A scoping review of communication outcomes measures in augmentative and alternative communication

ABSTRACT

Although outcomes are a critical component of evidence-based practice, measuring augmentative and alternative communication (AAC) outcomes remains problematic. This is, in part, because there is no consensus on how to operationally define AAC communication outcomes. To gain greater insight into AAC communication outcomes, we used the communicative competence framework to determine which areas of AAC intervention have received the greatest attention and how these outcomes are being measured. The following data were charted from the 77 studies that met the inclusion criteria for the scoping review: study design, study participants, study communication target (e.g., language, word learning, etc.), and communication outcome measurements. Across the included studies, researchers used a variety of standardized and non-standardized measures to assess outcomes. Seventy-seven percent of the studies assessed social skills and 62% assessed linguistic skills. A limited number of studies measured operational (14%), strategic (4%), and psychosocial (18%) skills. Using the communicative competence framework enabled us to identify gaps in the research that has been conducted to date.

KEYWORDS:

• Adults
• outcomes
• augmentative
• alternative communication

Augmentative and alternative communication (AAC) provides individuals who are unable to effectively use their natural speech an alternative way to engage with the people in their community and influence their environment (Beukelman & Mirenda, 2020). Given that communication is a dynamic interaction, successful communication relies on both the individual using an AAC system as well as their communication partner. Individuals who are prescribed AAC often receive treatment that focuses on helping them learn to effectively communicate using their AAC system. To determine whether individuals who use AAC benefit from intervention, researchers and clinicians must document AAC communication outcomes.¹ Outcomes measurement is a critical component of evidence-based practice because it lets researchers and clinicians determine “what works, for whom, under what conditions” (Smith, 2000). Given the heterogeneity of AAC users and their communication needs, there are a wide range of communication goals that may be addressed within intervention. In order to determine what works for whom, under what conditions researchers and clinicians must use a common set of definitions. This is challenging given that the focus of AAC intervention, which is assessed on an individual basis, is influenced by the skills, goals, and needs of the individual. To date, there is no consensus on how to operationally define AAC communication outcomes. This may be due, in part, to a lack of validated data collection tools as well as the diversity of the population who uses assistive technology such as AAC (e.g., Smith, 2000). Therefore, within this study we used the communicative competence framework (Light, 1989; Light & McNaughton, 2014) to determine which areas of AAC intervention have received the greatest attention and how these outcomes are being measured.

Communication competencies as a framework for measuring AAC communication outcomes

In the field of augmentative and alternative communication (AAC), the outcomes that researchers choose to measure often differs across studies. This is because the different stakeholders in the field of AAC (e.g., AAC users, AAC device manufacturers, funding agencies, practicing speech-language pathologists) value different types of outcomes (Scherer, 2017). While a researcher may be interested in documenting a person’s ability to use their device during social settings to interact with others, a funding agency may be more interested in documenting the economic benefit of funding a device (Lenker et al., 2005; Smith, 2016). Within the field of communication, there are numerous ways in which communication outcomes may be measured including administering standardized language assessments, using surveys and questionnaires and training observers to document behaviors of interest such as turn-taking or conversational breakdowns. Identifying and/or developing a common set of definitions for describing AAC communication outcomes will enable researchers and clinicians to make comparisons across research studies and facilitate the development of measurement tools that support the evaluation of meaningful outcomes using approaches that provide useful information for all of the stakeholders, which should ultimately improve outcomes for people of all ages that need AAC to communicate.

Classification systems may offer a viable framework for enabling researchers to determine what areas of AAC intervention have received the greatest attention and how outcomes are being measured in these areas. The International Classification of Functioning, Disabilities and Health (ICF) is the leading outcome measurement for how individuals with disabilities, including people who require AAC, interact with the world (World Health Organization [WHO], 2001). The ICF framework addresses an individual’s functioning and disability within the context of their participation in the community and at home (WHO, 2001). Using communication competencies (Light, 1989; Light & McNaughton, 2014), which describes the functionality and adequacy that facilitate participation in the community and at home to characterize an individual’s outcome is in line with the ICF framework’s focus on participation in everyday life. For people who require AAC, communication competence is built upon an individual’s knowledge and skills across four domains: linguistic competence, operational competence, strategic competence, and social competence; and communication competence considers the psychosocial factors that impact effective communication (Light, 1989; Light & McNaughton, 2014).

Linguistic competence is an individual’s knowledge and understanding of the spoken and written language used in their community and home, as well as their knowledge of the linguistic code used on their AAC device (e.g., selecting the appropriate symbols to create the phrase “I like to sing”). Operational competence is the “technical operation of AAC strategies and techniques” (Light & McNaughton, 2014, p. 3). For people who use high-tech AAC systems, such as speech generating devices, this includes turning the system on and off, navigating between pages, and utilizing the appropriate selection technique (e.g., direct selection, scanning, eye gaze). Strategic competence is the ability to compensate for limitations of the device. For example, a person who requires AAC may need to explain their AAC system and how they use it to an unfamiliar communication partner. Social competence involves pragmatic (e.g., communicating for a variety of functions, turn taking, terminating topics, etc.) and sociorelational skills (i.e., “the interpersonal aspects of communication that form the foundation for developing effective relationships;” Light & McNaughton, 2014, p. 3).

The communicative competence framework acknowledges that while each of these competencies are separate, they are intertwined and impact one another. For example, an individual with high linguistic competency (e.g., creating sentences using an AAC system), but low strategic competence (e.g., the ability to clearly communicate a message when the relevant vocabulary is not found on the system) will struggle to communicate when they don’t have access to relevant vocabulary. In addition to the knowledge, judgment, and skills that communication functionality and adequacy are built upon, the communicative competence framework considers the psychosocial factors that may impact communicative competence, such as individual attitudes toward AAC, environmental support and barriers, and policy and practice support and barriers (Light & McNaughton, 2014). Consideration of the psychosocial impact is critical given that the attitudes and perceptions of the AAC user, as well as their family members, can act as a facilitator or barrier to an individual’s overall use of the AAC device (Baxter et al., 2012; Moorecroft et al., 2019). An appealing aspect of the communication competence framework is that it takes all of the skills necessary for effective communication on an AAC system into consideration.

Current study

In summary, although outcomes are a critical aspect of evidence-based practice, there is currently no consensus on how to operationally define AAC communication outcomes. Within this study, we used the communicative competence framework to determine a) which areas of AAC intervention have received the greatest attention and b) how outcomes are being measured in these areas. We determined that a scoping review was the ideal methodology as this process would let us evaluate the previous literature to assess outcome characteristics of other researchers' work. Scoping reviews, which are conducted in a systematic way using clearly defined methodological frameworks (Arksey & O’Malley, 2005; Tricco et al., 2018) and reporting guidelines (Tricco et al., 2018), are designed to provide overviews of broad topics that may involve a sizable and diverse set of literature (e.g., Arksey & O’Malley, 2005; Tricco et al., 2018). Scoping reviews are widely used in AAC research; researchers have published scoping reviews on topics ranging from the effectiveness of AAC interventions for individuals with autism (Schlosser & Koul, 2015) to the impact that AAC intervention has on receptive language abilities of children with developmental disabilities (Dada et al., 2021).

