2000). Semantic feature analysis as a treatment for aphasic dysnomia: a
replication. Aphasiology, 14, 133–142.
Conley, A., & Coelho, C.A. (2003). Treatment of word retrieval impairment in chronic Broca’s aphasia. Aphasiology, 17,
203–211.
Davis, G. A. (2000). Aphasiology: Disorders and clinical practice. Boston: Allyn & Bacon.
*Davis L.A., & Stanton, S.T. (2005). Semantic Feature Analysis as a Functional Therapy Tool. Contemporary Issues in
Communication Science and Disorders, 32, 85–92.
Dell, G. S., Schwartz, M. F., Martin, N., Saffran, E. M., & Gagnon, D. A. (1997). Lexical access in aphasic and nonaphasic
speakers. Psychological Review, 104, 801–838.
Dell, G. S., Lawler, E. N., Harris, H. D., & Gordon, J. K. (2004). Models of errors of omission in aphasic naming.
Cognitive Neuropsychology, 21, 125–145.
*DeLong, C.., Nessler, C., Wright, S., & Wambaugh, J. (2015). Semantic Feature Analysis: Further Examination of
Outcomes. American Journal of Speech-Language Pathology, 24 (4), 864 -879.
Doesborgh, S. J., Van De Sandt-Koenderman, M. W., Dippel, D. W., Van Harskamp, F., Koudstaal,
P. J., & Visch-Brink, E. G. (2004). Effects of semantic treatment on verbal communication and linguistic
processing in aphasia after stroke: A randomized controlled trial. Stroke, 35, 141–146.
Drew, R. & Thompson, C. (1999). Model-Based Semantic Treatment for Naming Deficits in Aphasia. Journal of Speech
Language And Hearing Research, 42(4), 972 - 989.

Journal of Speech, Language, and Hearing Research
23
Durand, E. & Ansaldo, A.I. (2014) Recovery from Anomia Following Semantic Feature Analysis: Therapy-induced
Neuroplasticity Relies upon a Circuit Involving Language and Motor Processing Areas. The Mental
Lexicon. Special Issue: The Neural Correlates of Lexical Processing, 21, 195-215.
Eales, C. & Pring, T. (1998). Using individual and group therapy to remediate word finding difficulties. Aphasiology,
12(10), 913-918.
Edmonds, L.A., & Kiran, S. (2006). Effect of semantic naming treatment on cross linguistic generalization in bilingual
aphasia. Journal of Speech, Language, and Hearing Research., 49, 729–748.
*Falconer, C., & Antonucci, S.M. (2012). Use of semantic feature analysis in group discourse treatment for aphasia:
Extension and expansion. Aphasiology, 26 (1), 64–82. 20.
Fotiadou D., Northcott S., Chatzidaki A., & Hilari K. (2014) Aphasia blog talk: How does stroke and aphasia affect a
person’s social relationships? Aphasiology, 28, 1281-1300
Franklin, S., Buerk, F., & Howard, D. (2002). Generalised improvement in speech production for a subject with
reproduction conduction aphasia. Aphasiology, 16(10-11), 1087-1114.
Gast, D. (2010). Single subject research methodology in behavioural sciences. New York: Routledge.
Goldrick, M. (2006). Limited interaction in speech production: Chronometric, speech error, and neuropsychological
evidence. Language and Cognitive Processes, 21, 817–855.
Goodglass, H., Kaplan, E., & Barresi, B. (2001). The Boston Diagnostic Aphasia Examination: BDAE-3 Short Form.
Philadelphia, PA: Lippincott, Williams, and Wilkins.
Goodglass, H., & Wingfield, A. (1997). Anomia: Neuroanatomical and cognitive correlates. San Diego, CA: Academic
Press.
Greenwood, A., Grassly, J., Hickin, J., & Best, W. (2010). Phonological and orthographic cueing therapy: A case of
generalised improvement. Aphasiology, 24(9), 991-1016.
Haarbauer-Krupa, J., Moser, L., Smith, G., Sullivan, D.M., & Szekeres, S.F. (1985a). Cognitive rehabilitation therapy:
middle stages of recovery. In: Ylvisaker M, ed. Head Injury Rehabilitation: Children and Adolescents.
