Test Details

Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R)

The RAADS-R is a reliable instrument for assisting in the diagnosis of autistic adults, particularly addressing the gap in screening services for those with average or above IQ who may 'escape' traditional diagnosis methods. It has proven to be highly accurate in discerning between autism and other similar diagnoses, with excellent concurrent validity, sensitivity, specificity, and test-retest reliability. However, some studies have raised questions about its predictive validity, emphasizing the importance of using multiple measures and interviews during the diagnostic process.

Total scores on the RAADS-R range from 0 to 240, with a threshold of 65 indicating a likely autism diagnosis, while scores below 65 suggest a 97% chance of not being autistic. Recent average total scores for autistic males are 148, and for females, it is 160.

Social Responsiveness Scale (SRS-2)

The SRS-2 focuses on identifying and determining the severity of social difficulties within autism and differentiating it from similar disorders, such as ADHD and anxiety. It is sensitive to subtle symptoms, making it ideal for use with adults who may be autistic but have gone undiagnosed. The SRS-2 is also useful for addressing comorbid and/or differential diagnoses.

In a study by South et al (2017), it was reported that the average SRS-2 total percentile score for their sample of 40 autistic adults was 93. Subscale scores are intended to guide treatment more than assessment, though they hold diagnostic interest. Higher scores on a subscale indicate increased difficulty in that context.

A 2020 study found that the SRS-2 was strongly correlated with the ADOS2, considered the gold standard for assessing autistic adults, suggesting the SRS-2 may be useful in autism assessment. However, South’s 2017 study explained that when a person experiences high anxiety, the SRS-2 cannot be used in isolation for diagnosis—an obvious conclusion, as no diagnosis should rely solely on one test.

The SRS-2 has been studied in various cultures, including Korea, Turkey, Lithuania, Iran, and Uganda. For example, a 2021 study from Uganda found that the SRS-2 had reasonable reliability and validity when distinguishing between neurotypical and autistic individuals.

Comprehensive Autistic Trait Inventory (CATI)

The Comprehensive Autistic Trait Inventory (CATI) is a promising new test created in 2021 (English et al., 2021) that covers a broad range of autistic traits. Developed in response to concerns that popular tests such as the AQ, BAPQ, and SRS-2 may not capture some aspects of the autistic experience effectively, the CATI aims to fill crucial gaps in assessing various autistic trait dimensions, resulting in fewer measurement 'blind spots' compared to similar tests.

The authors of the CATI noted that autistic females are more likely to engage in 'masking' behaviors and other strategies to fit in with their peers, and that autistic males often had slightly higher scores on the communication and repetitive behavior subscales. Autistic women in their sample had slightly higher scores on the sensory sensitivity subscale.

Total scores range from 42 to 210, with scores of 134 and above indicating the likely presence of autistic traits. Subscale scores range from 7 to 35, with higher scores in a given trait area suggesting more neurodivergence.

The CATI's developers acknowledge that the test has yet to be validated by other researchers. However, their own studies found the CATI to be slightly more effective at classifying autistic and neurotypical people, and they reported excellent convergent validity—meaning that results from the CATI strongly agree with results from similar well-established tests.

In summary, the CATI is a well-written, promising new test that addresses a broader range of autistic traits, aiming to reduce measurement blind spots and improve the assessment of various autistic trait dimensions.

Modified Girls Questionnaire for Autism Spectrum Condition 2020 (GQ-ASC)

The GQ-ASC test, developed in 2011 and updated in 2020, aims to more accurately identify autistic cisgender and trans women, a population often under-diagnosed and misdiagnosed. This test gestures toward a more modern understanding of autistic adults' experiences, especially for those who are late-diagnosed. While the test is primarily designed for women, it remains useful for any adult suspecting they may be autistic. Users can answer the questions as written or use gender-neutral language where needed.

Scores range from 21 to 84, with a threshold of 57 or above indicating a high level of autistic traits. This threshold is sensitive to 80% of cases. Higher subscale scores signify an increased presence of a particular autistic trait. The GQ-ASC Imagination and Play subscale notably correlates strongly with the AQ Imagination subscale. The maximal scores on each subscale are 16, except for the imagination/play subscale, which is 20.

The 2020 update reported that the GQ-ASC, on its own, correctly identifies 80% of autistic cases. However, due to its recency, the test has not been extensively studied, and the validity has not been thoroughly assessed by other researchers.

Sensory Perception Quotient (SPQ)

The SPQ is a reliable and valid tool for measuring sensory sensitivities in adults. As the DSM-5-TR's B.4 focuses on sensory criteria, a thorough autism assessment should gather data on sensory hyper and hypo-sensitivities across all five major senses.

