INTRODUCTION
According to the American Psychiatric Association (2022), autism spectrum disorder (ASD) is a developmental deficiency affecting social and communicative skills and is characterized by the presence of fixated interests and repetitive actions. Estimates of ASD popularity vary across studies as some older reports suggested rates of 10-20 per 10,000 children, while the World Health Organization (2013) suggested that around 1 in 160 children globally are diagnosed with ASD. Also, Maenner et al. (2021) revealed that approximately 1 in 44 children in the United States are diagnosed with ASD at different stages of their childhood. However, as the child gets older, the diagnosis of ASD might be complicated because the child has developed some techniques in hiding some major diagnostic manifestations (Wicks and O’Reilly, 2013). This reveals the importance and necessity of investigating ASD across various contexts, considering cultural, social, and methodological factors. It is crucial to understand the global differences in prevalence rates and access to care, as well as to develop a more complete understanding of diagnostic tools for ASD. This is because successfully identifying individuals with ASD requires navigating a complex landscape of assessment methods, each with its own advantages, drawbacks, and important factors to consider.
Diagnostic instruments for ASD can be varied such as standardized tools, screening tools, clinical interviews, and emerging technologies like eye tracking and electrophysiological measures (Wagner et al., 2013). However, choosing the most appropriate method hinges on a multifaceted approach. The individual’s age, developmental level, symptom severity, cultural background, and accessibility of resources all play crucial roles in tailoring the assessment journey (Crane et al., 2018). Also, by critically evaluating the strengths and limitations of various methods, we embark on a path toward accurate and individualized diagnoses, paving the way for effective interventions and support for individuals on the autism spectrum (Gotham et al., 2009). Among the standardized tools available, the Gilliam Autism Rating Scale (GARS) deserves particular attention, and this will be covered in this paper.
The GARS is a highly utilized instrument that merits consideration because of its versatility and accessibility throughout a broad age spectrum spanning from 3 to 22 years (Gilliam, 2006). The assessment tool comprises 66 subscales and 56 items. The aim of it is to offer a comprehensive evaluation of fundamental behaviors associated with autism, including, but not limited to, restrictive and repetitive actions, social interaction and communication, emotional reactions, cognitive style, and speech skills (Gilliam, 2006). Therefore, the existing body of research on the implementation of GARS suggests that it is one of the most used scales internationally.
Since its introduction in the mid-1990s, the GARS has been modified three times, confirming its status as a prevalent informant rating instrument utilized in both clinical and research environments to identify and measure the severity of ASD (Gilliam, 2006, 2014). Initially, GARS was designed to correspond with the DSM criteria for ASD and related challenges in individuals between the ages of 3 and 22 years. However, DSM was assessed and developed over time until the final DSM-5 publication in 2013 (American Psychiatric Association, 2013), and therefore, its ASD diagnostic criteria have been modified. As a result, GARS has been revised to meet the diagnostic criteria (Gilliam, 2006).
In the literature, psychometric validation studies utilizing the original GARS have yielded contradictory findings, as evidenced by its inconsistent reliability, validity, and accuracy in detecting ASD (Gilliam, 2006; Karren, 2017). Also, studies examining the subsequent versions of GARS, such as GARS-2 and GARS-3, have reported more consistent and favorable results in terms of reliability and validity (Mazefsky and Oswald, 2006). For example, some studies using the Persian version of GARS-2 and the original version of GARS-3 found good psychometric properties, although some discrepancies were noted regarding the assignment of items to specific subscales (Samadi et al., 2020). However, apart from these studies, there is a lack of international peer-reviewed research on the measurement properties of the GARS versions collectively, as well as limited data on the diagnostic test accuracy for all three versions (Randall et al., 2018). Therefore, this study aims to address this gap by conducting a thematic review of GARS application. By employing the thematic and systematic review technique over the existing studies employing GARS, this paper aims to extract and systematically explore recurring themes related to its strengths, limitations, and cultural considerations. This in-depth exploration will facilitate a nuanced understanding of GARS’ usefulness as a tool for ASD assessment and contribute to the field of assessment by promoting practices that are better informed and more culturally sensitive.
Purpose of the present study
The primary objective of the study is to systematically and thematically analyze existing research that employs the GARS in diagnosing ASD. This includes understanding the application of GARS, its diagnostic traits, and its effectiveness in accurately identifying individuals with ASD. Additionally, the study will consider the impact of using GARS in different linguistic and contextual settings and how these factors may influence the psychometric properties of the tool. The goal is to shed light on the strengths and limitations of GARS, contributing to the broader dialog on ASD diagnosis and identification.
METHODS
Sample
The sample for this study was derived through a comprehensive and systematic search of four major databases: Web of Science, Scopus, Google Scholar, and PubMed. The objective was to identify studies that utilized the GARS across its versions (1-3). The search was extensive and aimed to capture a wide variety of research that employed GARS in the context of diagnosing ASD. The initial search yielded a total of 27 studies.