San Diego, CA: College-Hill Press, 287–310.
Haarbauer-Krupa, J., Moser, L., Smith, G., Sullivan, D.M., & Szekeres, S.F. (1985b). Cognitive rehabilitation therapy: late
stages of recovery. In: Ylvisaker M, ed. Head Injury Rehabilitation: Children and Adolescents. San
Diego, CA: College-Hill Press, 311–340.
*Hashimoto, N., & Frome, A. (2011). The use of a modified semantic features analysis approach in aphasia. Journal of
Communication Disorders, 44, 459–469.
*Hashimoto, N. (2012). The use of semantic-and phonological-based feature approaches to treat naming deficits in
aphasia, Clinical Linguistics & Phonetics, 26, 518-533.
Hashimoto, N. (2016). The role of effects of thematic relations on picture naming abilities across the lifespan. Aging,
Neuropsychology, & Cognition, 23, 499-512.
Helm-Estabrooks, N. (1992). Aphasia Diagnostic Profiles. Chicago: Riverside.
Hickin, J., Best, W., Herbert, R., Howard, D., & Osborne, F. (2002). Phonological therapy for word-finding difficulties: A
re-evaluation. Aphasiology, 16(10-11), 981-999.
Hilari, K., Needle, J. J., & Harrison, K. L. (2012). What are the important factors in health-related quality of life for people
with aphasia? A systematic review. Archives of Physical Medicine and Rehabilitation, 93, 86-95.
Hough, M. (1993). Treatment of Wernicke's aphasia with jargon: A case study. Journal Of Communication Disorders,
26(2), 101-111.
Howard, D. (2000). Cognitive neuropsychology and aphasia therapy: the case of word retrieval. In: Papathanasiou I, ed.
Acquired Neurogenic Communication Disorders: A Clinical Perspective. London, England: Whurr, 76–
99.
Howard, D. (1994). The Treatment of Acquired Aphasia. Philosophical Transactions Of The Royal Society B: Biological
Sciences, 346(1315), 113-120.
Howard, D., & Gatehouse, C. (2006). Distinguishing semantic and lexical word retrieval deficits in people with aphasia.
Aphasiology, 20, 921–950.
Howard, D., Patterson, K., Franklin, S., Orchard-lisle, V., & Morton, J. (1985). The facilitation of picture naming in
aphasia. Cognitive Neuropsychology, 2(1), 49-80.
Hung, J., Bauer, A., Grossman, M., Hamilton, R., Coslett, H., & Reilly, J. (2017). Semantic Feature Training in
Combination with Transcranial Direct Current Stimulation (tDCS) for Progressive Anomia. Frontiers in
Human Neuroscience, 11.
Kertesz, A. (2007). Western Aphasia Battery—Revised. San Antonio, TX: Harcourt Assessment.
Kintz, S., Wright, H., & Fergadiotis, G. (2016). Semantic knowledge use in discourse produced by individuals with anomic
aphasia. Aphasiology, 30 (9), 1012-1025.
Kiran, S., & Bassetto, G. (2008). Evaluating the Effectiveness of Semantic-Based Treatment for Naming Deficits in
Aphasia: What Works? Seminars in Speech and Language, 29 (1), 71-82.
Kiran, S., & Viswanathan, M. (2008). Effect of model based treatment on oral reading abilities in severe alexia: A case
study. Journal of Medical Speech Pathology,16 (1), 43-59.
Kiran, S., & Roberts, P.M. (2010). Semantic feature analysis treatment in Spanish-English and French English bilingual
aphasia. Aphasiology, 24, 231–261.
Kladouchou, V., Papathanasiou, I., Efstratiadou, E., Christaki, V., & Hilari, K. (2017). Treatment integrity of elaborated
semantic feature analysis aphasia therapy delivered in individual and group settings. International Journal
Of Language & Communication Disorders.
*Knoph, M. I. N., Lind, M., & Simonsen H. G. (2015). Semantic feature analysis targeting verbs in a quadrilingual speaker
with aphasia, Aphasiology, 29 (12), 1473-1496.

