Show the full module text

Module 10 — OAQ-G2 — A Longer Look at Alexithymia

A thirty-seven item questionnaire that asks about alexithymia across seven areas rather than three, what its four extra domains add and confuse, and why a tool with no published validation can still be worth an hour of your thinking.

Autistic Self-Discovery · Part Four — Feeling, Naming, Connecting · 13 min read, about 26 min with the workbook

The big idea

If you took the TAS-20 in Module 9 and came away feeling the questions had missed something — that your particular version of this was wider than three subscales could hold — this module is the longer look. Thirty-seven questions instead of twenty, seven areas instead of three. It is also, and this is the part nobody else will tell you, a questionnaire that has never been tested.

That is what thirty years of scrutiny buys you: not a perfect instrument, but a precise map of where it fails. The TAS-20 has a weak subscale and we know exactly how weak, in which population, by how much. For the OAQ-G2 there is no map at all — not a good report and not a bad one, just no report. That is the difference between a measurement and a set of questions.

Step 1 — The lesson

A. Half the room

Alexithymia is the gap between having a feeling and having a word for it. Module 9 covers it properly: the delay of hours between something happening and knowing what it did to you, the question "how are you" that you cannot answer honestly because the information is not there yet. If you have not read that module, read it first. This one assumes it.

What matters here is the size of the overlap. A systematic review and meta-analysis pooled fifteen studies comparing 366 autistic adults with 348 non-autistic comparison participants, and found that 49.93 per cent of the autistic group met the standard alexithymia threshold against 4.89 per cent of the comparison group.¹ That is the source of the figure you see quoted everywhere as "about half". It is worth holding onto exactly what it says. Half. Not all. The authors were careful about this: alexithymia is common in autistic people rather than universal, and autistic people who also have it may form a distinct subgroup with distinct needs.

Diagram — A · About half, not all. The circles are not the same size and the middle is not the whole of either. Roughly half of autistic adults score above the alexithymia line, which means roughly half do not — and plenty of alexithymic people are not autistic at all.

That overlap is not a footnote. In 2013 Geoffrey Bird and Richard Cook argued that where autistic people show difficulty recognizing emotion or reduced empathy, those difficulties belong to co-occurring alexithymia rather than to autism itself. Alexithymia correlated above 0.60 with accuracy at reading emotion in the voice while autism severity predicted nothing, and empathic brain response tracked alexithymia in autistic and non-autistic participants alike.² If that argument is right — and it has reorganised this corner of the field — then knowing where you sit on alexithymia tells you more about your emotional life than knowing you are autistic does.

The word itself is older than any questionnaire. Peter Sifneos coined it in 1973 to describe patients who showed "a relative constriction in emotional functioning, poverty of fantasy life, and inability to find appropriate words to describe their emotions".³ Notice the middle clause. Poverty of fantasy life was in the original description, alongside the missing words. That detail is going to matter, because the questionnaire this module is about takes it seriously and the field standard does not.

B. Thirty-seven questions, seven doors

The Online Alexithymia Questionnaire was written in 2005 by Jason Thompson and revised in 2007 into the version you will meet, the G2. It has 37 statements, each answered on a five-point agreement scale, and its published maximum is 185. His stated reason for building it was that existing scales asked poorly targeted questions about imagination, confused not recognizing a feeling with not having one, and were locked away from the public behind licences.⁴ That third complaint was fair. It is free, and it always has been.

The 37 items divide into seven areas. In the original scoring they are Difficulty Identifying Feelings (six items), Difficulty Describing Feelings (four items), Vicarious Interpretation of Feelings (three items), Externally-Oriented Thinking (seven items), Restricted Imaginative Processes (seven items), Problematic Interpersonal Relationships (six items) and Sexual Difficulties and Disinterest (four items).⁴ Six and four and three and seven and seven and six and four is thirty-seven, which is at least internally consistent.

Diagram — B · Seven doors, one number. Every area feeds the same total, and the total is the only thing most people look at. Three of the seven are shared with the TAS-20. The other four are additions that no published study has ever checked, which means roughly half your total comes from questions of unknown quality.

Compare that to the field standard. The Toronto Alexithymia Scale, built by Bagby, Parker and Taylor in 1994, has twenty items in three subscales, selected from a longer pool by keeping only the items that held together statistically, then cross-validated in separate clinical and non-clinical samples.⁵ The OAQ-G2 keeps those three and adds four more.

