What Adult Autism Screening Tools Can and Cannot, Tell Us

A client walks into session and says:

“I took several autism tests online, and they all say I’m autistic.”

What do we do with that information?

Dismissing the results entirely is unlikely to be helpful. The client may have spent months—or years, trying to understand experiences that have never quite made sense to them. Completing screening measures may be part of a much larger process of trying to understand themselves.

But accepting a screening score as diagnostic creates a different problem. Adult autism screening measures can be useful clinical tools. They can identify patterns worth exploring, provide language for experiences clients have struggled to describe, and help clinicians consider whether further assessment may be appropriate.

But different measures ask different questions, were developed for different purposes and populations, and have important limitations. The most useful way to think about them is this:

A screening measure can help us ask better questions. It cannot answer the entire diagnostic question.

Screening and Diagnosis Are Not the Same Thing

Before looking at individual measures, it is important to distinguish screening from diagnosis.

A screening instrument generally identifies characteristics associated with a condition or people for whom additional evaluation may be warranted. A comprehensive autism assessment asks a much larger question: whether a person's developmental history, current presentation, functional experiences, and available evidence are consistent with diagnostic criteria while also considering alternative and co-occurring explanations.

That distinction becomes particularly important in adult mental health settings because many experiences associated with autism overlap with other clinical presentations.

Social difficulties can occur with autism, social anxiety, trauma, depression, ADHD, personality disorders, and other conditions. Difficulty with change can have multiple explanations. Sensory sensitivities are not exclusive to autism. Neither are social exhaustion, emotional dysregulation, repetitive behaviors, or feeling different from other people.

A questionnaire score therefore needs context.

For counselors working with adults, particularly women and people whose autism may not have been recognized earlier, five measures are especially useful to understand:

AQ, RAADS-R, RAADS-14, GQ-ASC, and CAT-Q.

They are not interchangeable.

And none should be treated as a diagnostic verdict.

The Autism Spectrum Quotient (AQ)

The Autism Spectrum Quotient (AQ) was developed by Baron-Cohen and colleagues (2001) as a self-report measure of autistic traits in adults. The wording matters:

autistic traits.

The AQ contains 50 items examining five broad areas: social skills, attention switching, attention to detail, communication, and imagination (Baron-Cohen et al., 2001). Brown and colleagues (2020) similarly describe it as a self-administered measure designed to assess the degree to which an adult endorses traits associated with autism.

For counselors, the AQ may be useful when a client is beginning to explore whether longstanding experiences could fit within an autistic pattern. Individual responses can create openings for conversations about social interaction, communication, routines, attention, and developmental experiences.

But a high AQ score does not mean:

“This person has autism.”

A more appropriate interpretation is:

“This person endorsed a number of autistic traits. What can we learn by exploring those responses further?”

There is another important consideration for the population we are discussing.

Brown et al. (2020) noted that autistic traits as displayed by girls and women may not always be effectively measured by the AQ because of sensitivity concerns.  That does not make the AQ useless.

It means that a score, whether high or low, should not replace clinical curiosity.

The RAADS-R

The Ritvo Autism Asperger Diagnostic Scale–Revised (RAADS-R) is one of the adult autism measures clients are increasingly likely to encounter online. It is also a good example of why clinicians need to look beyond a cutoff score. The original validation study reported striking psychometric findings, including high sensitivity and specificity (Ritvo et al., 2011).

Those numbers can make the RAADS-R sound almost definitive.

But the context in which a measure was validated matters. The original validation procedures were not equivalent to someone independently completing an online questionnaire and receiving an automated result. The RAADS-R was studied within a broader clinical and diagnostic context.

Subsequent research has also raised concerns about relying on RAADS-R scores as stand-alone predictors of adult autism diagnosis. In clinical referral populations, its performance has not always replicated the very high discriminative ability reported in the original validation work (Jones et al., 2021). This does not mean the RAADS-R has no clinical value.

It means we should use it for what it can provide:

Information to explore, not a verdict to accept.

