The Gilliam Autism Rating Scale-Third Edition (GARS-3) is a norm-referenced screening instrument used to identify individuals with autism spectrum disorders (ASD), based on definitions from the American Psychiatric Association (2012) and the Autism Society (2012). This article provides fundamental information about autism, instructions for administering and scoring the scale, guidance for interpreting results, and technical details about the instrument. In this article we will discuss background information on autism and suggest applications of the scale.
Description of the GARS-3:
The GARS-3 is a norm-referenced screening tool designed to identify individuals aged 3 to 22 exhibiting severe behavioral issues indicative of autism. It comprises 58 clearly articulated items, divided into six subscales, each detailing specific, observable, and measurable behaviors. These subscales are:
Restricted/Repetitive Behaviors:
This subscale, encompassing 13 items, gauges stereotyped behaviors, fixated interests, routines, or rituals. Examples include staring at hands or objects, flicking fingers rapidly, and engaging in repetitive or ritualistic behaviors.
Social Interaction:
With 14 items, this subscale measures deficits in social behaviors such as failure to initiate conversations, minimal expressed pleasure during interactions, and a lack of interest in others.
Social Communication:
Comprising 9 items, this subscale assesses responses to social situations and understanding of social interaction and communication. Examples include difficulty understanding jokes, teasing, and predicting social consequences.
Emotional Responses:
This subscale, consisting of 8 items, evaluates extreme emotional responses to everyday situations, including excessive need for reassurance, upset reactions to changes in routines, and extreme reactions to loud, unexpected noise.
Cognitive Style:
With 7 items, this subscale measures idiosyncratic fixated interests, characteristics, and cognitive abilities, such as exceptionally precise speech, excessive focus on single subjects, and intense, obsessive interests in specific intellectual topics.
Maladaptive Speech:
Comprising 7 items, this subscale describes deficits and idiosyncrasies in verbal communication, including word or phrase repetition, flat tone or affect in speech, and uttering meaningless words or phrases.
Each subscale aligns with the ASD definition by the Autism Society (2012) and diagnostic criteria for autistic disorder outlined in the Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5; American Psychiatric Association).
Uses of the GARS-3:
The GARS-3 serves various purposes, including identifying individuals with Autism Spectrum Disorder (ASD), assessing the severity of ASD symptoms, documenting progress from intervention programs, targeting goals for Individualized Education Programs (IEPs), and acting as a research tool.
Identifying Individuals With Autism Spectrum Disorder:
The GARS-3 aids in identifying individuals with ASD by providing objective scores that assist in distinguishing those likely to have autism from those who do not. Its reliability and validity make it valuable alongside other assessment data, behavioral observations, case histories, and parent interviews for diagnosing autism.
Assessing Serious Behavioral Problems:
For individuals with severe behavioral problems, the GARS-3 offers specific information on stereotyped behaviors, communication impairments, and social interaction difficulties, which are relevant regardless of diagnosis. It aids in describing the nature of the person’s issues and guiding treatment decisions.
Documenting Behavioral Progress:
The GARS-3’s objective scales facilitate documenting behavioral progress, aiding in program planning, decision-making, and educational placement. It’s particularly useful for annual evaluations in special education and for accumulating data for decisions regarding extended-year schooling.
Targeting Goals for IEPs:
By identifying specific behavioral strengths and weaknesses, the GARS-3 assists in setting goals and intervention targets for individual students. Improvement in the behaviors described within the subscale items can be utilized as goals for IEPs.
Data Collection for Research:
The GARS-3 provides valid and reliable data for research purposes due to its objective measurement of behaviors based on frequency. Its format allows accurate measurement by various stakeholders, making it a promising instrument for researchers, both within and outside the school environment.
Administration Procedures:
Each rater administering the GARS-3 requires a copy of the Summary/Response Form and a writing instrument. The examiner should write the individual’s name and the date of the rating on the form’s front page. As the administration and scoring instructions are consistent across all subscales, the process for completing ratings will be explained using the Restricted/Repetitive Behaviors subscale as an example. Refer to the table below for visual aid during this discussion.
Raters should rate items on a scale of 0 to 3 based on how well they describe the individual’s behavior. Raters should circle the number that best matches their observations of the person’s typical behavior in ordinary circumstances. It’s crucial to rate every item, and if unsure, observers should delay the rating and observe the individual for six hours to make an informed assessment. Each item must receive a score.