To ensure the integrity and uniqueness of the studies, the Mendeley reference management software was used to remove duplicates. This process reduced the number of studies to 19. A thorough screening process was then undertaken to assess the relevance and appropriateness of each study. This included assessing the objectives, methodology, and outcomes of the studies. One study was removed at this stage due to it being an editorial piece rather than original research, resulting in a final set of 18 studies to be included in the thematic review.
Instrument
The instrument under investigation across all selected studies was the GARS. The GARS, available in three versions (GARS-1, GARS-2, and GARS-3), is a prevalent screening and diagnostic tool used in the identification of ASD (Gilliam, 2014). Each version of the GARS incorporates different aspects of ASD and has been refined based on feedback and advancements in understanding ASD. This instrument’s use in various studies provides a perspective on its effectiveness, limitations, and areas of improvement.
Design
The design of this study was rooted in a thematic analysis framework, a qualitative method used to identify, analyze, and interpret patterns or “themes” within the collected data. This method offers a flexible and useful research tool that can provide a rich, detailed, yet complex account of the data. In the context of this study, the thematic analysis allowed for an in-depth examination of the application, strengths, and limitations of the GARS across different populations and contexts.
Procedures
The analysis phase began with a thorough reading of all the studies to identify meaningful units of text relevant to the research topic. This process was meticulous, ensuring that all pertinent information was captured. Units of text dealing with the same issue were grouped together into categories and given provisional definitions. The data were then systematically reviewed to ensure that each category had a name, definition, and exhaustive set of data to support it. Figure 1 explains the used procedure.
Flow diagram of the study selection process for the thematic analysis. Abbreviations: ASD, autism spectrum disorder; GARS, Gilliam Autism Rating Scale.
As the figure shows, in order to enhance the reliability of this review, we conducted a comprehensive search across four major databases: Web of Science, Scopus, Google Scholar, and PubMed, using keywords such as “Gilliam Autism Rating Scale,” “GARS,” “ASD diagnosis,” “psychometric properties,” and “thematic analysis.” This initial search identified 100 records. After removing duplicates, 90 records remained for screening. Following a detailed screening process, 63 records were excluded based on relevance and methodological rigor, leaving 27 full-text articles to be assessed for eligibility. Ultimately, 9 articles were excluded for reasons such as being non-original research or not meeting the inclusion criteria, resulting in 18 studies being included in the qualitative synthesis. This thorough selection process ensured that the review incorporated relevant, recent, and seminal works, addressing potential limitations related to the reliance on existing and preselected assessment instruments and providing a comprehensive understanding of the current state of research in the field.
Finally, to establish the coherence and replicability of the themes, a second researcher independently recoded the initial set of data, focusing specifically on the first question, which resulted in a high level of inter-rater reliability, indicative of the consistency and dependability of the thematic categorizations (Randall et al., 2018). To further ensure the accuracy of our categorization process, levels of agreement for individual categories were meticulously calculated, following the guidelines proposed by previous methodological studies (Samadi et al., 2022a). These categories were then systematically grouped into key themes that reflected the patterns and commonalities identified across the studies. This method allowed us to draw well-supported conclusions from the thematic analysis, ultimately providing a comprehensive understanding of the application and effectiveness of GARS in diagnosing ASD. The final step of the analysis involved synthesizing these themes into broader insights, which facilitated a nuanced interpretation of the data and highlighted the tool’s strengths and limitations in various diagnostic contexts (Dua, 2014; Volker et al., 2016).
RESULTS
Characteristics of the included studies
Table 1 presents an overview of various studies conducted on different versions of the GARS ranging from GARS-1 to GARS-3. The studies aim to evaluate the utility, effectiveness, and limitations of GARS when used with different populations across varying contexts.
Characteristics of the included studies.