Methods

Inclusion/exclusion criteria

We included peer-reviewed research articles published between 1990 and 2021. To be included in the review the study participants needed to be using some type AAC system ranging from no-tech AAC systems like the Picture Exchange Communication System (Bondy & Frost, 1994) to speech-generating devices, there was a focus on face-to-face communication, and the study included a measure of communication outcomes. Studies were excluded if the participants were not using AAC or if they focused on manual sign, if the patients were in acute care or were in-patient, and if the study measured communication outcomes related to social media. We chose to exclude research studies focused on patients in acute care because our initial review revealed that the majority of these studies focused on patients in comas. We chose to exclude communication outcomes related to social media because of potential differences in how we communicate using social media as compared to how we communicate during face-to-face interactions.

Identifying and selecting studies

We followed the PRISMA guidelines for conducting a scoping review. A search across electronic databases was used to collect published studies. The search terms used were: Assistive Technology or AAC, and Communication, and Outcomes. Articles published between 1990 and 2021 were extracted from the following databases: Compendex, IEEE, PsycINFO, PubMed, Scopus, Web of Science. Covidence, a scoping review management software, was used to organize the studies for this review. The initial search yielded 4158 articles, which were imported into Covidence. Of the 4158 articles, 2,903 were duplicates yielding a total of 1255 unique articles to review. In the first round of screening, the title and abstract of each article was reviewed by three of the authors leaving a total of 176 articles for a full text review. During the full text review, 77 of the 176 articles were determined to meet the inclusion criteria (see Figure 1).

Figure 1. PRISMA flowchart.

Data charting

We extracted the following data from the 77 studies that met the inclusion criteria: study design, study participants, communication target of the study (e.g., language, word learning, etc.), and communication outcome measurements. Measurements unrelated to communication were not extracted from articles.

Study design

The articles were differentiated based upon the study design employed by the researcher and the definitions found in Aggarwal and Rangathan’s (2019a, 2019b) study design series were used to guide this coding. Articles were initially classified as qualitative or analytical (i.e., quantitative). Analytical studies were further classified as observational or interventional based on the presence of a treatment from the researcher.

Participant characteristics

The following demographic information was collected from each study: participant(s) age, sex, and diagnosis. For studies that used single-subject research designs if there were fewer than four participants, the research assistants listed each participant’s age. Researchers calculated the mean age when it was not specifically reported (and there were more than three participants). If specific participant age information was not provided, the information that the study provided was noted. When coding for participant diagnosis, all present diagnostic information was listed for the participants (including whether no official diagnosis was provided), participant information regarding delays/impairments were listed.

Study target

For the purpose of this review, study target was defined as the focus of the study (i.e., the facet of communication impacted by the AAC intervention that the researchers sought to evaluate). When multiple communication targets were included in the same study all targets were listed.

Communication outcomes measures

Researchers coded all of the measures that were used to assess communication outcomes. After the initial coding, outcome measures were further classified by measurement type: a) standardized assessments, b) surveys, questionnaires & structured observations, and c) non-standardized measures (i.e., trained observation or language sample analysis). Next, each measurement was classified to determine which aspect of communicative competence the outcome measure was used to assess: linguistic, social, operational, strategic, and psychosocial. When multiple outcome measurements were used within the same study, all measures were coded.

Communication competence

Interrater reliability

Sixty-five percent of the articles were screened by the first and fourth author during the initial screening (i.e., title and abstract review). There was an agreement of 87%. Disagreements were resolved with a third reviewer and during this phase the exclusionary criteria was refined to include patients in acute care or in-patient (e.g., patients with comas), which accounted for the majority of disagreements. Two undergraduate research assistants were trained on coding the articles for the full text screening and data charting. The reliability for the data charting of all variables was 96%, which was calculated based on a random 20% sample of articles. All disagreements were discussed until consensus was achieved.

Data analysis plan

First, we broadly summarize the studies included in this scoping review including the types of study designs used, ages and populations studied, and communication targets. Next, we described how AAC communication outcomes are being measured across research studies by calculating the number of studies that used a) standardized assessments, b) surveys, questionnaires, and structured observations, and c) non-standardized assessments. Finally, we classified the communication outcomes according to the communicative competence framework (Light, 1989; Light & McNaughton, 2014).

Results

Ninety-four percent of the studies (n = 72) used an analytical study design and 6% (n = 5) used a qualitative study design. Sixty-seven percent of the studies (n = 52) assessed communication outcomes in children (i.e., participants 18 years of age and under), 24% (n = 19) assessed communication outcomes in adults (i.e., participants over the age of 18 years), and 7% assessed outcomes in a sample that included adults and children (n = 6). Fifty-two percent of the studies (n = 40) had participants with a variety of diagnoses (or participants with multiple diagnoses), 35% of the studies (n = 19) focused solely on autistic participants, 4% (n = 3) focused on participants with cerebral palsy, 4% (n = 3) focused on participants with aphasia (n = 3), and 3% (n = 2) focused on participants with developmental delays. The remaining studies focused on participants with Down syndrome (n = 1), laryngectomies (n = 1), traumatic brain injury (n = 1), hypotonia (n = 1), Huntington’s disease (n = 1), amyotrophic lateral sclerosis (n = 1), and Rett’s syndrome (n = 1). Three studies did not report on the specific population studied.

Studies often had more than one communication target and the targets varied widely across research studies, with the largest number of studies targeting social communication, requesting, vocabulary and syntax. Fifty-five percent of the studies (n = 42) targeted social-communication skills, with 8% of these studies (n = 10) specifically targeted requesting, 17% of the studies (n = 13) targeted vocabulary and 14% (n = 11) targeted syntax. Other study targets included, among other things, literacy (n = 3), natural language production (n = 4), and communication modality (n = 4). See Appendix A for a detailed table of each study target.

Research Question 1: Which areas of AAC intervention have received the greatest attention (using the communicative competence framework as a structure for analysis)?

When we classified communication outcomes according to the communicative competence framework (Light, 1989; Light & McNaughton, 2014) we found that 75% (n = 60) assessed social skills, 62% of the studies (n = 48) assessed linguistic skills, 18% (n = 14) assessed psychosocial skills, 14% (n = 8) assessed operational skills, and 4% (n = 3) assessed strategic skills. Forty-nine percent (n = 38) of the studies measured one communicative competency domain, 32% (n = 25) measured two domains, 17% (n = 13) measured three domains, and 1% (n = 1) measured four domains.

Research Question 2: How are outcomes measured in those areas?

A wide variety of measures were used to assess communication outcomes across the studies. Forty-one percent of the studies (n = 32) used standardized assessments, 46% (n = 36) used surveys, questionnaires, and structured observations, and 94% used non-standardized assessments. Most standardized assessments were used in one or two studies with the exception of the Mullen Scales of Early Learning (Mullen, 1995) and Peabody Picture Vocabulary Test (Dunn et al., 1997) which were each used in 5% of the studies (ns = 5). The MacArthur Bates Communicative Developmental Inventory (Fenson, 2007; Fenson et al., 1993) and Vineland Adaptive Behavior Scales (Sparrow et al., 2005) were each used in 3% of the studies (ns = 3). A similar pattern was observed for surveys, questionnaires, and structured observations with only the Communication Matrix (n = 4; Rowland, 2011), Early Social Communication Scale (n = 3;Mundy et al., 2003), and Quebec User Evaluation of Satisfaction with Assistive Technology − 2nd edition (n = 3; Demers et al., 2002) being used in more than two studies. The two most widely used outcome measures were trained observation (i.e., training research assistants to reliably document communication outcomes of interest such as requesting, functional responses, number of communication acts, time of device use, initiations, efficiency, and communication repair) which was used in 63% of the research studies, and language sample analysis, which was used in 25% of the research studies. See Table 1 for a detailed description of each of the assessments used.