The original 2014 paper where the SP-Q was developed and tested reported good concurrent validity and the ability to differentiate between autistic and neurotypical adults. They also stated that elevated sensory sensitivity is associated with a person having 'more autistic traits'. This 2017 study found the SP-Q reliable and useful in detecting sensory over-responsivity, while a 2020 study demonstrated its excellent convergent validity with the respected Sensory Gating Inventory (SGI).

Furthermore, a 2020 follow-up study (Taylor, et al.) on the SPQ revealed that females with greater degrees of autistic characteristics also had greater hypersensitivity when compared to neurotypical and broad autism phenotype control groups. This finding lends support to autism's criteria B.4 (sensory sensitivity) and is likely reasonable to generalize to all autistic individuals, not just women as studied.

Toronto Alexithymia Scale (TAS-20)

The TAS-20 is one of the most common measurements for 'alexithymia', which means having difficulty identifying and describing one's own emotional experience. A recent meta-analysis (Kinnaird et al., 2019) found that approximately 50% of autistic people have alexithymia, compared to 5% in neurotypical populations. Compelling research has also suggested that difficulty with eye-contact is related to alexithymia rather than autism, contrary to popular belief.

A 2018 review and meta-analysis indicated that emotional processing difficulties may reflect co-occurring alexithymia instead of being a core feature of autism. Autistic individuals with alexithymia could represent a specific subgroup of autism who may benefit from tailored interventions. Given the prevalence of alexithymia in the autistic population, it's an important potential diagnostic marker—and it's typically very helpful for individuals to be aware of it.

Total scores on the TAS-20 range from 20 to 100, with higher scores indicating a higher likelihood of alexithymia. Scores below 51 likely indicate no alexithymia, scores of 52 to 60 suggest possible alexithymia, and scores above 60 indicate that alexithymia is very likely present. Higher subscale scores indicate which particular area a person has the most difficulty with, though there are no official threshold numbers available for subscales.

In 2020, several of the TAS-20's original authors reviewed the extensive research conducted on the TAS-20 in the 25 years since its initial creation. They concluded that it is still viewed as a reliable and valid measure of alexithymia.

Autism Quotient, Adult version (AQ)

The AQ is a widely used and respected test for adults of normal intelligence where autism is suspected, looking at a broad range of autistic traits. It's well known, very frequently used in research, and often one of the first screening tests completed by a person who suspects they may be autistic. This test was created to be a descriptive measure of autistic characteristics, rather than a purely diagnostic measure. Due to its extensive research, we have many data points about how AQ scores correlate to scores on other psychometrics related to autism.

Total scores range from 0 to 50, and a score of 32 or more was initially suggested as indicative of clinically significant autism traits. However, more recent research (Broadbent, 2013) found that total AQ scores of 29 or higher are more accurate. The original study (Baron-Cohen, 2001) reported average AQ scores of 35.1 for autistic men and 38.1 for autistic women. Subscales were each represented by 10 questions, and a higher score (closer to 10) on each subscale indicates more neurodivergence in that area.

A systematic review of literature on the AQ from Ruzich (2015) looked at 73 articles with almost 9000 participants. They reported the average total score for non-clinical populations was 16.9, and the average total for autistic people was 35.2. A 2017 Swedish study concluded that the AQ has "adequate sensitivity and specificity to distinguish people with ASD from those without ASD."

Wheelwright (2006) suggested a total AQ score of 23 is useful for evaluating sub-threshold autistic traits, which Ruzich (2015) endorsed. The AQ has been studied quite a bit in multicultural populations, such as Sweden, Turkey, China, India, and so on, with results ranging from positive to very positive. Many attempts have been made to create shorter or better versions of the AQ-50, but it still has good value in assessing adults due to its sensitivity and the significant number of comparisons to other tests possible because of the extensive research involving the AQ.

Adult Repetitive Behaviour Questionnaire-2 (RBQ-2A)

The RBQ-2A is a newer test (Barrett et al., 2015) that assesses subtypes of restricted and repetitive behaviour (RRB) in adults. It is of particular use with adults who may be autistic because RRB is one of the key criteria when diagnosing autism. Further, most existing research of this kind investigated RRB in children, so having this newer test for adults is more appropriate than merely adapting an existing test for children for use with adults.

Total scores range from 20 to 60, and 26 is the threshold above which indicates that a substantial level of RRBs are present. The average total score of autistic adults in Barrett’s sample was 36, whereas the average total for non-autistic adults was 25. No significant gender differences were observed (table 7).

The original research from 2015 on developing the RBQ-2A is an adaptation of a 2006 version that focused on repetitive behaviours in children. Barrett was involved in a 2017 study on adapting the RBQ-2A for use with adolescents and parent-reporting, a use in which they found the RBQ to be reliable. In 2018, Barrett lead a study comparing the RBQ-2A to the AQ, reporting that it is 'a reliable and valid self-report measure of repetitive and restrictive behaviours.'