Journal of Speech, Language, and Hearing Research
24
Knoph, M., Simonsen, H., & Lind, M. (2017). Cross-linguistic transfer effects of verb-production therapy in two cases of
multilingual aphasia. Aphasiology, 31(12), 1482-1509.
*Kristensson, J., Behrns, I., & Saldert, C. (2015). Effects on communication from intensive treatment with semantic feature
analysis in aphasia. Aphasiology, 29 (4), 466–487.
Laine, M., & Martin, N. (2006). Anomia: Theoretical and Clinical Aspects. Hove, UK: Psychology Press.
Law, S.P., Wong, W., Sung, F., & Hon, J. (2006). A study of semantic treatment of three Chinese anomic patients.
Neuropsychological Rehabilitation, 16, 601–629.
Leonard, C., Rochon, E., & Laird, L. (2004). The benefits of orthographic cueing to naming deficits in aphasia: A case
study. Brain and Language, 19, 185-186.
Leonard, C., Rochon, E., & Laird, L. (2008). Treating naming impairments in aphasia: Findings from a phonological
components analysis treatment. Aphasiology, 22, 923–947.
Levelt, W. J. M. (1999). Models of word production. Trends in Cognitive Sciences, 3, 223–232.
Levelt, W. J. M., Roelofs, A., & Meyer, A. S. (1999). A theory of lexical access in speech production. Behavioral and
Brain Sciences, 22, 1–75.
Linebaugh, C.W., Shisler, R.J., & Lehner, L., (2005). Cueing hierarchies and word retrieval: A therapy program.
Aphasiology, 19(1), 77-92.
Lorenz, A., & Ziegler, W. (2009). Semantic vs. word-form specific techniques in anomia treatment: A multiple single-case
study. Journal of Neurolinguistics, 22, 515–537.
Lorenz, A. & Nickels, L. (2007). Orthographic cueing in anomic aphasia: How does it work?. Aphasiology, 21(6-8), 670-
686.
Lowell, S., Beeson, P.M., & Holland, A.L. (1995). The efficacy of a semantic cueing procedure on naming performance of
adults with aphasia. American Journal of Speech-Language Pathology, 4, 109–114.
Maddy, K.M., Capilouto, G.J., McComas, K.L. (2014). The effectiveness of semantic feature analysis: An evidence-based
systematic review. Annals of Physical and Rehabilitation Medicine, 57, 254–267.
Maher, L. M., & Raymer, A. M. (2004). Management of anomia. Topics in Stroke Rehabilitation, 11, 10–21.
*Marcotte, K., & Ansaldo, AI. (2010). The neural correlates of semantic feature analysis in chronic aphasia: discordant
patterns according to the etiology. Seminars in Speech and Language. Specials Issue on Treatment of
Chronic Aphasia: International Perspectives, 31(1), 52–63.
Marshall, J., Pound, C., White-thomson, M., & Pring, T. (1990). The use of picture/word matching tasks to assist word
retrieval in aphasic patients. Aphasiology, 4(2), 167-184.
Martin, N., Fink, R., & Laine, M. (2004). Treatment of word retrieval deficits with contextual priming. Aphasiology, 18(5-
7), 457-471.
*Massaro, M., & Tompkins, C. (1994). Feature analysis for treatment of communication disorders in traumatically brain-
injured patients: an efficacy study. Clinical Aphasiology, 22, 245–256.
*Mehta S. V., & Isaki E. (2016). A modified Semantic Feature Analysis Approach with two individuals with chronic
aphasia. Contemporary Issues in Communication Science and Disorders, 43, 129 -138.
Mitchum, C., Haendiges, A., & Berndt, R. (1993). Model-guided treatment to improve written sentence production: A case
study. Aphasiology, 7(1), 71-109.
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D.G. (2009). Preferred reporting items for systematic reviews and meta-
analyses: The PRISMA statement. Annals of Internal Medicine, 151, 264–269.
Moher, D., Liberati, A., Tetzlaff, J., & Alman, D. G. (2010). Preferred reporting items for systematic reviews and meta-
analyses: the PRISMA statement. International Journal of Surgery, 8, 336-341.
Munro, P., & Siyambalapitiya, S. (2016). Improved word comprehension in Global aphasia using a modified semantic
feature analysis treatment. Clinical Linguistics & Phonetics, 31(2), 119-136.
*Neumann, Y. (2017). A case series comparison of semantically focused vs. phonologically focused cued naming treatment
in aphasia. Clinical Linguistics & Phonetics, 1-27.
Nickels, L. (2002). Therapy for naming disorders: revisiting, revising, and reviewing. Aphasiology, 16, 935–979.
Nickels, L. & Best, W. (1996). Therapy for naming disorders (Part I): Principles, puzzles and progress. Aphasiology, 10(1),
21-47.
Nickels, L. & Best, W. (1996). Therapy for naming disorders (Part II): Specifics, surprises and suggestions. Aphasiology,
10(2), 109-136.
Northcott, S., Moss, B., Harrison, K., & Hilari, K. (2016). A systematic review of the impact of stroke on social support and
social networks: associated factors and patterns of change. Clinical Rehabilitation, 30(8), 811-831.
Olsen, E., Freed, D. B., Marshall, R. C., (2012). Generalisation of personalised cueing to enhance word finding in natural
settings. Aphasiology, 26(5), 618-631.
Paradis, M., Libben, G., & Hummel, K. (1987). The bilingual aphasia test (English version). Hillsdale, NJ: Lawrence
Erlbaum Associates.
*Peach, R., & Reuter, K. (200). A discourse-based approach to semantic feature analysis for the treatment of aphasic word
retrieval failures. Aphasiology, 24, 971–990.
Raymer, A. M., Beeson, P., Holland, A., Kendall, D., Maher, L. M., Martin, N., et al. (2008). Translational research in
aphasia: From neuroscience to neurorehabilitation. Journal of Speech, Language, and Hearing Research,
51(1), 259–275.
Raymer, A., Thompson, C., Jacobs, B., & Le Grand, H. (1993). Phonological treatment of naming deficits in aphasia:
Model-based generalization analysis. Aphasiology, 7(1), 27-53.
Renvall, K., Laine, M., & Martin, N. (2007). Treatment of anomia with contextual priming: Exploration of a modified
procedure with additional semantic and phonological tasks. Aphasiology, 21(5), 499-527.