Table 1. Communication outcome assessments used across research studies.

Assessment Type	Research Study (ies)
Standardized Assessments
Autism Diagnostic Observation Measure	Lerna et al. (2014)
Clinical Evaluation of Language Fundamentals	King et al. (2020), Meinzen-Derr et al. (2021)
Communication and Symbolic Behavior Scale	Parsons et al. (2020)
Expressive One-Word Picture Vocabulary Test	King et al. (2020), Olive et al. (2008)
Gray Silent Reading Test	Lund and Light (2006)
Griffiths’ Mental Developmental Scale	Lerna et al. (2014)
MacArthur Bates Communicative Developmental Inventory	Boesch et al. (2013), King et al. (2020), Romski et al. (2008)
Mullen Scales of Early Learning	Barker et al. (2013), King et al. (2020), Parsons et al. (2020), Rose et al. (2020)
Peabody Picture Vocabulary Test-4^th Edition	Barker et al. (2013), Lund and Light (2006), Olive et al. (2008), King et al. (2020)
Preschool Language Scale	Hampton et al. (2020), King et al. (2020)
Preschool & Primary Inventory of Phonological Awareness Scale	Mathisen et al. (2009)
Receptive one-word picture vocabulary test	Olive et al. (2008)
Reynell Developmental Language Scale	Mathisen et al. (2009)
Sequenced Inventory of Communication Development	Romski et al. (2008)
Test of Auditory Comprehension of Language	Lund and Light (2006)
Vineland Adaptive Behavior Scales	King et al. (2020), Lerna et al. (2014), Thunberg et al. (2009)
Western Aphasia Battery	Johnson et al. (2008)
Surveys, Questionnaires, & Structured Observations
Structured Observation
ASHA Functional Assessment of Communication Skills	Johnson et al. (2008), Lund and Light (2006)
Clinical Assessment of Language Comprehension	Romski et al. (2008)
Early Social Communication Scale	Almirall et al. (2016), Hampton et al. (2020), Yoder and Lieberman (2010)
Pragmatic Observation Measure	Parsons et al. (2020)
Structured Play Assessment	Almirall et al. (2016)
Test of Playfulness	Parsons et al. (2020)
Caregiver observation
Diary of Communicative Development	Sevcik et al. (2004)
Survey or Questionnaire
Quebec User Evaluation of Satisfaction with Assistive Technology − 2nd edition	Chan et al. (2010), Desai et al. (2014), Borgestig et al. (2021)
Communicative Effectiveness Survey	Childes et al. (2017)
Communication Survey	Hamm and Mirenda (2006)
Health-Related Quality of Life – KIDSCREEN 10 version	Borgestig et al. (2021)
Psychosocial Impact of Assistive Devices Scale	Borgestig et al. (2021)
School Function Assessment	Desai et al. (2014)
Ages and Stages Questionnaire	Thomas-Stonell et al. (2016)
Communication Matrix	Borgestig et al. (2021), Desai et al. (2014), Mohan et al. (2019), Sreekumar et al. (2019)
Focus on Communication Outcomes Under Six	Thomas-Stonell et al. (2016)
Pragmatics Checklist	Meinzen-Derr et al. (2019)
Interviews
The Pragmatics Profile of Everyday Communication Skills in Children	King et al. (2020)
Social Network: A Communication Inventory for Individuals with Complex Communication Needs and their Communication Partners
Other
ASHA Quality of Communication Life Scale	Johnson et al. (2008)
Augmentative Communication Employment Training and Supports Survey	Bryen et al. (2004)
Communication Effectiveness Index	Johnson et al. (2008)
Effectiveness Framework of Functional Communication	Ferm et al. (2010)
Goal Attainment Scaling	Desai et al. (2014)
Quality of Life Profile: People with Sensory and Physical Disabilities	Hamm and Mirenda (2006), Lund and Light (2006)
Nonstandardized Assessments
Trained observation	Alzrayer (2020), Alzrayer et al. (2021), Babb et al. (2019, 2020, 2021), Baker et al. (2022), Bayldon et al. (2021), Bock et al. (2005), Boesch et al. (2013), Brock et al. (2017), Caron et al. (2020), Chan et al. (2010), Chapin et al. (2022), Conklin and Mayer (2011), Diehl and Wallace (2018), Falk et al. (2010), Flores et al. (2012), Ganz and Simpson (2004), Ganz et al. (2008, 2009), Gevarter, Groll, et al. (2020), Gevarter, Horan, et al. (2020), Gevarter et al. (2021), Hagan and Thompson (2014), Icht et al. (2020), Ivy et al. (2020), Johnson et al. (2008), Lancioni et al. (2009), Lerna et al. (2014), Light et al. (2021), McMillan and Renzaglia (2014), Mohan et al. (2019), Mongeau and Lussier-Desrochers (2017), Olive et al. (2008), Quinn et al. (2020), Ricci et al. (2017), Seale et al. (2020), Snodgrass et al. (2013), Son et al. (2006), Tan and Alant (2018), Therrien and Light (2016), Thunberg et al. (2007, 2009, 2011), Timpe et al. (2021), Trembath et al. (2009), Snodgrass et al. (2013), Walters et al. (2021), Wandin et al. (2021)
Language sample analysis	Almirall et al. (2016), Barker et al. (2013), Brock et al. (2017), Ganz and Simpson (2004), Griffith et al. (2014), Hampton et al. (2020), Hayes and Traughber (2021), Hux et al. (2010), Kasari et al. (2014), Light et al. (2021), Lund and Light (2006, 2006), Mathisen et al. (2009), Meinzen-Derr et al., (2017, 2019, 2021), Mooney et al. (2018), Romski et al. (2008), Senner et al. (2019), Sevcik et al. (2004), Soto et al. (2020)
Semi-structured interviews/Quesstionnaires	Clarke et al. (2011), Griffith et al. (2014), Hagan and Thompson (2014), Hampton et al. (2020), Keskinen et al. (2012), Timpe et al. (2021)

Owing to the large variability in the types of outcomes assessed across research studies, it was not possible to summarize how each type of outcome was assessed. Therefore, we focused on the most frequently assessed outcomes. In the social competency domain, researchers most frequently assessed requesting (n = 16), turn-taking (n = 13), initiations (n = 5), and joint attention (n = 4). All of the research studies assessing requesting used observation (Alzrayer, 2020; Alzrayer et al., 2021; Bock et al., 2005; Boesch et al., 2013; Flores et al., 2012; Ganz et al., 2009; Gevarter, Horan, et al., 2020; Gevarter et al., 2021; Ivy et al., 2020; Kasari et al., 2014; Lancioni et al., 2009; Lerna et al., 2014; Olive et al., 2008; Ricci et al., 2017; Snodgrass et al., 2013; Son et al., 2006). Language sampling and observation were used to assess turn-taking; five studies used observation (Bayldon et al., 2021; Chapin et al., 2022; Therrien & Light, 2016; Thunberg et al., 2007; Timpe et al., 2021), and eight studies used language sample analysis (Borgestig et al., 2021; Hux et al., 2010; Light et al., 2021; Lund & Light, 2006; Mathisen et al., 2009; Meinzen-Derr et al., 2017, 2019, 2021). With regard to initiations, four studies used observation (Bayldon et al., 2021; Conklin & Mayer, 2011; Lerna et al., 2014; McMillan & Renzaglia, 2014) and one study used language sample analysis (Hux et al., 2010). Two of the four studies assessed joint attention using the Early Social Communication Scale (Almirall et al., 2016; Hampton et al., 2020), one using the Clinical Assessment of Language Comprehension (Romski et al., 2008), and one using observation (Lerna et al., 2014).