A rigorous study from 2019 looked specifically at the validity and psychometric properties of the RBQ-2A, which they found acceptable. The authors did suggest a slightly modified version may be superior, made simply by removing several test questions they believed weren't as useful (due to having poor loadings in their factorial validity). A fascinating study from 2021 examined the "effects of alexithymia, intolerance of uncertainty, and anxiety on the relationship between sensory processing differences and restricted and repetitive behaviours in autistic adults". The authors utilized the RBQ-2A and reported that it has good psychometric properties for use in autistic samples as well as having "acceptable-good" internal consistency.

Camouflaging Autistic Traits Questionnaire (CAT-Q)

The CAT-Q measures a variety of autistic adult's social "masking" strategies, sometimes also called "camouflaging". Adults who appeared to function well enough socially through into adulthood without being diagnosed as autistic earlier in life are almost certainly utilizing substantial levels of masking. It's a common experience for late-diagnosed adults to have become so used to masking that it's quite difficult for them to 'take the mask off', and to feel somewhat at a loss about their social identity. This 2018 test was created based on the experiences of autistic adults' social "masking" strategies, also known as 'social camouflaging'.

The threshold over which total scores indicate a person socially masks autistic traits is 100. Higher subscale scores represent greater levels of camouflaging. In this context, assimilation refers to the strategies a person uses in a social situation to try and fit in during uncomfortable situations, and compensation refers to the variety of strategies an autistic person might employ in social situations to compensate for their social difficulties (Hull, 2018).

Internal consistency for the CAT-Q has been found to be very good, and test-retest reliability was acceptable (r=.77). Further, this same study looked at how the CAT-Q compared to other more established tests (i.e., convergent validity), and they reported that "the total CAT-Q score and all CAT-Q factors were significantly positively correlated with autistic-like traits and social anxiety in autistic and non-autistic samples, with the exception of the Masking factor, which was not significantly related to autistic-like traits in the autistic sample." See here for a more thorough discussion on reliability and validity for the CAT-Q.

The CAT-Q has been translated and studied numerous times. For example, a 2022 Italian translation also found the CAT-Q to have great internal consistency and test-retest reliability. For convergent validity, they compared the CAT-Q against the well-known Adult Autism Subthreshold Spectrum (AdAS Spectrum) and found it to be strongly correlated. Seeing results confirmed in other cultures is useful because my clients who do these assessments come from all sorts of backgrounds.

Prodromal Questionnaire-Brief Version (PQ-B)

The PQ-B is a self-report that is effective and efficient to use when screening for prodromal (i.e., early signs of) psychosis syndromes. It focuses on 'positive' psychotic symptoms (i.e., additive symptoms like delusions) rather than 'negative' symptoms. This screener is intended to be followed by a clinical interview, and the clinical interview I do during my autism assessments does attend to both a family and personal history of mental illness and psychosis.

Scores range from 0 to 105, with scores of 24 or above being the threshold which indicates a person may be at risk of developing psychosis. Some questions in the PQ-B can be confounded by trauma, social anxiety, as well as sensory sensitivities, and the interview addresses this.

The initial study found the PQ-B to have good concurrent validity, and researchers stated it was an effective and efficient screener, though it should be used with a follow-up interview. A 2016 study completed in China reported that the PQ-B is a useful screening instrument, though some questions seemed to have more predictive validity than others. This 2016 follow-up study looked at the PQ-B's psychometric qualities and found it has good internal consistency, "a high degree of overlap between self-reported clinical high-risk symptoms", and said it was an easy-to-use and reliable instrument for screening high-risk early symptoms of psychosis.

Adverse Childhood Experiences Quiz (ACE)

The ACE quiz is a well-known self-report using 10 questions about difficult early life experiences, mostly related to abuse, neglect, substance use, and loss. The CDC was heavily involved in the research, and there is lots of data showing the extraordinary relationships between ACE scores and adult health and social outcomes. This quiz is included to consider comorbidities and differentials.

Autism and trauma seem to have strong links that research is only now beginning to quantify. For example, “autistic children are more reactive to stressful events and, because they lack the coping skills that help them calm down, perhaps predisposed to PTSD” (Hoover, 2018). Until recently, the research consensus was that autistic people had a prevalence of PTSD similar to the rest of the population: 3%, and that this would “be one of the only psychiatric conditions that’s no more common in people with autism than in their typical peers,” (Kerns, 2018, here or here). Autistic children are bullied 3 to 4 times more than NT children, are at enhanced risk for ACEs (especially parental divorce and insufficient income), and having a higher ACE score often means a delay in when the person gets their autism diagnosis and any related support. Having a higher ACE score puts a person at an “elevated risk for comorbid psychiatric and medical health problems,” (Hoover, 2018).