Journal of Speech, Language, and Hearing Research
25
*Rider, J.D., Wright, H.H., Marshall, R.C., Page, J.L. (2008). Using semantic feature analysis to improve contextual
discourse in adults with aphasia. American Journal of Speech-Language Pathology, 17, 161–172.
Robey, R. R., & Schultz, M. C. (1998). A model for conducting clinical outcome research: An adaptation of the standard
protocol for use in aphasiology. Aphasiology, 12 (9), 787–810.
Robey, R. R. (2004). A five-phase model for clinical-outcome research. Journal of Communication Disorders, 37 (5), 401 -
411.
Robson, J., Pring, T., Marshall, J., Morrison, S., & Chiat, S. (1998). Written communication in undifferentiated jargon
aphasia: a therapy study. International Journal of Language & Communication Disorders, 33(3), 305-
328.
*Sadeghi Z, Baharloei N, Moddarres Zadeh A, Ghasisin L. (2017). Comparative Effectiveness of Semantic Feature
Analysis (SFA) and Phonological Components Analysis (PCA) for Anomia Treatment in Persian
Speaking Patients With Aphasia. Iranian Rehabilitation Journal, 15 (3), 259-268.
Schlosser, R.W., Lee, D.L., & Wendt, O. (2008). Application of the percentage of non - overlapping data (PND) in
systematic reviews and meta-analyses: a systematic review of reporting characteristics. Evidence Based
Communication Assess Intervention, 2, 163–87.
Schwartz, M.F., & Brecher, A. (2000). A model-driven analysis of severity, response characteristics, and partial recovery in
aphasics' picture naming. Brain and Language, 73(1), 62–91.
Schwartz, M. F., Dell, G. S., Martin, N., Gabl, S., & Sobel, P. (2006). A case-series of test of the interactive two-step model
of lexical access: Evidence from picture naming. Journal of Memory and Language, 54, 228–264.
Scottish Intercollegiate Guidelines Network. (2010). Retrieved from http://www.sign.ac.uk/pdf/sign118.pdf.
Scruggs, T., Mastropieri, M., & Casto, G. (1987). The quantitative synthesis of single subject research: methodology and
validation. Remedial and Special Education, 8 (2), 24–33.
Tate, R., McDonald, K., Perdices, M., Tougher, L., Schultz, R., & Savage, S. (2008). Rating the methodological quality of
single subject designs and n-of-1 trials: introducing the Single-Case Experimental Design (SCED) Scale.
Neuropsychological Rehabilitation, 18, 385–401.
*van Hees, S., Angwin, A., McMahon, K., & Copland, D. (2013). A comparison of semantic feature analysis and
phonological components analysis for the treatment of naming impairments in aphasia.
Neuropsychological Rehabilitation, 23(1), 102–132.
van Hees, S., Mcmahon, K., Angwin, A., De Zubicaray, G., & Copland, D. A. (2014a). Neural activity associated with
semantic versus phonological anomia treatments in aphasia. Brain Language, 129, 47–57.
van Mourik, M., Verschaeve, M., Boon, P., Paquier, P., & van Harskamp, F., (1992). Cognition in global aphasia:
Indicators or therapy. Aphasiology, 5, 491-499.
Wambaugh, J. L., Linebaugh, C. W., Doyle, P. J., Martinez, A. L., & Kalinyak-Fliszar, M. M. (2001). Effects of two cueing
treatments on lexical retrieval in aphasic speakers with different levels of deficit. Aphasiology, 15(10/11),
933–950.
*Wambaugh, J. L., & Ferguson, M. (2007). Application of semantic feature analysis to retrieval of action names in aphasia.
Journal of Rehabilitation Research and Development, 44, 381–394.
Wambaugh, J.L., Mauszycki, S., Cameron, R., Wright, S., & Nessler, C. (2013). Semantic feature analysis: incorporating
typicality treatment and mediating strategy training to promote generalization. American Journal of
Speech-Language Pathology, 22, 334–369.
*Wambaugh, J.L., Mauszycki, S., &Wright, S. (2014). Semantic feature analysis: Application to confrontation naming of
actions in aphasia. Aphasiology, 28 (1), 1-24.
Wallance, S.E., Kimelman, M.D.Z. (2013). Generalization of word retrieval following semantic feature treatment.
NeuroRehabilitation, 32(4), 899-913.
Wilshire, C. E. (2002). Where do aphasic phonological errors come from? Evidence from phoneme movement errors in
picture naming. Aphasiology, 16, 169–197.
Wisenburn, B., & Mahoney, K. (2009). A meta-analysis of word-finding treatments for aphasia. Aphasiology, 23(11), 1338-
1352.
Ylvisaker, M., & Szekeres, S. (1985). Cognitive-language intervention with brain injured adolescents and adults. Paper
presented at: Annual Convention of the Illinois Speech-Language-Hearing Association. Chicago, Illinois.