In the linguistic domain, researchers tended to focus on general language skills (n = 14), semantics, and syntax. Eleven of the 14 studies that assessed general language skills used standardized assessments, such as the Mullen, Preschool Language Scale, and Clinical Evaluation of Language Fundamentals (Barker et al., 2013; Boesch et al., 2013; Hampton et al., 2020; King et al., 2020; Lerna et al., 2014; Lund & Light, 2006; Mathisen et al., 2009; Meinzen-Derr et al., 2021; Parsons et al., 2020; Romski et al., 2008; Rose et al., 2020). The other three studies used the Communication Matrix (Borgestig et al., 2021), a semi-structured interview (Hagan & Thompson, 2014), and the Focus on Communication Outcomes Under Six (Thomas-Stonell et al., 2016). The most widely assessed area of syntax was mean length of utterance. All six studies that assessed mean length of utterance used language sample analysis (Hayes & Traughber, 2021; Kasari et al., 2014; Light et al., 2021; Mathisen et al., 2009; Meinzen-Derr et al., 2017, 2019, 2021). Kasari et al. (2014) used language sample analysis to assess the number of unique word combinations and Mathisen et al. (2009) used language sample analysis to assess production of one-word utterances. Johnson et al. (2008) used observation to assess sentence production and Mongeau and Lussier-Desrochers (2017) used observation to assess production and comprehension of word combinations and sentences.

With regard to semantics, eight studies used observation to assess the number of target words identified (Caron et al., 2020; Trembath et al., 2009), number of functional vocabulary words (Gevarter, Horan, et al., 2020), number of social/action words (Gevarter, Groll, et al., 2020), correct/incorrect use of pronouns (Olive et al., 2008), words per minute (Babb et al., 2021; Wandin et al., 2021), intelligible words/word approximations (Ganz et al., 2008), symbol production (Johnson et al., 2008), and generalization of symbols to novel stimuli (Snodgrass et al., 2013). Ten studies used language sample analysis to assess expressive vocabulary (Ganz & Simpson, 2004), the number of different words (Almirall et al., 2016; Barker et al., 2013; Kasari et al., 2014; Meinzen-Derr et al., 2017, 2019, 2021), the total number of words (Meinzen-Derr et al., 2017, 2019; Sevcik et al., 2004), type-token ratio (Mathisen et al., 2009), and intelligible words (Ganz & Simpson, 2004; Mathisen et al., 2009). One study used the Diary of Communicative Development to assess vocabulary (Sevcik et al., 2004), and two studies used standardized assessments, such as the Peabody Picture Vocabulary Test to assess vocabulary (Lund & Light, 2006; Olive et al., 2008). Finally, three of the six studies assessing communication effectiveness used observation (Thunberg et al., 2007, 2009, 2011). Childes et al. (2017) used the Communicative Effectiveness Survey, Ferm et al. (2010) used the Effectiveness Framework of Functional Communication and Johnson et al. (2008) used the Communicative Effectiveness Index.

Discussion

In this scoping review, we used the communicative competence framework to gain insight into which areas of AAC intervention have received the greatest attention and how outcomes are being measured in these areas. When outcomes were categorized according to the communicative competence framework over half of the research studies assessed participants’ linguistic skills and social skills. However, only a limited number of studies assessed participants operational (n = 8), strategic competence (n = 3) or psychosocial factors (n = 14). This gap in research would not have been apparent if we had not used the communicative competence framework for considering outcomes measurements. As noted within the communicative competence framework, in order to effectively communicate using an AAC system, individuals must not only understand the linguistic system of their culture and AAC device as well as the social aspects of communication, they must also be able to effectively operate the system and compensate for limitations of an AAC system. In addition, it is critical to understand the psychosocial impact that the device has on the AAC user and their quality of life.

Assessing operational and strategic competency, as well as the psychosocial impact that an AAC device has, is especially important since these factors often appear to be barriers for AAC use. For example, (Moorecroft et al. 2019 systematic review of barriers and facilitators to the provision and use of low-tech and unaided AAC systems for people with complex communication needs and their families found the most prominent barriers included the attitudes of professionals, family members and the AAC user. This is consistent with Baxter et al. (2012) systematic review of the barriers and facilitators to the use of high-tech AAC devices, which revealed that family attitudes and perceptions served as a barrier or facilitator of AAC use. These findings highlight the need for researchers to consider all aspects of communication competence, since AAC outcomes need to consider not only the ability of an individual to effectively communicate using a system but also their ongoing use and satisfaction with the system in place. If an individual abandons their system, they will no longer have a system that will enable them to communicate and, ultimately, participate in their community.

The two most widely used outcome measures were trained observation (used in 63% of the research studies) and language sample analysis (used in 25% of the research studies). The widespread use of language sample analysis and trained observation to assess a variety of communication outcomes is consistent with recent discussions around the paucity of standardized measures available to assess communication outcomes in AAC users (e.g., Brady et al., 2013; Broomfield et al., 2019; Enderby, 2014; Light & Mcnaughton, 2015). The large number of studies that relied on trained observation protocols developed by researchers demonstrates the innovative approach that researchers have taken to enable them to document the communicative outcomes of interest in a systematic way. While useful, a drawback to this approach is that the operational definitions used across trained observations may differ making comparisons across research studies challenging. Another drawback to this approach is moving these measures into clinical practice. Once research determines that the process was beneficial to learn about communication outcomes, that needs to be translated into clinical practice. The wide use of language sample analysis aligns with recent publications highlighting the clinical utility of language sample analysis for assessing language outcomes in individuals who use AAC for years (e.g., Kovacs & Hill, 2015; Mooney et al., 2021; Van Tatenhove, 2014). However, researchers continue to differ in how language sampling is conducted due to variable configurations across speech-generating devices. As noted by Mooney et al. (2021) “vocabulary configurations on a speech-generating device may or may not have the option to insert bound morphemes, making it challenging to determine how to consider one’s use of, or lack of, bound morphemes” (p. 49). A common set of assessment tools would facilitate consistency in terms of what aspects of communication are being measured as well as how they are measured, enabling comparison across research studies. However, the heterogeneity of individuals who use AAC, which ranges from nonspeaking autistic children who are in the process of developing language to adults with ALS who are losing the ability to communicate verbally, makes developing a common set of tools especially challenging.

Although none of the standardized measures used across research studies were specific to AAC users, 11% (n = 9) of the research studies used surveys and questionnaires specifically developed for AAC users. The Communication Matrix, which was used in four studies, is a “communication skills assessment” designed for individuals who use any type of communication behavior including alternative forms. The matrix focuses on two major aspects of communication – why an individual is communicating and how they are communicating (Rowland, 2011). Unlike the Communication Matrix, the Psychosocial Impact of Assistive Devices Scale, which was used in one study, and the QUEST-2, which was used in three studies, assesses the impact of the device on an individual’s life and the patient’s satisfaction with the device. Finally, the Augmentative Communication Employment Training and Supports Survey was used in one study to assess employment outcomes.

Taken together, these findings indicate that researchers are relying on non-standardized measures to assess a wide variety of language skills including pragmatics, semantics, and syntax communication outcomes in people who use AAC. Using the communicative competence framework revealed patterns that were not apparent when we approached communication outcomes from the viewpoint of research question 1 (i.e., how outcomes are being assessed).