| Study | Study aim | Population characteristics | Used version | Measurability | Measurability shortcomings | Measurability strengths | Translation language |
|---|---|---|---|---|---|---|---|
| South et al. (2002) | To examine the validity of the GARS when used with a sample of 119 children with strict DSM-IV diagnoses of autism. | 119 children with strict DSM-IV diagnoses of autism. | GARS-1 | GARS consistently underestimated the likelihood that autistic children in this sample would be classified as having autism. | GARS has a low sensitivity of the summary scores, potentially leading to underdiagnosing autism. | The GARS is reliable and has good internal consistency among the separate scales. | English-US |
| Lecavalier (2005) | To evaluate the construct and diagnostic validity, inter-rater reliability, and effects of participant characteristics of the GARS. | Children and adolescents with ASDs. | GARS-1 | A three-factor solution explaining 38% of the variance was obtained. Internal consistency for the three behavioral subscales was good. | The Developmental Disturbance subscale did not contribute to the Autism Quotient (AQ) and was poorly related to other subscales. | The GARS showed good internal consistency for the three behavioral subscales. | English-US |
| Choi and Kim (2005) | To examine the reliability and validity of the GARS in the Republic of Korea. | 79 children and adolescents, including 56 individuals with autism and 23 individuals without disabilities, in the Republic of Korea. | GARS-1 | Demonstrated strong internal consistency reliability. The GARS scores correlated highly with the ABC scores, providing evidence of concurrent validity. | Not explicitly mentioned. | High reliability, strong concurrent validity with the ABC scores, and good discriminant validity. | Korean |
| Al Jabery (2008) | To enrich the special education literature in Jordan with a new instrument designed to be used with children with autism between the age of 3 and 13 years. | Children with autism between the age of 3 and 13 years in Jordan. | GARS-2 | Fairly robust in indicating and predicting the likelihood of autism and differentiating autism from mental retardation. | Limitations related to sample size, translation accuracy, and age range. | Supported by significant validity and reliability indicators. Can be used to make a differential diagnosis of autism from mental retardation. | Arabic-Jordanian |
| Pandolfi et al. (2010) | To assess the constructs of GARS-2 using data from the standardization sample. | Matched subgroups of the standardization sample. | GARS-2 | A correlated three-factor solution was found, but more than a third of the GARS-2 items were assigned to the wrong subscale. | The Developmental Disturbance subscale did not contribute to the AQ and was poorly related to other subscales. | The study provides recommendations for instrument revision. | English-US |
| Jhin et al. (2011) | To standardize and verify the reliability and validity of the Korean-Gilliam Autism Rating Scale-2 (K-GARS-2). | 109 patients with autism aged 3 to 22 years in Korea. | GARS-2 | High reliability and validity. The K-GARS-2 standard scores highly correlated with scores of CARS. The Autism subscale standard score and the AQ also showed a strong correlation. | Not explicitly mentioned. | High reliability and validity. Can be used for a wide age range (3-22 years old). | Korean |
| Li (2005) | To begin the process of developing two measures of autism for use with Chinese-speaking individuals. | A group of 20 Chinese immigrants. | GARS-2 | The Chinese translations of CARS2-QPC and GARS-2 measure what they were intended to evaluate. They have a good internal consistency. | Small sample size, utilization of nonclinical sample, restricted educational range of the sample, and limited age range of participants. | The Chinese versions of CARS2-QPC and GARS-2 are valid instruments for measuring characteristics of ASDs. | Chinese-Mandarin |
| Diken and Gilliam (2012) | To evaluate the psychometric characteristics of the Turkish version of GARS-2 (TV-GARS-2) in order to standardize it into Turkish. | Individuals diagnosed with autism, intellectual disability, hearing impairment and typically developing children. | GARS-2 | The Turkish version of GARS-2 demonstrated strong psychometric characteristics, including high reliability and validity. | Not explicitly mentioned. | TV-GARS-2 is a valid and reliable scale that can be used in different settings. | Turkish |
| Jackson et al. (2013) | To adapt GARS-2 from English into Spanish. | Children with autism or other significant developmental disabilities. | GARS-2 | The Spanish version of GARS-2 demonstrated a high degree of internal consistency and good discriminative validity. It measures the same constructs as the original version. | Small sample size and the generalizability of the results. | This culturally and linguistically appropriate assessment tool could be an important instrument for autistic children with Spanish-speaking parents. | Spanish |
| Samadi and McConkey (2014) | To evaluate the use of a rating scale for autism in Iran. | Parents of 656 children: 442 who had been diagnosed with autism, 112 intellectually disabled, and 102 normally developing. | GARS-1 | Iranian parents identified items relating to impaired social interaction and repetitive behaviors as more indicative of autism rather than those relating to communication and language. | Limitations relate to the fact that parents of children with ASD had been previously given a diagnosis, possibly affecting their responses. | The GARS proved to be a useful rating scale for use with families in Iran. A short ten-item version showed promise as a possible screening tool. | Persian |
| Dua (2014) | To examine the factor structure, internal consistency, and screening sensitivity of GARS-2 ratings completed by special education teaching staff. | Sample of students (N = 216) with ASDs. | GARS-2 | The exploratory factor analysis supported the retention of three factors labeled Stereotyped and Repetitive Behaviors, Social Avoidance and Withdrawal, and Atypical Language and Communication. | The sensitivity estimates for the Autism Index suggest that the instrument results in a high percentage of false negative results for ASD. | This culturally and linguistically appropriate assessment tool could be an important instrument for autistic children with Spanish-speaking parents. | English-US |