Journal of Speech, Language, and Hearing Research
26
Figure 1: The feature analysis charts for nouns and verbs
Noun SFA Verb SFΑ
Boyle, 2004; Coelho et al., 2000 Wambaugh & Ferguson, 2007; Wambaugh et al., 2014
Noun
Picture
Group
Is a ........
Action
Does.......
Use
Is used for....
Location
Is found.....
Properties
Looks like....
Has......
Association
Goes with.....
Verb
Picture
Subject
Who......?
Purpose of
Action
Why.......?
How
What......?
Location
Where.....?
Properties
Looks like....
Has......
Related
Objects or
Actions
What..........?

Figure 2: Identification process of articles from electronic databases
SFA: Semantic Features Analysis
PCA: Phonological Components Analysis

Journal of Speech, Language, and Hearing Research
28
Table 1: Critical appraisal and methodological quality of studies (n=17) based on Single Case Experimental Design Scale (SCED)
Items of SCED
Scale Clinical
History Target
Behaviours Design Baseline Treat
ment
Phase
Raw
Data
Record
Inter-
Rater
Reliability
Independ
ence of
Assessors
Statistical
Analysis Replication Generalization
Total
Score
of
SCED
Scale
Phase of
treatment
1. Massaro &
Tomkins, 1994
YES YES MBAB YES YES YES YES YES NO YES YES 10
Pre-efficacy 1
2. Boyle &
Coelho, 1995
YES YES ABA YES YES YES YES NO NO NO YES 8
Pre-efficacy 1
3. Coelho et al.,
2000
YES YES ABA YES YES YES YES NO NO NO YES 8
Pre-efficacy 1
4. Boyle, 2004 YES YES MBAB YES YES YES YES YES YES YES YES 11
Pre-efficacy 1
5. Davis &
Stanton, 2005
YES YES MBAB YES YES YES YES NO NO NO YES 8
Pre-efficacy 1
6. Wambaugh
& Ferguson,
2007
YES YES MBAB YES YES YES YES YES YES YES NO 10
Pre-efficacy 1
7. Rider et al.,
2008 YES YES MBAB YES YES YES YES YES YES YES NO 10
Efficacy
8. Antonucci,
2009
YES YES ABA YES YES YES YES YES YES YES Partly 10.5
Pre-efficacy 2
9. Marcotte
& Ansaldo,
2010
YES AB Not a single case study but an observation control study No
Pre-efficacy 1
10. Peach &
Reuter, 2010 YES YES
Single case
time series
across
behaviors
YES YES YES YES YES YES YES Variable 10.5
Pre-efficacy 1
11. Hashimoto
& Frome, 2011
YES YES MBAB YES YES YES YES YES YES YES YES 11
Pre-efficacy 1
12. Falconer &
Antonucci,
2012
YES YES ABA YES YES YES YES NO YES YES YES 9
Pre-efficacy 2

Journal of Speech, Language, and Hearing Research
29
Items of SCED
Scale Clinical
History Target
Behaviours Design Baseline Treat
ment
Phase
Raw
Data
Record
Inter-
Rater
Reliability
Independ
ence of
Assessors
Statistical
Analysis Replication Generalization
Total
Score
of
SCED
Scale
Phase of
treatment
13. Hashimoto,
2012 YES YES MBAB YES YES YES YES YES YES YES Partly 10.5
Pre-efficacy 2
14. van Hees et
al., 2013 YES YES ABA YES YES YES NO NO YES YES NO 8
Pre-efficacy 2
15. Wambaugh
et al., 2014
YES YES MBAB YES YES YES YES YES YES YES NO 10
Pre-efficacy 2
16. Kristensson
et al., 2015
YES YES MBAB YES YES YES YES YES YES YES NO 10
Pre-efficacy 1
17. DeLong et
al., 2015 YES YES
MBAB YES YES YES YES YES YES YES Variable 10.5
Pre-efficacy 1
18. Knoph et
al.,2015
YES YES AB YES YES YES YES NO YES YES YES 10
Pre-efficacy 2
19. Mehta &
Isaki, 2016
YES YES ABA YES YES YES NO YES YES YES NO 9
Pre-efficacy 2
20. Neumann,
2017
YES YES MBAB YES YES YES NO NO YES YES YES 9
Pre-efficacy 2
21. Sadeghi et
al., 2017
YES YES AB YES YES YES NO NO YES YES YES 8
Pre-efficacy 2
SCED: Single Case Experimental Design
MBAB: Multiple baseline across behaviors study, involving multiple assessments pre- treatment, post-treatment and follow up
AB: Pre- / post- treatment study
ABA: Pre- / post- treatment / follow up study

Journal of Speech, Language, and Hearing Research
30
Table 2: Study characteristics: number of participants, type of SFA treatment,
dosage, duration and amount of treatment
Study
n
Type of
SFA
Language
Treatment dosage and
duration
Total amount of
treatment (mins)
1. Massaro &
Tompkins, 1994
2
Noun
SFA
Monolingual
21 sessions
CNC
2. Boyle & Coelho, 1995
1
Noun
SFA
Monolingual
3*60min sessions/wk
6 weeks
1080
3. Coelho et al., 2000
1
Noun
SFA
Monolingual
3*60min sessions/wk
7 weeks
1260
4. Boyle, 2004
2
Noun
SFA
Monolingual
3*50- 75 min sessions/wk
8 weeks
≈1500
5. Davis & Stanton,
2005
1
Noun
SFA
Monolingual
2* 60 min sessions/wk
6 weeks
720
6. Wambaugh &
Ferguson, 2007
1
Verb
SFA
Monolingual
3*45 - 60 min sessions/wk
4 weeks
≈630
7. Rider et al., 2008
3
Noun SFA
Monolingual
2-3 * 60min sessions/wk
5 weeks
or 80% naming accuracy
across 2 sessions
≈750
8. Antonucci, 2009
3
Group
Approach
Discourse
SFA
Monolingual
2*60 -90min sessions/wk
7 weeks
≈1050
9.Marcotte & Ansaldo,
2010
1
Nouns &
Verb
SFA
Monolingual
3*60min sessions/wk
3 weeks
540
10. Peach and Reuter,
2010
2
Discourse
SFA
Bilingual
P1: 14 *50 min per
sessions
10 weeks
P2: 13*50 min per
sessions
10 1⁄2 weeks
≈675
11. Hashimoto &
Frome, 2011
1
Modified
Noun
SFA
Monolingual
2*60min sessions/wk
12 weeks
1440
12. Falconer &
Antonucci, 2012
4
Group
Approach
Discourse
SFA
Monolingual
2* 90 - 120 min
sessions/wk
7 weeks
& daily practice of
homework
≈1470