There is no cut-off score with the ACE quiz, but there is data from enormous samples about how health outcomes in adulthood decline as ACE scores increase. 61% of adults experienced at least one ACE, and 16% have experienced 4 or more of the ACE types. This doesn’t mean a child was abused on 4 separate occasions; it means that (for example) they were abused, had an alcoholic parent, often didn’t have enough to eat, and had someone in their home go to jail.

In general, higher ACE scores can leave a person feeling vulnerable, which leads a significant portion of them to embrace risky social or health behaviours in attempts to seek comfort from traumatic experiences. The kinds of risky behaviours involved often lead to disease, disability, poor health, and early death. Females and some minority groups are most at risk of experiencing 4 or more ACEs. See here for more information.

International Cognitive Ability Resource (ICAR16)

IQ tests are sometimes included in autism assessments due to wanting test data to inform one of the DSM-5’s specifiers “With or without accompanying intellectual impairment.” In the case of late-diagnosed autistic adults, my current opinion is that if a person gets to adulthood without someone (a parent, teacher, doctor, etc) insisting they get pretty thoroughly tested, then the adult in question likely has at least average intelligence. Thus, extensive IQ testing is usually unnecessary in our case. No online IQ test is great; to test intelligence you need a lengthy, thorough test like the WAIS-IV. If we had further reason to test IQ, I have a second, more thorough test I could administer.

However, I offer the option to clients if they’d like to do the ICAR16, which is a short series of letter-number series tasks to measure fluid reasoning, whereas verbal reasoning, matrix reasoning, and 3D reasoning tasks measure visual-spatial reasoning. It was created by the University of Cambridge and Northwestern University’s Psychology department. The ICAR16 is a valid measure of nonverbal intelligence. Scores are out of 16 and are presented to you as a percentile.

Adult ADHD Self-Report Scale (ASRS-5)

The ASRS-5 was updated in 2017 from the previous ASRS 2005 version. It can be used as a starting point to help recognize the signs/symptoms of Adult ADHD but is not meant to replace a full ADHD assessment. It can also be useful as part of an autism-only assessment.

Scores can range from 0 to 24. Scores above 14 indicate a person likely does have ADHD, and scores below 14 indicate they probably do not. This screener has been widely studied, including a 2021 paper which found the ASRS-5 to be reliable and psychometrically sound when looking at ADHD in adolescents and adults. Additionally, it has been translated into many languages (e.g., French, Turkish, Serbian) and research on these translations also found good specificity, validity, and reliability.

Weiss Functional Impairment Rating Scale Self-Report (WFIRS-S)

This test was originally developed for ADHD, helping examine the difference between ADHD symptoms and “actual impairment,” which are related but different concepts. A person could have one and not the other, or both. Since its inception, this test has also been used in non-ADHD research, and it's suggested to be relevant to autistic people as well. It is used in this assessment to give a more specific look at a person’s level of functioning, which can be relevant to autism as well as ADHD.

The original scoring instructions suggest that any subscale’s score (average) that is over 1.5 suggests “impairment” in that context.

Kessler Psychological Distress Scale (K10)

The K10 is a simple and widely utilized measure of adult distress, both for general and clinical populations. It uses 10 questions about a person's emotional state on a 5-point Likert scale. It is used here to give an indication of a person's current wellbeing, which has an impact both on recommendations given as well as a look at their state of mind when answering questions in the other tests.

Total scores range from 10 to 50, with higher scores indicating feeling more distressed. Total scores 25 or above indicate clinically significant levels of distress. Scores over 30 are into the moderate/severe range. Elevated subscale scores aren't necessarily indicative of autism, though elevated scores do indicate that area of a person's life is causing distress. The cause of the relevant concerns may be based in autistic traits, though more detail would be required to know with certainty.

This study reported that the K10 has good psychometric properties, particularly that it has good convergent validity, internal consistency, and accuracy. Another study also reported that the K10 has high levels of validity and reliability when looking for affective disorders (e.g., depression), though they used a slightly higher than normal cutoff score of 27.

PTSD Checklist for DSM-5 (PCL-5)

The PCL-5 is a 20-item self-report measure that assesses the 20 DSM-5 symptoms of PTSD. The PCL-5 can be used to screening for PTSD and making a provisional PTSD diagnosis, though a full assessment for PTSD would include an interview as well.

Total scores range from 0 to 80. A total score threshold has been suggested at approximately 31 to 33, but diagnosing PTSD is more complex than a single score and has specific requirements within each cluster of the criteria. Again, an interview is required for a formal diagnosis.