Limitations

While the information obtained from this scoping review provides important insight into how AAC outcomes are being measured across research studies, there are limitations to this study. The first being in how we chose to categorize AAC outcome measures. It is possible that using a different categorization method for research question 1 may have enabled the study team to gain greater insight from the data obtained. Future research should work to tie together the classification systems such as communicative competence with the International Classification of Functioning, Disabilities and Health (ICF), which addresses an individual’s functioning and disability within the context of their participation in the community and at home (WHO, 2001). Using a combination of these systems to examine AAC communication outcomes would enable researchers to capture how communication targets such as employment outcomes related to participation in the community and at home. Another limitation is the broad view of this study. While this view was beneficial to lay the foundation for the need for better clarification and definition of AAC outcome measures, a focus within one diagnostic population may have yielded insight into the benefits and drawbacks of particular measures for that population. This study also did not review each of the measures used to determine if they consider various aspects of the communicative competence framework. Finally, though we extracted data from 77 studies, we may have missed some relevant papers as a result of our search criteria and inclusion/exclusion criteria. We may have identified additional insights if we had searched additional databases (e.g., Google Scholar), or included gray literature (e.g., peer reviewed abstracts, conference proceedings, magazine articles). Broadening the databases and the inclusion criteria may have increased the breadth of outcome measures addressing AAC communication in contemporary research studies.

Conclusion

In conclusion, the findings from this scoping review highlight the wide number of studies assessing linguistic and social outcomes in people who use AAC. However, the use of communicative competence framework revealed significant gaps in research. Given that operational and strategic competence as well psychosocial factors impact AAC use, we believe that it is critical for researchers to begin to assess all measures of communicative competence in order to provide stakeholders with a comprehensive understanding of the impact of AAC on individuals’ communication.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Given the complex nature of communication interactions, we chose to focus this scoping review on the measurement of communication outcomes and development of the AAC user’s communication competency rather than the communication partner.