Is that a reasonable thing to do? Genuinely, it might be. Sifneos's original description had four components, and one of them was constricted imaginal processes — the poverty of fantasy life. The TAS-20 dropped it. In 2024 Taylor, Porcelli and Bagby published a defense of the original four-component construct, arguing that there is insufficient evidence to justify removing the imaginal component from the definition of alexithymia at all.⁶ On the other side, David Preece and colleagues built the 24-item Perth Alexithymia Questionnaire on an attention-appraisal model that deliberately leaves fantasising out, on the grounds that difficulty fantasising is better understood as a correlate of alexithymia than as a part of it.⁷

So the specialists who have spent thirty years on this cannot agree on what alexithymia includes. That is not a scandal — constructs get argued about, and that is how they improve. It does mean that a questionnaire adding four extra domains is taking a position in an unfinished argument. The OAQ-G2 does not tell you it is doing that. It just adds the numbers up.

C. The four extra doors

Take the additions one at a time, because each one behaves differently when the person answering is autistic.

Restricted Imaginative Processes. This is the domain with the best theoretical claim to be there and the worst fit with the evidence about autistic people. The old stereotype of an impoverished autistic imagination has not held up. A study of 312 people found that higher autistic traits went with lower fluency on divergent thinking tasks — fewer ideas — but with an unusually high number of atypical, original responses, which the authors read as a possible adaptive advantage rather than a deficit.⁸ Fewer ideas, stranger ideas. A questionnaire that asks whether you produce a lot of daydreams will score that as restriction. And in a sample of 223 autistic adults recruited through the SPARK research cohort, 42 per cent scored above the cut-off on a measure of maladaptive daydreaming — immersive inner worlds vivid enough to interfere with daily life.⁹ Whatever is going on with autistic imagination, "restricted" is not a safe summary of it.

Diagram — C · The four extra doors. Every line in the left column is how the questionnaire records an answer. Every line in the right column is an equally likely reading of the same answer. A questionnaire that has never been validated has no way of telling you which column you are in, so it silently picks the left one.

Sexual Difficulties and Disinterest. This is the domain that appears in no consensus definition of alexithymia anywhere. It is also the one most likely to add points to an autistic person's total for reasons that have nothing to do with emotion words. In a survey of 2,386 people — 1,183 autistic and 1,203 non-autistic, aged 16 to 90 — autistic participants were around eight times more likely to describe themselves as asexual or as having an "other" sexuality, and reported lower rates of sexual activity, though the majority were sexually active and interested in relationships.¹⁰ Being asexual is an orientation. It is not a difficulty identifying your feelings. A questionnaire that cannot tell those apart will hand asexual autistic people extra points for existing.

Problematic Interpersonal Relationships. Here the association is real. In 200 healthy young adults, every alexithymia dimension was positively associated with interpersonal distress, and after adjusting for negative mood, difficulty identifying feelings remained the main predictor.¹¹ So the link exists. What that study cannot tell you is the direction, and what the OAQ-G2 cannot tell you is whether your relationship difficulties come from alexithymia or from being an autistic person in a world built for someone else. Fold the consequence into the measure and you get a number that goes up twice for the same life.

Vicarious Interpretation of Feelings. Three items about reading emotion in other people. This is the one addition where Bird and Cook's argument cuts in the questionnaire's favor, since difficulty reading others does track alexithymia rather than autism.² It is also three items, which is not enough to carry a subscale.

None of this is theoretical. An autistic writer took the OAQ-G2 in 2013 and published her working. She scored 141, well into the top band, then went through the domains one at a time. She found the externally-oriented thinking questions confusing — "I think in concrete terms, but I also spend a lot of time examining my thoughts and feelings" — flagged the sexual difficulties items as not part of the formal definition of alexithymia, and wondered whether being single would skew them. Her conclusion is where this module lands too: a reasonable amateur measure and a useful starting point, with no scientific validation behind it.¹² She worked that out by reading carefully. Nothing has changed since.

D. What the number is not

Here is the section that matters most, and it is short, because the honest version of it is short.

There is no published validation of the OAQ-G2. Not a weak one, not a small one, not one in an obscure journal. A title-and-abstract search of PubMed for the exact phrase "online alexithymia questionnaire" returns zero results. No peer-reviewed paper reports its internal consistency, its test-retest reliability, its factor structure, its correlation with the TAS-20, or its performance against any clinical criterion. There is no normative sample. There are no norms for autistic adults, or for anyone else.

The author's own page said, back in 2007, that "clinical application, and validation trials are presently underway, and the results of these studies will be collated and published in due course".⁴ That was nineteen years ago. Nothing appears to have been published. This is not an accusation of bad faith — validation studies are slow, expensive and frequently abandoned, and an instrument given away free has no budget behind it. It is simply the fact, and you are entitled to it before you read a number off a screen.