| Volker et al. (2016) | To examine the factor structure, internal consistency, and screening sensitivity of GARS-2. | Individuals with autism or other significant developmental disabilities. | GARS-2 | Demonstrated high internal consistency among the separate scales and strong discriminative validity. | Some items yielded low factor loadings and may include complex wording. | Factor structure was broadly replicated, especially on items relating to social interaction and social communication. | English-US |
| Tsirgiotis et al. (2022) | To investigate specific behaviors in which differences lie between autistic males and females. | Profiles of 777 children were analyzed. | GARS-3 | Males demonstrated greater difficulty in six CARS2-ST items and seven behaviors on GARS-3, mostly reflecting restricted and repetitive behaviors. | The use of existing and preselected assessment instruments restricts research to the existing, possibly androcentric conceptualization of ASD. | Large sample size, the use of empirically validated ASD instruments, and sophisticated statistical methods. | English-Australia |
| Samadi et al. (2022a) | To evaluate the level of overlap between two different versions of the GARS (GARS-2 and GARS-3). | A group of 148 cases with ASD and developmental disabilities in the Kurdistan Region of Iraq. | GARS-2 and GARS-3 | GARS-3, updated based on the DSM-5, tends to diagnose children with accompanying diagnoses and different levels of symptoms severity of ASD at different age levels. | Limitations related to the sample size, the pre-informed status of parents about their children’s diagnosis of ASD, and lack of confirmed diagnosis of ASD and ID. | Helps professionals and policymakers in low- and middle-income countries understand the updated versions of the available scales and depend on the older version. | Kurdish |
| Samadi et al. (2022b) | To develop a suitable screening procedure for use in the Kurdistan Region of Iraq. | 735 children, including those with a pre-existing diagnosis of ASD, those with a confirmed diagnosis of a developmental disability, and those who were typically developing. | GARS-3 | A ten-item screening tool was developed that had a reasonable sensitivity and specificity in distinguishing possible cases of autism from typically developing children. | Screening tools need to be used proactively rather than retrospectively. The screening tool needs to be embedded with systems that allow for further diagnostic assessments. | Provides an example of how screening procedures can be developed in the context of support services in low- and middle-income countries. | Kurdish |
| Samadi et al. (2022c) | To evaluate the Kurdish translation of the GARS—third edition (GARS-3). | 735 children (original study incorrectly stated 789), including 442 diagnosed with ASD, 165 with an intellectual disability, 49 with communication disorders, and 133 typically developing children. | GARS-3 | The factor structure of GARS-3 was broadly replicated, especially on items relating to social interaction and social communication. | The main limitations were that parents of individuals with ASD had already received a diagnosis, thus possibly affecting their responses, and that there was no confirmed diagnosis of ASD or ID. | Confirms the universality of autism symptoms but also the different emphasis Kurdish parents may place on them. | Kurdish |
| Isbell (2022) | To examine the factor structure, internal consistency, and screening sensitivity of GARS-3. | Sample of students with ASD and other developmental disabilities. | GARS-3 | The exploratory factor analysis supported the retention of three factors labeled Stereotyped Behaviors, Social Avoidance, and Atypical Language and Communication. | The sensitivity estimates for the Autism Index suggest that the instrument results in a high percentage of false negative results for ASD. | The study provides some psychometric support for the use of GARS-3 as a screening tool. | English-US |
| Camodeca (2023) | To examine the factor structure, internal consistency, and screening sensitivity of GARS-3. | Sample of clinically referred children suspected of autism [mean age = 8.98; autism (AUT), n = 87; not autism (NOT), n = 99]. | GARS-3 | The exploratory factor analysis supported the retention of three factors: Stereotyped/Repetitive Behavior, Social Avoidance and Withdrawal, and Atypical Language and Communication. | The sensitivity and specificity of the Autism Index were lower than predicted. | The study provides preliminary evidence for the psychometric properties of GARS-3. | English-US |
Abbreviations: ABC, autism behavior checklist; ASD, autism spectrum disorder; CARS, childhood autism rating scale; GARS, Gilliam Autism Rating Scale; ID, intellectual disability.
The first study by South et al. (2002) used GARS-1 to examine its validity in a sample of 119 children with strict DSM-IV diagnoses of autism. The study found that GARS-1 consistently underestimated the likelihood of a child being classified as having autism, indicating potential issues with sensitivity. In contrast, a study by Lecavalier (2005) evaluated the construct and diagnostic validity, inter-rater reliability, and effects of participant characteristics of GARS-1 in a sample of children and adolescents with ASD. The study found that GARS-1 demonstrated good internal consistency among the separate scales, suggesting strong reliability. Moving on to GARS-2, a study by Pandolfi et al. (2010) found that more than a third of the GARS-2 items were assigned to the wrong subscale, indicating potential issues with the instrument’s validity.
Studies on GARS-3, the most recent version of the scale, have focused on its construct validity, internal consistency, and screening sensitivity (Dua, 2014; Isbell, 2022). However, the sensitivity estimates for the Autism Index suggest that GARS-3 may result in a high percentage of false negative results for ASD, indicating a need for further refinement of the tool. The table also highlights the extensive global application of the GARS, with translations available in multiple languages, including English, Korean, Arabic-Jordanian, Chinese-Mandarin, Turkish, and Persian. This reflects the universal significance of GARS in ASD diagnosis and the importance of cultural and linguistic adaptation of diagnostic tools.