Appendix A

Table

Author (year)	Study design	Age group/sex	Participant Characteristics	Study Target	Communication Competence Classification	Outcome (Measurement tool)
Almirall et al. (2016)	analytical, interventional, randomized control trials	age range = 5–8 years of age/10 females, 51 males	autism	social-comm.	linguistic	total social communicative utterances (language sample) number of direct root words (language sample)
					social	initiating joint attention (ESCS) initiating behavior regulation (ESCS) unique play acts (Structured Play Assessment)
Alzrayer et al. (2021)	analytical, interventional pre-post studies	age range = 4–5 years/1 female, 3 males	autism	requesting	social	number of requests per minute (observation)
Alzrayer (2020)	analytical, interventional pre-post studies	mean age = 3 years, 8 months (range = 3–5 years)/1 female, 3 males	autism	requesting	social	PECS based requests (observation) SGD based requests (observation)
Babb et al. (2019)	analytical, interventional, pre-post studies (longitudinal study)	18 years/1 male	autism	comm.	social	number of communication acts during vocational tasks (observation)
Babb et al. (2020)	analytical, interventional pre-post studies	mean age = 17 years, 3 months (range = 14–20 years)/4 males	Down syndrome, autism	task analysis	social	percent of task analysis completed individually (observation)
Babb et al. (2021)	analytical, interventional pre-post studies	age = 19 years/1 male	Down syndrome	language output	linguistic	number of intelligible words per minute (observation) average number of intelligible words per minute (observation)
Baker et al. (2021)	analytical, observational, cohort	mean age = 7 years, 7 months (range = 8 years–10 years, 4 months)/8 females, 6 males	autism	device access	social	time of device use (observation)
Barker et al. (2013)	analytical, interventional, pre-post studies (longitudinal study)	mean = 4 years, 1 month (range = 36–71 months)/17 females, 66 males	autism, down syndrome, global developmental delay, spina bifida, cerebral palsy, TBI, other disorders	language growth	linguistic	receptive language (Mullen & Peabody Picture Vocabulary Test-4^th Ed.) expressive language (Mullen) number of different words (language sample)
Bayldon et al. (2021)	analytical, interventional pre-post studies	ages = 8 years, 2 months, 8 years, 11 month/1 female, 1 male	cerebral palsy	pragmatics	social	Initiations (observation) turn taking (observation)
Bock et al. (2005)	analytical, interventional, cross-over (alternating)	4-year-olds/6 males	non-verbal with developmental delay	acquisition and generalization of requesting	social	correct response (observation)
Boesch et al. (2013)	analytical, interventional, cross-over	ages = 6 years, 7 years, 10 years/1 female, 2 males	autism	social-comm. & natural speech production	linguistic	expressive language (MCDI) receptive language (MCDI)
					social	total number of social communicative behaviors (observation) requesting (observation) gesture use (MCDI)
Borgestig et al. (2021)	analytical, interventional pre-post studies	mean age = 11 years, 4 months (range = 3–26 years)/10 females, 7 males	Cerebral palsy, Rett syndrome, metabolic disorder	comm.	linguistic	expressive comm. (Communication Matrix)
					psychosocial	quality of life (KIDSCREEN-10) functional independence with and without device (Psychosocial Impact of Assistive Devices Scale) Device satisfaction (QUEST 2.0) Communication Matrix
					operational	question response accuracy designed to assess navigational errors (observation)
					psychosocial	instances of frustration (observation)
					social	total conversation time (language sample analysis) number of conversational turns (language sample analysis) number of discourse units (language sample analysis)
Bryen et al. (2004)	analytical, interventional, pre-post studies	mean age = 30 years, 6 months (range = 24–37 years)/1 female, 5 males	Not reported	employment	linguistic	employment communication skills (ACETS Employment Survey)
					social	employment communication skills (ACETS Employment Survey)
Caron et al. (2020)	analytical, interventional pre-post studies	mean age = 12 years, 2 months (range = 7 years, 5 months–20 years, 7 months/1 female, 4 males	Down syndrome, Intellectual and developmental disability	vocabulary	linguistic	number of target words identified (observation)
Chan et al. (2010)	analytical, interventional, cross-over studies	19 years/1 male	hypotonia	utterances	operational	types of errors (observation) speed (observation)
					psychosocial	level of fatigue (questions adapted from Borg scale) satisfaction with device (QUEST 2.0)
Chapin et al. (2022)	analytical, interventional, pre-post studies	age range = 3 years, 11 months, 4 years, 11 months, 5 years, 6 months/1 female, 2 males	autism	turn-taking	social	number of communicative turns taken (observation)
Childes et al. (2017)	analytical, interventional, nonrandomized controlled trials	mean age = 64 years (range = 47–81 years)/12 females, 22 males	laryngectomy	AT use	psychosocial	positive and negative experiences with the device (self-report)
					social	communication effectiveness score across a range of contexts (Communicative Effectiveness Survey)
					strategic	number of communication repair strategies (self-report)
Clarke et al. (2011)	analytical, interventional, pre-post studies	mean age = 11 years, 4 months (range = 7–16 years)/3 females, 7 males	Cerebral palsy, Trisomy 8, Worester Drought syndrome, dyspraxia	children’s view on short-term impact of new comm. aid	psychosocial	self-efficacy (self-report) self-esteem (self-report) other’s attitudes (self-report)
Conklin and Mayer (2011)	analytical, interventional, pre-post studies (longitudinal study)	Ages = 23 years, 38 years, 51 years/2 females, 1 male	severe mental retardation, Cerebral palsy, Down syndrome	PECS use	social	percentage of independent initiations (observation) percent of off-task behavior (observation)
Desai et al. (2014)	analytical, interventional, pre-post studies (longitudinal study)	13 years/1 male	Cerebral palsy and autism	comm. and participation in school activities	linguistic	communication modality (Communication Matrix) attainment of communication goals identified by SLP (Goal Attainment Scaling)
					psychosocial	teacher satisfaction with device (QUEST 2.0)
					social	participation in non-academic and social aspects of school (School Function Assessment) communication function (Communication Matrix)
Diehl and Wallace (2018)	analytical, interventional, pre-post studies (longitudinal study)	ages = 21 years, 62 years/2 males	TBI	comm. modalities and breakdowns	Linguistic social	number of modalities used to produce target words (observation) communication repair (observation)
Falk et al. (2010)	analytical, interventional, pre-post studies	2 adults (age and gender not reported), 1 child (age and gender not reported), 1 adolescent (age and gender not reported)	adults had no diagnosis, child and adolescent had hypotonic Cerebral palsy and have severe phonatory impairment	comparing performance of prototype device to commercial devices	linguistic operational	percentage of correctly identified vocalizations by the system (observation) percentage of correctly rejected user-generated artifacts (observation) average task time (observation)
					psychosocial	Fatigue (observation and self-report)
Ferm et al. (2010)	analytical, interventional, nonrandomized controlled trials	mean age = 61 years (range = 51–71 years)/2 females, 3 males	Huntington’s disease	comm. effectiveness	linguistic	participant’s understanding of options presented (Effectiveness Framework of Functional Communication)
					social	comm. effectiveness (Effectiveness Framework of Functional Communication) comm. duration (observation)
Flores et al. (2012)	qualitative, case series	mean age = 9 years (range = 8–11 years)/5 males	autism, Intellectual disability	Utility of iPAd for making requests	social	number of requests (observation)
Ganz and Simpson (2004)	analytical, interventional, pre-post	Ages = 5 years, 8 months, 7 years, 2 months, 3 years, 9 months years/1 female, 2 males	autism, developmental delay	evaluate the utility of PECS to increase the number of words spoken and complexity of word utterances	linguistic	average number of intelligible words per trial (language sample analysis) percentage of trials in which non-vocalizations occurred (language sample analysis) syntax (language sample analysis) expressive vocabulary (language sample analysis)
					operational	participants proficiency using PECS (observation)
Ganz et al. (2008)	analytical, interventional, pre-post studies	ages = 4 years, 5 months, 3 years, 1 month, 5 years, 1 month/1 female, 2 males	autism, developmental delay	impact of PECS on comm.	linguistic operational	intelligible words (observation) word approximations (observation) independent and prompted exchanges to request (observation)
					social	independent exchange to request (observation)
Ganz et al. (2009)	analytical, interventional, pre-post studies	ages = 3 years, 2 months, 6 years, 5 years, 6 months/3 males	autism	impact of PECS on comm.	social	number of times spoken words used to request (observation) number of times pictures used to request (observation) maladaptive behaviors (observation)
Gevarter, Horan, et al. (2020)	analytical, interventional, cross-over	mean age = 3 years, 4 months (range = 3 years–5 years)/1 female, 4 males	autism	vocabulary	linguistic	number of functional vocabulary words (observation)
Gevarter, Groll, et al. (2020)	analytical, interventional, pre-post studies	mean age = 4 years, 2 months (range = 3 years,1 month–5 years, 7 months)/6 males	autism, deaf	social language production	social linguistic	Number of requests (observation) Number of action/social words (observation)
Gevarter et al. (2021)	analytical, interventional, pre-post studies	ages = 3 years, 8 months, 4 years, 5 months, 4 years, 10 months/3 males	autism	requesting and responding	social	item requests (observation) functional responses (observation)
Griffith et al. (2014)	analytical, interventional, case series	mean age = 58 years (range = 42 years–70 years)/3 females, 1 male	Broca’s aphasia	how the type of visual supports and the presence of linguistic supports influence the comm.	linguistic psychosocial strategic	expressive modality units (language sample analysis) perceived helpfulness of AT support (informal interviews – self report) percentage of repaired trouble sources (language sample analysis)