If a client brings in a high RAADS-R score, the next clinical question is not necessarily:

“How high was your score?”

It may be:

“Which questions made you think, ‘That describes me’?”

The RAADS-14 Screen

The RAADS-14 Screen is a shorter instrument derived from the RAADS-R and developed specifically as a screening measure for autism in adult psychiatric populations (Eriksson et al., 2013). Its purpose is particularly relevant to counselors. The RAADS-14 is intended to help identify people who may warrant further autism assessment.

That is screening. It is not diagnostic confirmation.

This also gives us an opportunity to distinguish two concepts clinicians frequently encounter when reading assessment research. Sensitivity refers to how effectively a measure identifies people who actually have the condition. Specificity refers to how effectively it identifies people who do not have the condition. For a screening measure, high sensitivity can be particularly valuable because we do not want to miss people who may warrant additional evaluation.

But there is a tradeoff.

When specificity is lower, some people who screen positive will not ultimately meet diagnostic criteria. That is not necessarily evidence that the screening tool “failed.” It reflects what screening is designed to do.

A positive screen means “look more closely,” not “diagnosis confirmed.”

The GQ-ASC

The Girls Questionnaire for Autism Spectrum Condition (GQ-ASC) is especially relevant to this discussion because it attempts to capture aspects of autistic presentation in girls and women that may be less adequately represented by traditional autism measures. Brown et al. (2020) evaluated a modified version of the GQ-ASC specifically as an autism screening measure for adult women. Their study included 672 cisgender and trans women between 18 and 72 years old, including 350 autistic and 322 nonautistic participants.

The researchers identified five components:

Imagination and play — experiences involving fantasy, fiction, and reflection on childhood imaginative play.

Camouflaging — effortful attempts to reduce the visibility of autistic characteristics.

Sensory sensitivities — hyper- and hyposensitivities across sensory modalities.

Socializing — barriers involving social understanding and participation.

Interests — including age-advanced and nonstereotypically feminine interests.

These areas are clinically interesting because they may capture experiences that do not necessarily resemble stereotypical presentations of autism. For example, some of the camouflaging items reflected consciously observing and copying other women or adopting different personas in different situations. The socializing component included being able to socialize successfully for a period of time but subsequently feeling exhausted.

That is a very different clinical picture from simply asking:

“Does this person have social skills?”

The better question may be:

“What is required for this person to successfully navigate a social interaction?”

Autistic and nonautistic women differed significantly across all five components and on the total GQ-ASC score. Brown et al. (2020) reported an area under the receiver operating characteristic curve of 0.89, indicating strong discrimination between the groups in their sample.

The authors concluded that the modified GQ-ASC showed promise as a screening tool for adult autistic women and suggested that it could potentially help practitioners identify women who may benefit from referral for further assessment.

But there are important limitations.

The study used online self-report data. Participants were recruited through social media and online autism communities and support groups. The AQ was used as part of the process of distinguishing the autistic and nonautistic groups despite the researchers themselves noting sensitivity concerns with the AQ for women. The researchers also did not collect information about co-occurring intellectual disability.

Brown et al. therefore called for additional research and refinement, including research with clinical populations.  That makes the GQ-ASC promising and particularly relevant to the population we are discussing.

It does not make it a stand-alone diagnostic instrument.

The CAT-Q

The Camouflaging Autistic Traits Questionnaire (CAT-Q) asks a somewhat different question from the other measures in this article. It is not primarily an autism screener. It was developed to measure social camouflaging behaviors (Hull et al., 2019).

The 25-item CAT-Q includes three broad domains:

Compensation — strategies used to compensate for difficulties navigating social situations.

Masking — strategies used to conceal autistic characteristics or present oneself in ways that more closely conform to social expectations.

Assimilation — strategies related to trying to fit into social situations.

That distinction is extremely important.

A high CAT-Q score does not establish:

“This person is autistic.”