A comparison of GARS versions: GARS-1, GARS-2, and GARS-3
The GARS has been a pivotal instrument in the field of ASD for diagnosing and identifying individuals with ASD. Since its inception, the GARS has undergone several iterations, each with its unique characteristics and sets of improvements. This essay provides an in-depth comparison of the three versions of the GARS—GARS-1, GARS-2, and GARS-3—examining their evolution, their strengths, and limitations.
The original GARS, often referred to as GARS-1, was the starting point of this diagnostic journey. As a pioneering tool, it provided a structured approach to ASD screening and diagnosis (South et al., 2002; Lecavalier, 2005). However, despite its innovative approach, GARS-1 was not without its limitations. Studies revealed that it consistently underestimated the likelihood of a child being classified as having autism, indicating potential issues with sensitivity (South et al., 2002). This critical feedback paved the way for the development of an improved version, GARS-2.
GARS-2 was designed to address the limitations identified in its predecessor. It incorporated a broader range of ASD behaviors and symptoms, and its psychometric properties were extensively evaluated (Pandolfi et al., 2010; Volker et al., 2016). Despite these enhancements, some issues persisted. Researchers noted that GARS-2, much like GARS-1, had sensitivity issues and a tendency to assign items to the wrong subscale, affecting its overall validity (Pandolfi et al., 2010). These persistent challenges led to the development of the third iteration, GARS-3.
GARS-3 represents the most recent effort to refine this diagnostic tool. The scale was updated to bolster its psychometric properties and improve its accuracy in identifying individuals with ASD (Dua, 2014; Volker et al., 2016; Isbell, 2022). GARS-3 was designed to reflect the latest knowledge about ASD, integrating new insights about the disorder. Despite these advances, research on GARS-3 is still limited, and its effectiveness is yet to be fully established (Dua, 2014; Isbell, 2022).
Across all these versions, the GARS has demonstrated significant versatility. It has been translated into multiple languages and used in various cultural and linguistic contexts, including Australia, China, Jordan, Korea, Iraq, Iran, and Turkey (Choi and Kim, 2005; Li, 2005; Al Jabery, 2008; Jhin et al., 2011; Diken and Gilliam, 2012; Samadi et al., 2022b; Tsirgiotis et al., 2022; Camodeca, 2023). This global application underscores the universal significance of GARS in ASD diagnosis and the importance of cultural and linguistic adaptation in diagnostic tools.
To sum up, while each version of the GARS shares a common goal of accurately identifying and diagnosing ASD, they each present unique characteristics and have undergone different levels of validation and refinement. The evolution of the GARS versions reflects the ongoing advancements in ASD understanding and the continuous effort to improve diagnostic tools. However, as our knowledge and understanding of ASD continue to grow, so too should the tools we use to identify and diagnose it. Thus, further research and refinement of the GARS are necessary to ensure its effectiveness and accuracy in diagnosing ASD across various cultural and linguistic contexts.
Themes on GARS
Table 2 provides a thematic analysis of the GARS based on various studies. The first theme discusses the psychometric properties of GARS, with studies examining the validity, reliability, and factor structure across different versions (GARS-1, GARS-2, and GARS-3). The results indicate mixed outcomes regarding these properties. The second theme delves into the cultural context and translation of GARS, highlighting its universal applicability and the importance of cultural and linguistic adaptation. Studies have translated GARS into various languages and applied it in different cultural settings, such as Australia, China, Jordan, Korea, Iraq, Iran, and Turkey. The third theme explores the impact of GARS on ASD diagnosis and identification, with research data indicating its utility in distinguishing individuals with ASD from those with other developmental disabilities and typically developing individuals. The final theme outlines the limitations of GARS and recommendations for improvement, including low sensitivity, non-optimal cut-off scores, issues with item assignment, and suggestions for improvements such as revising certain items, considering age effects, and expanding the sample characterization. Before delving into the discussion of the themes below, Figure 2 provides a clear overview of the distribution of these themes.
Themes for the thematic analysis.
| No. | Theme | Supporting data | Sample sources |
|---|---|---|---|
| 1 | Psychometric properties of GARS | Multiple studies examined the validity, reliability, and factor structure of different versions of GARS, including GARS-1, GARS-2, and GARS-3, often revealing mixed results. | Dua (2014), Isbell (2022), Lecavalier (2005), Pandolfi et al. (2010), South et al. (2002), Volker et al. (2016) |
| 2 | Translation and cultural context | Studies that translated GARS into different languages and utilized it in various cultural settings, such as Australia, China, Jordan, Korea, Iraq, Iran, and Turkey. This theme highlights the universal applicability of GARS and the importance of cultural and linguistic adaptation. | Al Jabery (2008), Camodeca (2023), Choi and Kim (2005) Diken and Gilliam (2012), Jhin et al. (2011), Li (2005), Samadi et al. (2022b, c), Tsirgiotis et al. (2022) |
| 3 | Impact on ASD diagnosis and identification | Research data indicating GARS’ utility in distinguishing individuals with ASD from those with other developmental disabilities and typically developing individuals. These studies often discuss the sensitivity, specificity, and accuracy of GARS in identifying ASD. | Dua (2014), Isbell (2022), Tsirgiotis et al. (2022), Volker et al. (2016) |
| 4 | Limitations and recommendations for improvement | Multiple studies discussed the limitations of GARS, such as its low sensitivity, non-optimal cut-off scores, and issues with item assignment. These studies often provided suggestions for improving the tool, such as revising certain items, considering age effects, and expanding the sample characterization. | Dua (2014), Isbell (2022), Pandolfi et al. (2010), South et al. (2002) |
Abbreviations: ASD, autism spectrum disorder; GARS, Gilliam Autism Rating Scale.