Hagan and Thompson (2014)	qualitative, interventional, case report	40 years/1 female	ADHD, Intellectual disability, nonverbal	impact of using an SGD on the comm. abilities and pragmatic skills	linguistic	comm. modality (semi-structured interview) expressive language (semi-structured interview) receptive language (semi-structured interview)
					social	conversation skills (semi-structured interview) community participation (semi-structured interview)
					strategic	repairs communication topic (observation)
Hampton et al. (2020)	analytical, interventional, randomized controlled trials	mean age = 43 months (range = 36–55 months)/15 females, 53 males	autism	language production and pragmatics,	linguistic	expressive language (Preschool Language Scale-5^th edition) receptive language (Preschool Language Scale-5^th edition)
					social	joint attention (Early Social Communication Scale) social communicative utterances (language sample analysis)
Hamm and Mirenda (2006)	qualitative, case series	mean age = 28 years, 8 months (range = 19 years, 9 months −24 years, 9 months)/2 females, 6 males	autism, Intellectual disability, Cerebral palsy, Rett syndrome, deaf/blindness	post-school outcomes of individuals with complex comm. needs	linguistic psychosocial social	communication modality (Communication Survey) quality of life (Quality of Life Profile: People with Sensory and Physical Disabilities) quality of communication (Communication survey)
Hayes and Traughber (2021)	analytical, interventional pre-post studies	19 school age children/not reported	Not reported	language output	linguistic	mlu (language sample analysis) total utterances (language sample analysis)
Hux et al. (2010)	analytical, interventional, cross-over	61 years/1 male	1 participant with aphasia (9 neurotypical participants served as communication partners)	conversational skills	linguistic social	conceptual complexity level (language sample analysis) number conversational turns (language sample analysis) number utterances serving as initiations and responses (language sample analysis) number of content units successfully transmitted during conversational exchange (language sample analysis)
Icht et al. (2020)	analytical, interventional, non-randomized controlled trials	mean age = 20 years (range = 9–32 years)/4 female, 8 males	SMA, cerebral palsy, Rubinstein-Taybi syndrome, Pelizaeus-Merzbacher syndrome	memory	linguistic	hit rates (observation)
Ivy et al. (2020)	analytical, interventional pre-post studies	ages = 4 years, 5 months, 5 years, 3 months, 7 years, 2 months/1 female, 2 males	Baraitser-Winter syndrome, developmental delay, hearing loss, cytomegalovirus, Lennox-Gastaut syndrome, hydrocephalus, movement disorder	independent exchanges and operational skills	social operational	independent requests (observation) percentage of trials with independent completions of three phases pickup, reach, release (observation)
Johnson et al. (2008)	analytical, interventional, pre-post studies	ages = 57 years, 67 years, 77 years/2 females, 1 male	aphasia and apraxia of Speech	functional comm. skills	linguistic	symbol identification (observation) symbol production (observation) sentence production (observation) comprehension (Western Aphasia Battery) repetition (Western Aphasia Battery) naming (Western Aphasia Battery) reading/writing (Western Aphasia Battery) spontaneous speech (Western Aphasia Battery)
					psychosocial	impact of communication disorder from perspective of communicatively-impaired person (American Speech-Language Hearing Association Quality of Communication Life Scale)
					social	everyday comm. activities (American Speech-Language Hearing Association Functional Assessment of Communication skills) comm. effectiveness (Communicative Effectiveness Index)
Kasari et al. (2014)	analytical, interventional, pre-post study, randomized control, longitudinal Study	mean age = 6 years,3 months(range = 5 years–8 years)/10 females, 51 males	autism	spontaneous comm.	linguistic	number of unique word combinations (language sample) total number of different root words mlu (language sample)
					social	spontaneous communicative utterances (language sample) socially communicative utterances (language sample) comments (language sample) requests (language sample) protests (language sample)
Keskinen et al. (2012)	analytical, interventional, pre-post studies	mean age = 26 years (range = 14 years–37 years)/9 males	autism. speech, physical, intellectual and hearing disability, visual impairment, behavior disorders	comm.	psychosocial	user experience with the device (interview)
King et al. (2020)	analytical, observational, longitudinal	15 years/1male	apraxia, hearing loss, dysarthria	language and functioning	linguistic	Receptive/expressive language (Clinical Evaluation of Language Fundamentals, Preschool Language Scale) receptive language (Clinical Evaluation of Language Fundamentals, Preschool Language Scale) Receptive/expressive vocabulary (Expressive One-Word Picture Vocabulary Test) Receptive/expressive vocabulary (Expressive One-Word Picture Vocabulary Test, Expressive Vocabulary Test, MCDI) receptive vocabulary MCDI, Peabody Picture Vocabulary Test
					social	adaptive functioning (Vineland, Social Network: A Communication Inventory for Individuals with Complex Communication Needs and their Communication Partners
Lancioni et al. (2009)	analytical, interventional, cross-over studies	mean age = 10 years,7 months (range = 5 years–18 years)/gender not reported	Encephalopathy, blindness, severe visual impairment, minimal residual vision, total blindness, typical visual functioning	requesting	social	number of requests (observation)
Lerna et al. (2014)	analytical, interventional, nonrandomized control trials, pre-post studies	age range = 18–60 months/gender not specified	autism	social-comm.	linguistic	expressive language (Griffiths’ Mental Developmental Scale) receptive language (Griffiths’ Mental Developmental Scale) comm. (Vineland & ADOS)
					social	personal/social (Griffiths’ Mental Developmental Scale) reciprocal social interaction (ADOS) social skills (Vineland) during unstructured free-play with examiner the following were measured: cooperative play, eye contact, joint attention, verbal requests, nonverbal requests, initiations
Light et al. (2021)	analytical, observational, case study	38 months/1 female	developmental delays, strabismus intermittent esotropia, rotational nystagmus	language production and pragmatics	linguistic	mean length of message (language sample analysis) number of letter correspondences acquired (observation) sound blending (observation)
					social	turns per minute (language sample analysis)
Lund and Light (2006)	analytical, observational, case report	mean age = 21 years, 7 months (range = 19 years–23 years)/7 males	Cerebral palsy	language, reading, pragmatics and educational and vocational achieve	linguistic	receptive language (Peabody Picture Vocabulary Test-Revised & Test of Auditory Comprehension of Language) reading Comprehension (Gray Silent Reading Test) linguistic complexity (language sample)
					psychosocial	vocational goals (interview) self-determination (Arc’s Self-Determination Scale) quality of life (Quality of Life Profile for People with Physical and Sensory Disabilities)
					social	communicative Interaction: turn reciprocity (language sample) fulfilled turn opportunities (language sample) percentage of turns containing more than one concept (language sample) number of concepts on multi-concept turns (language sample) functional communication-base needs (Functional Assessment of Communication Skills for Adults) functional communication – social (Functional Assessment of Communication Skills for Adults) adequacy of communication (Functional Assessment of Communication Skills for Adults) appropriateness of communication (Functional Assessment of Communication Skills for Adults) communication sharing (Functional Assessment of Communication Skills for Adults) educational Placement (level of participation) (Functional Assessment of Communication Skills for Adults)
Mathisen et al. (2009)	analytical, interventional, pre-post studies	3 years, 10 months/1 female	Cerebral palsy	emergent literacy and language	linguistic	receptive language (Reynell Developmental Language Scale) communication form (The Pragmatics Profile of Everyday Communication Skills in Children) emergent literacy skills (The Preschool and Primary Inventory of Phonological Awareness & The Sheffield Early literacy development profile) communication form (The Pragmatics Profile of Everyday Communication Skills in Children) response to communicative input (The Pragmatics Profile of Everyday Communication Skills in Children) verbal utterances (language sample analysis) non-verbal utterances (language sample) response questions (language sample analysis) 1-word utterances (language sample analysis) intelligible words (language sample analysis) mlu (language sample analysis) Brown’s stages (language sample analysis) type-token ratio (language sample analysis)
					Social	communication function (The Pragmatics Profile of Everyday Communication Skills in Children) interaction and conversation The Pragmatics Profile of Everyday Communication Skills in Children) contextual variation The Pragmatics Profile of Everyday Communication Skills in Children) mean turn length (language sample analysis)
McMillan and Renzaglia (2014)	analytical, interventional, pre-post studies	mean age = 9 years, 7 months (range = 8 years–11 years)/4 males	Intellectual disability, autism	frequency of device use	social	number of independent initiations (observation) number of independent responses (observation)
Meinzen-Derr et al. (2017)	analytical, interventional, pre-post studies	mean age = 7 years, 3 months (range = 5 years 4 months–9 years, 9 months)/5 females	deaf, hard of hearing	comm. competence	linguistic	mlu (language sample) number of different words (language sample) number of total words (language sample)
					social	mean turn length (language sample)