Journal of Speech, Language, and Hearing Research
31
Study
n
Type of
SFA
Language
Treatment dosage and
duration
Total amount of
treatment (mins)
13. Hashimoto, 2012
2
SFA
vs
PCA
Monolingual
2*45-60min sessions per
day
4 sessions/wk
until >80%
naming
accuracy across 3 sessions
2 – 7 ½ weeks
≈1470
14. van Hees et al.,2013
8
SFA
vs
PCA
Monolingual
3* 45-90min sessions/wk
4 weeks
≈810
15. Wambaugh et al.,
2014
4
Verb
SFA
Monolingual
3*60min sessions/wk
Until 90% accurate
naming of trained items in
2-3 probes or 4 weeks
720
16. Kristensson et al.,
2015
3
Nouns &
Verb
SFA
Monolingual
20* 60min sessions
5-6 weeks
≈1200
17. DeLong et al., 2015
5
Noun
SFA
Monolingual
3* 50 min sessions/wk
Max 20 treatment sessions
per treatment phase or
86% items correct in 2 of
3 consecutive probe
sessions
1000
18. Knoph et al.,2015
1
Verb
Quadrilingu
al SFA
Quadrilingual
29 sessions
3 days per week
2.5 weeks
1320
19. Mehta & Isaki,
2016
2
Noun
SFA
Monolingual
2*60min sessions/wk
8 weeks
720
20. Neumann, 2017
4
SFA
vs
PCA
2 Monolingual
2 Bilingual
2-3 * 120 min sessions/wk
Max 10 trained items
given or 40% or more
above baseline in naming
improvement in the
treated items on 2
consecutive probe sessions
2 to 6 ½ weeks
CNC
21. Sadeghi et al.,
2017
4
SFA
vs
PCA
Monolingual
7*45 min sessions
2 weeks
315
CNC: Cannot calculate

Journal of Speech, Language, and Hearing Research
32
Table 3: Participants’ demographic and stroke and aphasia characteristics (N=51)
Study
n
Participants
Age
(years)
Gender
Etiology
TPO
(months)
WAB AQ
Aphasia
Severitya
Aphasia
Type
Fluency
1. Massaro & Tompkins, 1994
2
P1
P2
24
28
M
F
TBI
TBI
60
144
NR
NR
Broca
NR
Non – Fluent
Non -Fluent
2. Boyle &Coelho, 1995
1
P1
57
M
L CVA
65
82 Mild
Broca
Non - Fluent
3. Coelho et al., 2000
1
P1
52
M
TBI
17
56.6 Moderate
NR
Fluent
4. Boyle, 2004
2
P1
P2
70
80
M
M
L CVA
LCVA
15
14
90.6 Mild
61.2 Moderate
Anomic
Wernicke
Fluent
Fluent
5. Davis & Stanton, 2005
1
P1
59
F
CVA
4
102b Moderate
NR
Fluent
6. Wambaugh & Ferguson, 2007
1
P1
74
F
L CVA
50
67.7 Moderate
Anomic
Non - Fluent
7. Rider et al., 2008
3
P1
P2
P3
73
55
62
M
F
M
L CVA
L CVA
L CVA
26
45
126
74.6 Moderate - Mild
76.5 Mild
66 Moderate
Transcortical Motor
Transcortical Motor
Broca
Non – Fluent
Non – Fluent
Non - Fluent
8. Antonucci, 2009
3
P1
P2
P3
NR
53
59
M
M
F
NR
NR
NR
NR
18
16
NR
63 Moderate
90.2 Mild
NR
Conduction
NR
NR
Fluent
Fluent
9. Marcotte & Ansaldo, 2010
1
P1
66
M
CVA
84
Severe
Broca
Non – Fluent
10. Peach & Reuter, 2010
2
P1
P2
77
62
F
F
L CVA
L CVA
4
14
90.2 Mild
70.3 Moderate
Anomic
Anomic
Fluent
Fluent
11.
Hashimoto & Frome, 2011
1
P1
72
F
CVA
NR
35 Severe
Broca
Non -Fluent
12. Falconer & Antonucci, 2012
4
P1
P2
P3
P4
35
55
31
62
M
M
M
F
M CVA
L CVA
TBI
M CVA
72
156
96
25
69.6 Moderate
61 Moderate
34 Severe
52.4 Moderate
Conduction
Conduction
Broca
Transcortical Motor
Fluent
Fluent
Non –Fluent
Non -Fluent
13. Hashimoto, 2012
2
P1
P2
66
33
F
F
L CVA
L CVA
60
18
49.5 Severe - Moderate
57.5 Moderate
Wernicke
Broca
Non – Fluent
Fluent
14. van Hees et al., 2013
8
P1
P2
P3
P4
P5
P6
P7
60
60
41
52
56
48
69
F
M
F
F
F
F
M
L CVA
L CVA
L CVA
L CVA
L CVA
L CVA
L CVA
38
57
170
55
25
17
36
77.2 Mild – Moderate
87.4 Mild
92 Mild
86.4 Mild
57.3 Moderate
81.7 Mild
73.4 Moderate
Conduction
Anomic
Anomic
Conduction
Anomic
Anomic
Anomic
Fluent
Fluent
Fluent
Fluent
Fluent
Fluent
Fluent