Follow these guidelines for ratings, ensuring that every item is rated accordingly.
| 0 | Not at all like the individual | You have never seen the person behave in this manner. |
| 1 | Not much like the individual | The person seldom behaves in this manner (perhaps one to two times per 6-hour period). |
| 2 | Somewhat like the individual | The person sometimes behaves in this manner (perhaps three to four times per 6-hour period). |
| 3 | Very much like the individual | The person frequently behaves in this manner (at least five to six times per 6-hour period). |
Each item on the scale is scored from 0 to 3 based on how well it describes the individual’s behavior. For instance, if Item 1 suggests that the individual spends the majority of their time in repetitive or stereotyped behaviors when left alone, a rating of 2 indicates that this behavior somewhat aligns with the individual’s actions, occurring three to four times in a 6-hour period. If continuous observation isn’t feasible, raters may seek advice from someone who observes the individual regularly or extrapolate ratings based on their own observations. Most raters tend to generalize their ratings as “not much,” “somewhat,” or “very much.” Each item within this subscale, as well as across the other five subscales, follows the same scoring procedure. In cases of uncertainty regarding the frequency of behaviors, raters should either seek further observation opportunities or provide ratings based on reliable information obtained indirectly. It’s crucial to score every item, and if any are left blank, the examiner must inquire why and ensure they are scored appropriately.
When Individuals Do Not Communicate
According to the Centers for Disease Control and Prevention (2012), around 40% of children diagnosed with ASD are nonverbal. Additionally, between 25% to 40% of these children may develop some vocabulary between the ages of 12 to 18 months but subsequently cease speaking. Moreover, some individuals with ASD may not start speaking until later in childhood. During the norming process of the GARS-3, data revealed that 25% of students diagnosed with ASD were nonverbal. Many of these individuals did not utilize sign language or augmentative communication devices. Consequently, it is anticipated that a considerable number of individuals referred for evaluation may face communication challenges.
Rating these individuals using the GARS-3 poses difficulties because the Maladaptive Speech subscale and Items 45, 47, and 51 on the Cognitive Style subscale necessitate some form of communication. If an individual is entirely nonverbal and has not communicated through signs or verbally with anyone within the past six months, the examiner or rater should exclude the Maladaptive Speech and Cognitive Style subscales from the evaluation. Additionally, they should indicate on the GARS-3 Summary/Response Form that the individual was nonverbal/noncommunicative, and only four subscales were completed.
Recording and Interpreting GARS-3 Results
This section provides guidance on scoring, recording, analyzing, and interpreting GARS-3 scores. Key topics covered include completing the Summary/Response Form, understanding normative scores, utilizing GARS-3 scores for identifying individuals with autism spectrum disorder (ASD), and exercising caution during result interpretation.
Completing the Summary/Response Form
The Summary/Response Form serves as a tool to record and summarize the rater’s evaluation of the individual’s behavior. It comprises seven sections:
- Identifying Information
- Subscale Performance
- Composite Performance
- Interpretation Guide
- Ratings
- GARS-3 Characteristics
The form’s primary purposes include rating the individual’s behavior in domains essential for autism diagnosis, summarizing GARS-3 performance, making diagnostic inferences, and documenting ASD criteria fulfillment.
Section 1: Identifying Information
Section 1 focusing on Identifying Information, where details like the individual’s name, date of rating, date of birth, age at rating, gender, school, and rater’s information are recorded. An example illustrating age calculation is provided to ensure accurate recording.
Section 2: Subscale Performance
In this section, the examiner records the raw scores, percentile ranks, and scaled scores for each subscale of the GARS-3. The total raw score for each subscale is calculated by summing the individual item scores within that subscale. All items within a subscale must receive a rating for the subscale to receive a raw score; if an item is left blank, the subscale cannot receive a raw score. The subtotal raw scores are then written at the bottom of each subscale. These raw scores are transferred to Section 2 of the GARS-3 Summary/Response Form. Once all subscale raw scores are recorded, they should be checked for accuracy. Next, the raw scores are converted to percentile ranks and scaled scores using the table in Appendix A in the manual.
For example, if Daniel’s raw score for the Restricted/Repetitive Behaviors subscale is 52, the examiner would find the raw score of 52 in the appropriate column of the table in Appendix A to determine the corresponding percentile rank (84) and scaled score (13). This process is repeated for each subscale, and Daniel’s scaled scores for the other subscales are recorded accordingly: Social Interaction (12), Social Communication (12), Emotional Responses (10), Cognitive Style (7), and Maladaptive Speech (10). The standard errors of measurement (SEMs) for each subscale corresponding to the scaled scores are preprinted in this section of the form.
Section 3: Composite Performance
The Autism Index is determined by summing the scaled scores for each subscale and consulting Appendix B in the manual. In Daniel’s case, since all six subscales were completed, the sum of his scaled scores is 64, corresponding to an Autism Index of 106. These scores are recorded in Section 3 of the Summary/Response Form.
In cases where all six subscales are not completed, such as when a person is noncommunicative, the examiner can still determine a reliable Autism Index by summing the scaled scores for the completed subscales (Restricted/Repetitive Behaviors, Social Interaction, Social Communication, and Emotional Responses). The procedure for determining the Autism Index remains the same, consulting Appendix B to find the appropriate column for the sum of the completed subscales’ scaled scores and locating the corresponding Autism Index.
Section 4: Guide for Interpretation
In Section 4, you’ll find descriptive categories corresponding to the Autism Index, ranging from Unlikely to Very Likely. This index provides an interpretation of the Autism Index score, helping determine the Severity Level for ASD. The Severity Level is represented by a numerical rating aligned with the diagnostic criteria index. For instance, if the score falls within the range, and according to DSM-5, the severity level is rated as 3, it indicates a need for very substantial support.