Meinzen-Derr et al. (2019)	analytical, interventional, pre-post studies	mean age = 5 years, 7 months (range = 3 years–10 years)/5 females, 6 males	bilateral hearing loss, autism	spoken language	linguistic	mlu (language sample) number of different words (language sample) number of total words (language sample)
					social	mean turn length (language sample) pragmatics (Pragmatics Checklist)
Meinzen-Derr et al. (2021)	analytical, interventional, randomized control studies	mean age = 6 years, 2 months/20 females, 21 males	Deaf and language deficits	Syntax, Semantics, Language production	linguistic	expressive language (Clinical Evaluation of Language Fundamentals 5^th edition) receptive language (Clinical Evaluation of Language Fundamentals 5^th edition) mlu (language sample analysis) number of different words (language sample analysis)
					social	mean turn length (language sample analysis)
Mohan et al. (2019)	qualitative, case report	4 years/1 male	autism	language skills	linguistic	ask questions (observation) answer questions (observation) proper punctuation in written language (observation) adaptive literacy skills (observation)
					operational	typing speed (observation)
					social	refuse (Communication Matrix) obtain (Communication Matrix) social (Communication Matrix) information (Communication Matrix) use of polite forms (observation)
Mongeau and Lussier-Desrochers (2017)	qualitative, case reports	mean age = 36 years, 8 months (range = 23–54 years)/4 females, 3 males	Intellectual disability, ataxia, Cerebral palsy	comm. skills and adaptive behaviors	linguistic	comprehend word combinations (observation) comprehend sentences [(observation) forming word combinations (observation) forming sentences (observation)
Mooney et al. (2018)	analytical, interventional, cross-over studies	mean age = 68 years (range = 58–80 years)/3 females, 3 males	primary progressive aphasia	lexical retrieval	linguistic	word production (story retell)
Olive et al. (2008)	analytical, interventional, pre-post studies	4 years/1 female	autism	requesting and general language	linguistic	expressive language (Expressive one-word picture vocabulary test-R & Expressive Vocabulary test) receptive language (Receptive one-word picture vocabulary test & Peabody Picture vocabulary test) Correct/Incorrect pronoun use (observation)
					social	challenging behavior (observation) requesting attention (observation)
Parsons et al. (2020)	analytical, interventional pre-post studies	mean age = 79 years/4 females, 11 males	autism	language, device use, and pragmatics	linguistic	expressive language (Mullen) receptive language (Mullen) speech (Communication and Symbolic Behavior Scale)
					social	pragmatics (Pragmatic Observation Measure) social skills (Communication and Symbolic Behavior Scale) play skills (Test of Playfulness)
						rate of symbolic communication acts (observation)
Ricci et al. (2017)	analytical, interventional, pre-post studies	mean age = 20 years, 6 months (range = 11 years–60 years)/gender not reported	Intellectual disability, blindness, deafness	requesting	social	frequency and types of requests (observation)
Romski et al. (2008)	analytical, interventional, pre-post study, longitudinal study	40 months/1 male	Cerebral palsy, intellectual disability, speech delay	expressive and receptive comm. gains	linguistic	vocalizations (language sample) intentional symbol use (language sample) receptive language (Sequenced Inventory of Communication Development & MCDI) expressive language (Sequenced Inventory of Communication Development & MCDI) emerging language skills (Clinical Assessment of Language Comprehension)
					social	joint attention (Clinical Assessment of Language Comprehension)
Rose et al. (2020)	analytical, interventional pre-post studies	mean age = 46.88 months (range = 31–67 months)/11 females, 37 males	autism	expressive language	linguistic	Expressive language (Mullen)
Seale et al. (2020)	analytical, interventional, non-randomized controlled trials	age range = 38 years–77 years/25 females, 15 males	ALS	communicative contexts	social	talk time (observation) presentation rate (observation) acceptance rates (observation) main talk rates (observation) repair talk rates (observation) symmetry (observation) task completion (observation)
Senner et al. (2019)	analytical, interventional, pre-post study	mean age = 7 years (range = 5 years–9 years, 6 months)/1 female, 3 males	autism, dysgenesis of the corpus callosum, tracheostomy, vocal fold paralysis, developmental delay	vocabulary	linguistic	unique words (language sample analysis)
Sevcik et al. (2004)	analytical, observational, case report	4 years/1 male	generalized developmental delay, seizure disorder	comm., engagement and utterance attempts	linguistic	vocabulary (Diary of Communicative Development) communicative use (Diary of Communicative Development) child utterance attempts (language sample analysis)
					social	engagement (language sample analysis) nonverbal communication (language sample analysis)
Snodgrass et al. (2013)	analytical, interventional, pre-post studies	9 years/1 male	Down syndrome, deaf-blindness, and severe intellectual disability.	symbol learning	linguistic social operational	generalization of symbol to novel stimuli (observation) requests (observation) reaching for symbol rather than engaging in non-communicative behavior (observation)
Son et al. (2006)	analytical, interventional, pre-post studies	ages = 5 years, 5 months, 3 years, 8 months, 3 years/2 females, 1 male	autism, developmental disorder, pervasive developmental disorder	acquisition of requesting	social	requesting (observation)
Soto et al. (2020)	analytical, interventional pre-post studies	mean age = 9 years, 9 months age (range = 8 years, 8 month–13 years, 9 months)/3 females, 5 males	Cerebral palsy, Pfeiffer syndrome, apraxia	conversational repair	social	recast/repair (language sample analysis)
Sreekumar et al. (2019)	analytical, interventional pre-post studies	7 years/1 female	spastic quadriplegic cerebral palsy, hearing loss	comm. skills	social	percentage of skill attainment (Communication Matrix)
Tan and Alant (2018)	analytical, interventional, pre-post studies	7 years/2 males	neurotypical communication partner, autism and speech disorder	comm.	social	communicative acts per minute (observation) relevance of communicative acts (observation)
Therrien and Light (2016)	analytical, interventional, pre-post studies	ages 4 years, 2 months, 4 years, 10 months/2 males	Prader Willi syndrome, developmental delay, 22q11.2 deletion syndrome, apraxia	social interaction	social	frequency in turn taking (observation)
Thomas-Stonell et al. (2016)	analytical, observational, case Reports	mean age = 2 years, 6 months age (range = 1 year, 4 months −4 years, 9 months)/1 female, 7 males	autism, oral motor/dyspraxia, cerebral palsy, downs syndrome, developmental delay, premature	comm.	linguistic	expressive language (Focus on Communication Outcomes Under Six) receptive language (Focus on Communication Outcomes Under Six) intelligibility (Focus on Communication Outcomes Under Six)
					psychosocial	confidence (Focus on Communication Outcomes Under Six)
					social	comm. (Focus on Communication Outcomes Under Six; Ages and Stages Questionnaire – Social/Emotional subtest) participation (Focus on Communication Outcomes Under Six)
Thunberg et al. (2007)	analytical, interventional, case control studies	mean age = 6 years (range = 5 years–7 years)/4 males	autism, oral motor/dyspraxia, Cerebral palsy, Down syndrome, developmental delay, premature	comm.	linguistic	mode of communication (observation)
					social	degree of engagement (observation) role in turn-taking (observation) comm. function (observation) comm. effectiveness (observation)
Thunberg et al. (2009)	analytical, interventional, pre-post studies	ages = 4 years, 11 months, 5 years, 6 months, 7 years, 6 months/3 males	autism, pervasive developmental disorder- not otherwise specified	comm.	linguistic social	communication mode (observation) communication (Vineland) comm. (Vineland) engagement (observation) rate of communication role in turn] (observation) comm. effectiveness (observation)
Thunberg et al. (2011)	analytical, interventional, case study, pre-post study	7 years/1 male	autism and learning disability	comm. across contexts	linguistic social	mode of communication (observation) degree of engagement (observation) role in turn-taking (observation) comm. function (observation) comm. effectiveness (observation)
Timpe et al. (2021)	analytical, interventional pre-post studies	ages = 3 years, 4 years, 5 months, 5 years 6 months/1 female, 2 males	Down syndrome, autism	turn taking and strategy use	linguistic psychosocial social	novel semantic concepts rates (observation) social validation (questionnaire) turn taking rates (observation)
Trembath et al. (2009)?	analytical, interventional, randomized controlled trials	age range = 3–5 years/(comm. partners 3 females, 3 males), 3 males (with autism)	3 (with autism), 6 (comm. partners)	comm.	social	successful comm. behaviors per minute (observation)
Walters et al. (2021)	analytical, interventional, pre-post studies	age range = 24–36 months/12 females, 33 males	developmental delay	language	linguistic	communication mode (observation) target words (observation)
Wandin et al. (2021)	analytical, interventional, pre-post studies	ages = 27 years, 29 years, 31 years/3 females	Rett syndrome	word production	linguistic	unique Synthesized words per minute (observation) synthesized words per minute (observation)
Yoder and Lieberman (2010)	analytical, interventional, randomized controlled trials	age range = 18–60 months/sex not specified	autism and developmental disorders	generalization	social	PECS exchanges (ESCS)

Note. ACETS = Augmentative Communication Employment Training and Supports; ADHD = attention deficit hyperactivity disorder; ADOS = Autism Diagnostic Observation Scale; Comm. = communication; AT = assistive technology; ESCS =  Early Social Communication Scales; MCDI = MacArthur Bates Communication Developmental Inventory; MLU = mean length of utterance; Mullen = Mullen Scales of Early Learning; PECS = Picture exchange communication system; TBI = traumatic brain injury; QUEST 2.0 = Quebec User Evaluation of Satisfaction with Assistive Technology − 2nd edition; SGD = speech generating device; Vineland = Vineland Adaptive Behavior Scales.