Journal of Speech, Language, and Hearing Research
33
Study
n
Participants
Age
(years)
Gender
Etiology
TPO
(months)
WAB AQ
Aphasia
Severitya
Aphasia
Type
Fluency
P8
65
M
L CVA
20
82.9 Mild
Anomic
Fluent
15. Wambaugh et al., 2014
4
P1
P2
P3
P4
48
53
55
60
F
M
M
M
L MCA
L PCA
L CVA
R MCA L MCA
276
66
79
21
77.4 Mild
83.4 Mild
53 Moderate
66.9 Moderate
Conduction
Anomic
Broca
Broca
Fluent
Fluent
Non – Fluent
Non - Fluent
16. Kristensson et al., 2015
3
P1
P2
P3
71
54
64
M
F
M
L PCA
L BG
L MCA
36
60
24
Moderate – Severe
Moderate – Severe
Mild – Moderate
Wernicke
Mixed
Broca
Fluent
Non- Fluent
Non – Fluent
17. DeLong et al., 2015
5
P1
P2
P3
P4
P5
62
54
30
53
65
F
M
M
F
F
L CVA
L MCA
L MCA
L MCA
L MCA
11
30
23
384
12
64.5 Moderate
58.3 Moderate
66 Moderate
78.4 Moderate
18 Very Severe
Conduction
Wernicke
Broca
Anomic
Global
Fluent
Fluent
Fluent
Fluent
Non – Fluent
18. Knoph et al., 2015
1
P1
59
F
L NR
7
Moderatec
NR
Non – Fluent
19. Mehta & Isaki, 2016
2
P1
P2
58
58
M
M
L CVA
L CVA
108
132
53 Moderate
60.2 Moderate
Wernicke
Conduction
Fluent
Fluent
20. Neumann, 2017
4
P1
P2
P3
P4
41
38
60
47
M
F
M
M
L CVA
L CVA
L NR
L NR
96
24
84
24
Moderated
Mildd
Mildd
Severe
d
Conduction
Anomic
Anomic
Anomic
Fluent
Fluent
Fluent
Fluent
21. Sadeghi et al., 2017
4
P1
P2
P3
P4
61
52
45
47
M
F
M
M
L CVA
L CVA
L MCA
L CVA
24
17
67
15
NR
NR
NR
NR
Broca
Broca
Anomic
Broca
Non – Fluent
Non – Fluent
Fluent
Non – Fluent
a: Aphasia severity based on Western Aphasia Battery-Revised (Kertesz, 2007) Aphasia Quotient. Retrieved October 1, 2015,
from http://www.pearsonclinical.com/language/products/100000194/western-aphasia-batteryrevised.html
b: Based on Aphasia Diagnostic Profiles score (Helm-Estabrooks, 1992)
c: Based on Bilingual Aphasia Test (Paradis, Libben, & Hummel, 1987)
d: Based on Boston Diagnostic Aphasia Examination – Short Form (Goodglass, Kaplan &Barresi, 2001)
NR: not reported; R: Right hemisphere; L: left hemisphere; TPO: Time Post Onset; MCA: Middle Cerebral Artery; CVA: Cerebral Vascular Accident; PCA: Posterior Cerebral Artery; BG:
Basal Ganglia

Journal of Speech, Language, and Hearing Research
34
Table 4: Summary of treatment outcomes
Study
n
Treated items
improved?
Maintenance
Generalization to
untreated items?
1. Massaro &
Tompkins, 1994
2
YES
YES
YES
YES
YES
YES
2. Boyle & Coelho 1995
1
YES
YES
YES
3. Coelho et al., 2000
1
YES
YES
YES
4. Boyle, 2004
2
YES
YES
YES
Unavailable
YES
YES
5. Davis &
Stanton, 2005
1
YES
YES
YES
6. Wambaugh
& Ferguson,
2007
1
YES
YES
NO
7. Rider et al., 2008
3
YES
YES
YES
YES
YES
NO
NO
NO
NO
8. Antonucci, 2009
3
YES
YES
NO
YES
YES
Unavailable
YES
NO
Unavailable
9. Marcotte & Ansaldo,
2010
1
YES
10. Peach and Reuter,
2010
2
YES
YES
NO
NO
Variable
Variable
11. Hashimoto &
Frome, 2011
1
YES
YES
YES
12. Falconer &
Antonucci, 2012
4
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
13. Hashimoto, 2012
2
YES
YES
YES
YES
NO
YES
14. van Hees et al., 2013
8
YES
YES
YES
YES
NO
NO
NO
NO
YES
NO
YES
YES
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
15. Wambaugh et al.,
2014
4
YES
YES
YES
YES
YES
YES
NO
NO
NO

Journal of Speech, Language, and Hearing Research
35
Study
n
Treated items
improved?
Maintenance
Generalization to
untreated items?
NO
NO
NO
16. Kristensson et al.
2015
3
NO
NO
NO
NO
NO
NO
NO
NO
NO
17. DeLong et al., 2015
5
YES
YES
YES
YES
NO
YES
NO
YES
NO
NO
NO
NO
NO
NO
NO
18. Knoph et al., 2015
1
YES
NO
19. Mehta & Isaki, 2016
2
YES
YES
YES
YES
20. Neumann, 2017
4
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
21. Sadeghi et al., 2017
4
YES
YES
YES
YES
YES
YES
YES
YES
Total
55
YES n=45 (81.82%)
NO n=10 (18.19%)
YES n=32 (58.18%)
NO n=15 (27.27%)
Unavailable n=2 (3.63 %)
NR n=6
YES n=22 (40%)
NO n=27 (49.09%)
Variable n=2 (3.63%)
Unavailable n=1 (1.81%)