Instead, it provides information relevant to a different question:

“How much effort is this person putting into managing how they appear socially?”

For a client whose external presentation seems inconsistent with their description of their internal experience, that can be valuable information. Someone may appear socially fluent while consciously calculating every part of the interaction. Someone may make eye contact while finding it distracting or uncomfortable.

Someone may appear relaxed while actively monitoring their posture, facial expression, tone of voice, and conversational timing. Someone may successfully navigate an entire social event and then need hours, or longer, to recover. The CAT-Q can help bring those less-visible processes into the clinical conversation.

But camouflaging should not itself be treated as proof of autism. Again:

The measure tells us something worth exploring. It does not tell us everything we need to know.

Five Measures, Five Different Questions

Measure	What It Can Help Explore	What It Does Not Establish
AQ	Self-reported autistic traits across several domains	An autism diagnosis
RAADS-R	Autistic characteristics that may warrant deeper exploration	Diagnosis based on a score alone
RAADS-14	Whether further autism assessment may be worth considering, particularly in psychiatric populations	Diagnostic confirmation
GQ-ASC	Characteristics described in autistic women, including camouflaging, sensory experiences, socializing, interests, and childhood imagination/play	A stand-alone autism diagnosis
CAT-Q	Camouflaging through compensation, masking, and assimilation	Whether someone is autistic

This is why asking:

“Which autism test is best?”

may not be the most useful question.

A better question is:

“What am I trying to understand?”

“But I Took Six Tests and They All Say I’m Autistic”

Now return to the client from the beginning. Suppose someone brings you screenshots from multiple online autism measures. Their scores are elevated. They have read extensively about autism. And they tell you:

“Everything finally makes sense.”

There are two responses we should probably avoid. The first is:

“Online tests aren't valid. You need a real assessment.”

That can unnecessarily dismiss what may represent months or years of meaningful self-exploration. The second is:

“Then you're autistic.”

That asks screening instruments to answer a question they cannot independently answer. There is another option.

Become curious.

Ask:

“What made you start looking into autism?”

“What did you recognize in yourself?”

“Which questions felt particularly familiar?”

“Did anything about your childhood begin to make more sense?”

“Are there parts of descriptions of autism that don't fit?”

“What would knowing whether you're autistic mean to you?”

Now the screening results have become what they may be most useful for a starting point for a richer clinical conversation.

A Low Score Deserves Context Too

There is another side to this issue that is easy to overlook. If a high score is not proof of autism, a lower score should not automatically become proof that autism is absent. Self-report measures depend on the respondent's interpretation of questions and awareness of their own experiences.

An adult who has spent decades adapting to social expectations may not recognize some of their behaviors as compensatory. They may assume everyone consciously studies conversations. They may assume everyone rehearses what they are going to say before making a telephone call. They may assume everyone returns home from social events completely exhausted. They may answer a question based on what they are capable of doing, rather than what doing it costs them.

And some measures may capture certain presentations better than others. Brown et al. (2020), for example, found that the GQ-ASC and AQ appeared to measure somewhat different constructs associated with autism and raised the possibility that this may partly reflect gendered differences in autistic presentation or in the development of the measures themselves.

So the question should not be:

“Did the client pass or fail the autism test?”

Instead ask:

“What information does this result add to everything else we know?”

Use the Answers, Not Just the Total

For counselors, one of the richest parts of a screening measure may not be the final score.

It may be the client's answers.

Suppose someone strongly endorses an item involving social exhaustion.

Ask about it.

What makes socializing exhausting?

When did that begin?

What happens during the interaction that other people might not see?

How much preparation is required?

What happens afterward?

Suppose someone endorses sensory difficulties.

Ask what sensory experiences are difficult.

When did they begin?

What does the client do to manage them?

Have they spent years forcing themselves to tolerate environments they find overwhelming because they assumed everyone experienced them that way?

The number summarizes the responses.

The client's story explains them.