Section 5: Rating Procedures
Contained within Section 5 are the individual items for each of the GARS-3 subscales. Instructions for rating these items have been explained earlier. At the bottom of each subscale, there’s a designated box for tallying the sum of the item scores.
Section 6. GARS-3 Characteristics
Section 6 of the GARS-3 provides a concise overview of its characteristics, offering examiners quick access to essential information without requiring the Examiner’s Manual. This summary is particularly useful in scenarios like IEP meetings or multidisciplinary team gatherings where technical details about the GARS-3 may be needed.
Section 7: Autism Spectrum Disorder Diagnostic Validation Checklist
It aids examiners in efficiently incorporating information about the GARS-3 into their written reports by referencing Section 7 of the form.
Normative Scores and Their Meaning
Within this section, you’ll find details on three kinds of normative scores provided by the GARS-3: percentile ranks, scaled scores, and Autism Indexes. It explains how these scores were calculated and offers guidance on their interpretation.
Percentile Ranks
These ranks, ranging from 0 to 99, indicate the percentage of the normative sample scoring at or below a specific point. They are commonly used but may not represent equal intervals of information, particularly as they move away from the mean. Percentile ranks for the GARS-3 are provided in Appendixes A and B in the manual.
Scaled Scores
Derived by transforming raw scores to a mean of 10 and a standard deviation of 3, scaled scores offer interval data that can be directly compared with other standard scores. They are particularly useful for comparing an individual’s performance across different subscales of the GARS-3 and can be manipulated statistically, making them ideal for research purposes.
Composite Autism Indexes
Comprising the sum of scaled scores from either four or six subscales, the Autism Index provides a standardized measure of autism spectrum disorder severity. It is calculated to have a mean of 100 and a standard deviation of 15. There are two indexes available: one for individuals with nonverbal or severely impaired communication skills and another for those with verbal communication skills. Both indexes are reliable and valid, aiding in diagnostic decisions and severity assessments of ASD.
Using the GARS-3 to Determine the Likelihood of Autism Spectrum Disorder
The Autism Index is a crucial measure for assessing the likelihood and severity of ASD. Scores below 55 indicate a significantly low likelihood of ASD, with the majority of the normative sample scoring 55 or higher. The following table provides guidelines for interpreting Autism Index scores and estimating the severity level of ASD. Very few cases in the norming sample had scores below 55, suggesting potential misdiagnosis or significant improvement in behavior.
| Guidelines for Interpreting Autism Index Scores | ||||
| Autism Index | ≤54 | 55-70 | 71-100 | ≥101 |
| Probability of ASD | Unlikely | Probable | Very likely | Very likely |
| DSM-5 severity level for ASD | Level 1 | Level 2 | Level 3 | |
| Descriptor | Not ASD | Minimal support required | Requiring substantial support | Requiring very substantial support |
An Autism Index ranging from 55 to 20 signifies being 2.10 to 3 standard deviations below the normative sample’s mean for individuals identified with autism spectrum disorder. Within the normative sample, 77 cases (4% of the sample) fell into this range, indicating individuals at the lower end of the autism spectrum. When the Autism Index falls between 55 and 70, the likelihood of an autism spectrum disorder diagnosis is probable. Individuals within this range are typically highly functioning and may exhibit behaviors more aligned with typical development. However, caution is warranted as other disabilities such as intellectual disability (ID), attention-deficit/hyperactivity disorder (AD/HD), and others may manifest similar behaviors to those with ASD. Therefore, additional evidence from various sources should be gathered to clarify the diagnosis when the Autism Index is within one standard deviation from 55.
An Autism Index between 71 and 100 indicates being 1 to 2 standard deviations below the normative sample’s mean, strongly suggesting an autism spectrum disorder diagnosis. Individuals scoring between 71 and 84 are positioned on the lower end of the autism spectrum, previously categorized as having “mild autism” or “high functioning autism.” They are typically verbal and academically capable but may struggle with social interaction and communication. Scores ranging from 85 to 100 also strongly indicate autism spectrum disorder, suggesting significant autistic behaviors impacting academic and social interactions. However, individuals within this range should still be able to make satisfactory progress with less intensive educational and behavioral support compared to those with higher scores.
Individuals with Autism Indexes of 101 or higher display significant autistic behavior, leaving little doubt about the diagnosis. The probability of these individuals having autism spectrum disorder is very high, indicating a need for substantial individual attention and intensive academic and behavioral programming to support functioning.
References:
- American Psychiatric Association. (2012). Proposed revision: 299.00, autistic disorder. Washington, DC:
- Autism Society of America. (2003). Autism facts. Retrieved from http://www.autism-society.org
- American Psychiatric Association. (2012). Proposed revision: 299.00, autistic disorder. Washington, DC: Author.
- American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author.
- Centers for Disease Control and Prevention. (2012). FACTS: Facts about ASD. Retrieved from http://www .cdc.gov/ncbddd/actea.