Journal of Speech, Language, and Hearing Research
36
Table 5: Time of Assessments after Therapy
Study
Number of
Assessments
Time of Assessment after Therapy
1. Massaro & Tompkins,
1994
2
Immediately after therapy
2 weeks
2. Boyle & Coelho 1995
3
Immediately after therapy
1 month
2 months
3. Coelho et al., 2000
3
Immediately after therapy
1 month
2 months
4. Boyle, 2004
2
Immediately after therapy
1 month
5. Davis & Stanton,
2005
5
Immediately after therapy
6 weeks
12 weeks
18 weeks
1 year
6. Wambaugh &
Ferguson, 2007
3
Immediately after therapy
2 weeks
6 weeks
7. Rider et al., 2008
2
Immediately after therapy
4 weeks
8. Antonucci, 2009
2
Immediately after therapy
6 weeks
9. Marcotte & Ansaldo, 2010
1
Immediately after therapy
10. Peach and Reuter, 2010
2
Immediately after therapy
4 ½ months
11. Hashimoto & Frome, 2011
2
Immediately after therapy
6 weeks
12. Falconer & Antonucci,
2012
2
Immediately after therapy
6 weeks
13. Hashimoto, 2012
2
Immediately after therapy
6 weeks
14. van Hees et al., 2013
2
Immediately after therapy
2-3 weeks
15. Wambaugh et al., 2014
3
Immediately after therapy
2 weeks
6 weeks
16. Kristensson et al. 2015
2
Immediately after therapy
10-12 weeks
17. DeLong et al., 2015
3
Immediately after therapy
2 weeks
6 weeks
18. Knoph et al., 2015
1
Immediately after therapy
19. Mehta & Isaki, 2016
2
Immediately after therapy
8 weeks
20. Neumann, 2017
2
Immediately after therapy
4-6 weeks
21. Sadeghi et al., 2017
1
Immediately after therapy

Journal of Speech, Language, and Hearing Research
37
Table 6: Clinical Efficacy: effect sizes and percent of non-overlapping data.
Study
Participants
Cohen’s d
PND
Magnitude of effect
1. Massaro & Tompkins, 1994
P1
P2
7.45c
3.54c
Medium effect
Less than small effect
2. Boyle & Coelho 1995
P1
16.36c
Large effect
3. Coelho et al., 2000
P1
4.41c
Small effect
4. Boyle, 2004
P1
P2
18.48a
CNC
100%
Large effect
Highly effective
5. Davis & Stanton, 2005
P1
CNC
91.67%
Highly effective
36. Wambaugh &
Ferguson, 2007
P1
6.35a
Small effect
7. Rider et al., 2008
P1
P2
P3
3.86b
5.54b
2.97b
Less than small effect
Small effect
Less than small effect
8. Antonucci, 2009
P1
P2
P3
CNC
ns
2.05b
CNC
CNC
CNC
Less than small effect
9. Marcotte & Ansaldo, 2010
P1
CNC
CNC
-
-
10. Peach & Reuter, 2010
P1
P2
1.79
85%
Less than small effect
Moderate effective
11. Hashimoto & Frome, 2011
P1
10.56b
Large effect
12. Falconer & Antonucci, 2012
P1
P2
P3
P4
3.44
4.16
0.03
1.28
Less than small effect
Small effect
Less than small effect
Less than small effect
13. Hashimoto, 2012
P1
P2
7.11
7
Medium effect
Medium effect
14. van Hees et al., 2013
P1
P2
P3
P4
P5
P6
P7
P8
5.86
3.79
4.54
7.79
ns
ns
ns
ns
Small effect
Less than small effect
Small effect
Medium effect
-
-
-
-
15. Wambaugh et al., 2014
P1
P2
P3
P4
6.87b
13.14b
1.58b
8.53b
Small effect
Large effect
Less than small effect
Medium effect
16. Kristensson et al., 2015
P1
P2
P3
1.06b
0.66b
0.64b
Less than small effect
Less than small effect
Less than small effect
17. DeLong et al., 2015
P1
P2
P3
P4
P5
3.03b
2.20b
4.68b
6.66b
CNC
CNC
Less than small effect
Less than small effect
Small effect
Small effect
-

Journal of Speech, Language, and Hearing Research
38
18. Knoph et al., 2015
P1
10.07
Large effect
19. Mehta & Isaki, 2016
P1
P2
CNC
CNC
CNC
CNC
-
-
20. Neumann, 2017
P1
P2
P3
P4
1.12b
9.58b
1.67b
11.21b
Less than small effect
Medium effect
Less than small effect
Large effect
21. Sadeghi et al., 2017
P1
P2
P3
P4
5.08 b
6.89 b
19.23b
11.89b
Small effect
Small effect
Large effect
Large effect
PND: percent of non-overlapping data; CNC: Cannot calculate, a: Calculated by first author of this paper, b: Average
calculation by first author of this paper, c : Calculated by first author of this paper based on substituted data from other
phases, ns: no substantial change