Distribution of themes in the thematic analysis of GARS. Abbreviations: ASD, autism spectrum disorder; GARS, Gilliam Autism Rating Scale.
As seen in the pie chart, the largest segment, Translation and Cultural Context (40%), emphasizes the importance of adapting the GARS across diverse linguistic and cultural settings. Psychometric Properties of GARS (30%) highlights the validity, reliability, and factor structure. Both Impact on ASD Diagnosis and Identification and Limitations and Recommendations for Improvement account for 15% each, focusing on the utility of GARS in distinguishing ASD and identifying areas for enhancement, respectively. This visualization offers a concise overview of the thematic distribution, aiding in the comprehensive understanding of GARS evaluation. Each theme is discussed in detail below.
Psychometric properties of GARS
The GARS, across its various iterations from GARS-1 to GARS-3, has been a widely used tool in the identification and diagnosis of ASD. The psychometric properties of this instrument have been the subject of several research studies, which have aimed to explore its validity, reliability, and factor structure.
The validity of the GARS has been examined in several studies. For instance, Lecavalier (2005) and South et al. (2002) examined the construct and diagnostic validity of GARS-1. Similarly, the construct validity of GARS-2 was investigated by Pandolfi et al. (2010) and Volker et al. (2016), while the construct validity of GARS-3 was examined by Volker et al. (2016) and Dua (2014).
Reliability is another critical psychometric property of the GARS. Studies by Lecavalier (2005), South et al. (2002), Pandolfi et al. (2010), Volker et al. (2016), and Dua (2014) have all assessed the internal consistency and inter-rater reliability of different versions of GARS. These studies have generally found that the GARS demonstrates good internal consistency, a key indicator of reliability.
The factor structure of the GARS has also been the focus of several studies. Lecavalier (2005) and South et al. (2002) conducted factor analyses on GARS-1, while Pandolfi et al. (2010) and Volker et al. (2016) conducted factor analyses on GARS-2. Most recently, the factor structure of GARS-3 was explored by Volker et al. (2016) and Dua (2014). These studies have generally found that the factor structures of the different versions of GARS align reasonably well with the intended constructs of the instrument. However, across these studies, the psychometric properties of the GARS have garnered mixed results. Some studies have reported strong psychometric properties, while others have identified potential issues with the instrument’s validity, reliability, and factor structure. These mixed findings underscore the need for continuous research and possible revisions of the GARS to ensure its effectiveness in accurately identifying individuals with ASD.
In short, the GARS, across its various iterations, has served as an essential tool in ASD identification and diagnosis. However, the mixed findings regarding its psychometric properties highlight the need for continued research to ensure its effectiveness. As our understanding of ASD continues to evolve, so too should the tools we use to identify and diagnose it. Further research into the psychometric properties of the GARS, as well as potential revisions of the instrument, will be crucial in this ongoing effort.
Translation and cultural context of GARS
The GARS has seen extensive use in a variety of cultural and linguistic contexts, underscoring its global significance in the screening and diagnosis of ASD. The application of GARS across different sociocultural environments has led to its translation into multiple languages and its usage in countries such as Australia, China, Jordan, Korea, Iraq, Iran, and Turkey (Choi and Kim, 2005; Li, 2005; Al Jabery, 2008; Jhin et al., 2011; Diken and Gilliam, 2012; Samadi et al., 2022b; Tsirgiotis et al., 2022; Camodeca, 2023). These studies highlight the importance of developing linguistically and culturally appropriate versions of GARS to ensure its efficacy in diverse settings. The process of translating and adapting the GARS has involved direct translation of the tool into the target language, back-translation to verify the accuracy of translation, and pilot testing to ensure cultural appropriateness.
These culturally and linguistically adjusted versions of GARS have been found to maintain good psychometric properties, demonstrating high internal consistency and validity in identifying ASD. For instance, the Arabic-Jordanian version of GARS-2 showed significant validity and reliability indicators (Al Jabery, 2008). The Chinese-Mandarin version of GARS-2 was found to have good internal consistency and concurrent validity (Li, 2005), and the Korean version of GARS-2 showed high internal consistency and discriminant validity (Choi and Kim, 2005).