Screening Should Expand Clinical Reasoning, Not Replace It

None of these measures eliminates the need for clinical reasoning. And none can independently reconstruct a developmental history. A comprehensive adult autism evaluation may include developmental information, behavioral observation, clinical interviews, psychiatric and medical history, self-report information, consideration of differential and co-occurring diagnoses, and other sources of information depending on the evaluator and individual circumstances.

The larger context matters because the same behavior can arise from different processes. A person may avoid social situations because of social anxiety. Another may avoid them because sensory and communication demands are exhausting. Another may have significant trauma associated with interpersonal situations. And another may be autistic and have social anxiety and have trauma. A screening score cannot make those distinctions for us.

Clinical reasoning still matters.

What Can I Do Differently on Monday?

You do not need five questionnaires waiting on your desk. Instead, remember three things.

First, know what the measure actually measures. An autism-trait screener, a psychiatric-population screener, a measure developed around presentations in women, and a camouflaging measure are not interchangeable simply because they all appear in conversations about adult autism.

Second, treat the score as the beginning of a conversation rather than the end of an assessment. Ask which responses resonated, when those experiences began, how they appear across settings, and what they cost the client.

Third, resist the urge to use screening measures as gatekeepers. A high score does not establish autism. A low score does not necessarily resolve every clinical question.

Screening should increase curiosity not end it.

The Bottom Line

Adult autism screening measures can be useful. They can provide language. They can reveal patterns. They can help organize developmental questions. They can identify experiences that deserve further attention. And they may help clinicians and clients decide whether formal assessment is worth considering.

What they cannot do is replace the larger clinical picture.

So when a client brings you an autism screening score, you do not have to dismiss it. And you do not have to diagnose from it. You can do something counselors already know how to do:

Get curious about the story behind the answer.

Screening ≠ diagnosis.A score is information—not a conclusion.

Continue the Series

Part 1 — RECOGNIZE: When Therapy Isn't Enough: Recognizing When Autism May Deserve Another Look

Part 2 — EXPLORE: What Adult Autism Screening Tools Can—and Cannot—Tell Us

Part 3 — REFER: Helping Adults Navigate Autism Assessment

Part 4 — ADAPT: Making Therapy More Accessible Without Waiting for a Diagnosis

References

Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The Autism-Spectrum Quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31(1), 5–17. https://doi.org/10.1023/A:1005653411471

Brown, C. M., Attwood, T., Garnett, M., & Stokes, M. A. (2020). Am I autistic? Utility of the Girls Questionnaire for Autism Spectrum Condition as an autism assessment in adult women. Autism in Adulthood, 2(3), 216–226. https://doi.org/10.1089/aut.2019.0054

Eriksson, J. M., Andersen, L. M. J., & Bejerot, S. (2013). RAADS-14 Screen: Validity of a screening tool for autism spectrum disorder in an adult psychiatric population. Molecular Autism, 4, 49. https://doi.org/10.1186/2040-2392-4-49

Hull, L., Mandy, W., Lai, M.-C., Baron-Cohen, S., Allison, C., Smith, P., & Petrides, K. V. (2019). Development and validation of the Camouflaging Autistic Traits Questionnaire (CAT-Q). Journal of Autism and Developmental Disorders, 49(3), 819–833. https://doi.org/10.1007/s10803-018-3792-6

Jones, S. L., Johnson, M., Alty, B., & Adamou, M. (2021). The effectiveness of RAADS-R as a screening tool for adult ASD populations. Autism Research and Treatment, 2021, Article 9974791. https://doi.org/10.1155/2021/9974791

Ritvo, R. A., Ritvo, E. R., Guthrie, D., Ritvo, M. J., Hufnagel, D. H., McMahon, W., Tonge, B., Mataix-Cols, D., Jassi, A., Attwood, T., & Eloff, J. (2011). The Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R): A scale to assist the diagnosis of autism spectrum disorder in adults: An international validation study. Journal of Autism and Developmental Disorders, 41(8), 1076–1089. https://doi.org/10.1007/s10803-010-1133-5