However, it is important to note that the process of translating and adapting GARS for use in different cultural contexts is not without challenges. The structure, content, and interpretation of the items may vary across cultures and languages, potentially affecting the overall utility and psychometric properties of the tool (Samadi et al., 2022c). Therefore, it is crucial to ensure that the translated versions of GARS maintain fidelity to the original instrument while being culturally sensitive and appropriate for the target population.
To conclude, the translation and cultural adaptation of GARS have significantly broadened its global applicability as a screening and diagnostic tool for ASD. However, ongoing research and refinement are necessary to ensure that these translated versions of GARS are culturally sensitive, accurate, and effective in identifying ASD in diverse cultural and linguistic settings.
Impact of GARS on ASD diagnosis and identification
The GARS plays a pivotal role in identifying and diagnosing ASD. The tool’s effectiveness in accurately distinguishing individuals with ASD from those exhibiting other developmental disabilities or typical development has been a focal point in numerous research studies (Dua, 2014; Volker et al., 2016; Isbell, 2022; Tsirgiotis et al., 2022).
The GARS has been found to demonstrate considerable discriminant validity. It has shown the capacity to differentiate between individuals diagnosed with ASD and those with other developmental disabilities or typically developing peers. As such, the GARS serves as a valuable assessment tool in the field of ASD diagnosis, providing insightful data to guide further diagnostic procedures and interventions (Dua, 2014; Isbell, 2022).
However, the degree to which the GARS accurately identifies individuals with ASD varies across studies. While some research indicates high sensitivity and specificity in distinguishing individuals with ASD (Dua, 2014), other studies suggest that the tool’s sensitivity may be lower than predicted, particularly when using the suggested cut-off scores (Isbell, 2022). This discrepancy underscores the importance of using GARS as part of a comprehensive diagnostic process rather than a standalone diagnostic tool.
The impact of the GARS on ASD diagnosis and identification is also influenced by the instrument’s ability to capture the multifaceted nature of ASD. The tool assesses a broad range of behaviors and symptoms associated with ASD, from social interaction and communication difficulties to restrictive and repetitive behaviors (Tsirgiotis et al., 2022). However, the tool’s sensitivity in capturing these varied symptoms, particularly in different cultural and linguistic contexts, remains an area of ongoing research.
To this end, the GARS plays a significant role in the identification and diagnosis of ASD. While it has demonstrated considerable discriminant validity, the tool’s sensitivity and specificity in identifying ASD vary across studies. Further research is needed to optimize the use of GARS in accurately diagnosing ASD, considering factors such as the multifaceted nature of ASD and the tool’s application in different cultural and linguistic contexts.
Limitations and recommendations for improvement of GARS
The GARS, through its various iterations, has proven instrumental in the diagnosis and identification of ASD. However, like any diagnostic tool, it has its limitations and areas for potential improvement, as highlighted in numerous research studies (South et al., 2002; Pandolfi et al., 2010; Dua, 2014; Isbell, 2022).
A common critique of the GARS centers around its sensitivity. Several studies have found that the tool’s sensitivity may be lower than initially predicted, particularly when utilizing the suggested cut-off scores (Isbell, 2022). This suggests that the GARS may not be as effective as desired in accurately identifying all individuals with ASD, potentially leading to underdiagnosis. Another notable limitation pertains to the GARS’ item assignment. In many instances, studies have found that more than a third of the GARS-2 items were assigned to the wrong subscale (Pandolfi et al., 2010). Such misassignments can potentially impact the validity of the results and the accuracy of the diagnosis.
Recommendations for improving the GARS often involve revising certain items and considering age effects. Studies suggest that eliminating redundant items, items with weak factor loadings, and considering the non-linear changes or small but potentially important differences across age groups could improve the instrument’s validity (South et al., 2002; Pandolfi et al., 2010). Additionally, the need to expand the sample characterization has been emphasized. Given that ASD symptoms can greatly vary across individuals, it is crucial for the GARS to be validated across a more diverse sample that represents these variations (South et al., 2002; Pandolfi et al., 2010).
Briefly, while the GARS serves as a valuable tool in identifying and diagnosing ASD, it is not without its limitations. Continued research and potential revisions of the GARS are necessary to ensure its effectiveness and accuracy in diagnosing ASD. These improvements would contribute to better identification and understanding of ASD, ultimately aiding in the provision of appropriate interventions and support for individuals with ASD.
DISCUSSION
The collective aim of the studies under consideration is to evaluate the effectiveness and applicability of the GARS in identifying and diagnosing ASD across various populations and developmental disabilities. These studies, employing varied methodologies, including thematic analysis, factor analyses, and clinical discriminant validity assessments, have provided nuanced insights into the functionality of GARS across editions and contexts.
The methodological approaches ranged from exploratory and confirmatory factor analyses (Volker et al., 2016; Isbell, 2022) to regression analyses and receiver operator characteristic curve analyses (Camodeca, 2023). The main findings from these studies reveal a complex picture of GARS’s utility. For instance, Tsirgiotis et al. (2022) found phenotypic differences in ASD presentations between males and females, while Samadi et al. (2022c) highlighted the challenges of updating diagnostic scales to match the evolving criteria for ASD diagnosis, particularly in low- and middle-income countries (LMICs).
The studies consistently underscore the importance of cultural and contextual factors in the application of GARS. For instance, the work of Samadi et al. (2022b) in the Kurdistan Region of Iraq illustrates not only the universality of autism symptoms but also the unique emphasis different cultures may place on them, indicating the need for culturally sensitive adaptations of the scale.
Despite the utility of GARS as highlighted by these studies, several limitations and concerns have been raised. For example, Isbell (2022) questions the clinical discriminant validity of GARS-3, while Camodeca (2023) points out the inadequate criterion validity of GARS-3 in complex community samples, suggesting that other published screeners may have better predictive utility.
The argument could be made that the variability in the psychometric properties of the GARS across studies indicates a need for caution in its use, particularly in diverse cultural contexts. The studies by Samadi et al. (2022b, c) in the Kurdistan Region amplify this concern, as they reveal the nuanced challenges in LMICs, such as the scarcity of resources and trained professionals, which can impact the diagnostic process.
Moreover, the evidence of sex/gender differences in ASD presentations such as in the study by Tsirgiotis et al. (2022) raises critical questions about the potential for GARS to overlook or misdiagnose certain populations, particularly females, whose symptoms may not align with the historically male-centric conceptualization of ASD.
It is apparent from these studies that while GARS has been a valuable tool, there is a pressing need for continuous refinement of the scale. The studies by Isbell (2022) and Volker et al. (2016) suggest that revisions are necessary to address issues with factor structure and item relevance. Furthermore, the question of how the GARS can be adapted effectively for use in diverse cultural contexts without compromising its psychometric integrity remains a challenging yet crucial endeavor.
In conclusion, the thematic analysis of the 18 studies presents a comprehensive yet complex picture of the GARS’ utility in diagnosing ASD. It is clear that while GARS has widespread use and applicability, its effectiveness is contingent upon continuous examination and adaptation, taking into account cultural, linguistic, and gender-related nuances. The studies collectively call for an informed and cautious application of GARS, with an emphasis on the development of additional research, culturally sensitive adaptations, and potential revisions to ensure the tool’s accuracy and relevance in varied contexts.
Implications
The findings from the thematic review of the GARS across its iterations have several implications for clinical practice and research. Given the mixed results concerning the psychometric properties of GARS, practitioners should exercise caution when relying solely on this tool for diagnosing ASD. The scale should be used as a component of a comprehensive assessment strategy, alongside other diagnostic tools and clinical evaluations. Furthermore, the cultural and linguistic adaptability of GARS, as demonstrated by its translations and applications in diverse contexts, underscores the importance of cultural competence in ASD assessment. Clinicians should be aware of the cultural nuances that may influence the interpretation of behaviors relevant to ASD diagnoses and should seek appropriate translations and adaptations of the GARS to maintain its reliability and validity across different populations.
Limitations
While the thematic review provides a broad overview of the application and utility of GARS, it is important to acknowledge the limitations inherent in the studies included. Several studies noted the potential for reduced sensitivity of the GARS, particularly in certain subpopulations, such as females with ASD, which may lead to underdiagnosis or delayed diagnosis. Additionally, the variability in the assignment of items to subscales across different versions of GARS may impact the tool’s consistency and the clinicians’ confidence in its diagnostic accuracy. Another limitation to consider is the reliance on existing and preselected assessment instruments, which may restrict research to the current conceptualization of ASD and potentially overlook emerging understandings of the spectrum. The studies also indicate the need for further research to establish normative data across diverse cultural and linguistic groups, given the potentially confounding effects of cultural interpretation on the assessment results.
CONCLUSION
The thematic review of the GARS across multiple studies indicates that while it is a valuable tool for the identification and diagnosis of ASD, its use is not without challenges. The scale exhibits mixed psychometric properties and requires continuous validation, especially in light of evolving diagnostic criteria and diverse cultural applications. Its effectiveness in accurately identifying ASD across various cultural and linguistic contexts must be considered, with a need for ongoing research and potential revisions to enhance its utility. The translation of GARS into multiple languages and its application across different cultures reflects its global significance in ASD diagnosis. However, the adaptation process must be rigorous to ensure cultural relevance and accuracy. Ultimately, the GARS should be utilized as part of a multifaceted diagnostic approach, informed by clinical judgment and supplemented by additional diagnostic tools to provide a comprehensive assessment of ASD. The collective body of research calls for a nuanced application of GARS, with an emphasis on culturally sensitive practices and the need for continuous improvement of the assessment tool.