Diagram — D · The bands on the live screener. These three bands are what the screener will show you, reproduced exactly. What is missing is the study that put the lines at 94 and 113 rather than anywhere else. Cut-offs on validated instruments come from testing hundreds of people against an independent assessment. These have no such paper trail.

Two further things are worth knowing about those bands. The first is arithmetic: 37 items scored one to five give a minimum of 37, not 0, so the range of "0 to 185" on the live page does not match the published scoring. The band boundaries themselves are unaffected. The second is that the live page names its seven areas slightly differently from the original, and gives no score range or band for any of them, so there is no way to read your own profile across the seven. You get one number.

It is worth seeing what a real validation looks like, so you can measure the gap. A meta-analytic factor analysis of the TAS-20 pooled 88 samples from 62 studies: 69,722 people, 25 countries, 16 languages. It confirmed the three-factor structure, and found reliability of 0.84 for difficulty identifying feelings, 0.75 for difficulty describing feelings and only 0.62 for externally-oriented thinking.¹³ A separate study compared 743 autistic adults with 721 general-population adults and found that same subscale had a reliability of 0.451 in the autistic sample, with seven of its eight items loading poorly.¹⁴

So what is a score of 128 worth? Not nothing. It is a record of how you answered 37 mostly sensible questions about your emotional life on one particular afternoon. What it is not is a quantity. You cannot compare it with anyone else's, you cannot compare it with your own from two years ago in any calibrated way, and the difference between 112 and 113 means nothing at all. Read the items, not the number.

INSIDE THE INSTRUMENT

OAQ-G2 — the Online Alexithymia Questionnaire, second generation

Where it came from

Written by Jason Thompson in 2005 and revised in May 2007, with nine items reworded to reduce similarity to the TAS-20. The stated rationale was that the TAS-20 and the Bermond-Vorst Alexithymia Questionnaire targeted imagination poorly, conflated failure to recognize a feeling with absence of feeling, and were not publicly accessible.⁴ It is not a New Path instrument and it is not a licensed commercial one; it is a self-published questionnaire that has been circulating free on the web for two decades.

What it is made of

Thirty-seven items on a five-point agreement scale, published maximum 185, with items 3, 7, 8, 23 and 32 reverse-scored and unanswered items defaulting to the midpoint. Seven areas: Difficulty Identifying Feelings (6), Difficulty Describing Feelings (4), Vicarious Interpretation of Feelings (3), Externally-Oriented Thinking (7), Restricted Imaginative Processes (7), Problematic Interpersonal Relationships (6), Sexual Difficulties and Disinterest (4).⁴ Published cut-offs: 94 and below non-alexithymic, 95 to 112 possible, 113 and above alexithymic. The live screener reproduces these boundaries with different band labels and a stated floor of 0.

How well it performs

Unknown, in the strict sense. No alpha, no omega, no test-retest coefficient, no confirmatory factor analysis, no convergent validity against the TAS-20 or the PAQ, no sensitivity or specificity against any criterion. A PubMed title-and-abstract search for the exact phrase returns nothing. The developer stated in 2007 that validation trials were underway;⁴ no results appear to have been published in the nineteen years since.

Where it was validated

Nowhere. There is no normative sample, no clinical sample, no autistic sample and no cross-cultural work in the peer-reviewed record, despite translation into several languages. Any percentile, band or comparison a site attaches to an OAQ-G2 score is unsourced.

What it cannot do

It cannot support a quantitative claim of any kind. Without reliability data there is no standard error of measurement, so neither a change score nor a difference between two people is interpretable. Treat the total as an ordinal impression at best.

Four of its seven domains sit outside the current consensus construct. The field is actively divided on whether constricted imaginal processes belong in the definition of alexithymia at all: Taylor, Porcelli and Bagby defend the original four-component model,⁶ while Preece and colleagues exclude fantasising as a correlate rather than a component and built the PAQ accordingly.⁷ Interpersonal problems and sexual difficulty appear in no consensus definition on either side.

Several of its extra domains are confounded with autism itself. Autistic adults are roughly eight times more likely to report asexual or other sexual identity in a sample of 2,386,¹⁰ and interpersonal difficulty is a diagnostic feature of autism rather than a marker of alexithymia. An autistic respondent can accrue points across three of the seven domains without any difficulty naming emotions, which makes the total circular if it is being used to examine alexithymia in an autistic person.

The restricted imagination domain runs against the evidence in autistic samples. Higher autistic traits predicted lower divergent-thinking fluency but a greater number of atypical responses in 312 participants,⁸ and 42% of 223 autistic adults exceeded a maladaptive daydreaming cut-off.⁹ Scoring low imaginative output as pathology misreads a well-documented pattern.

It shares the structural weakness of every alexithymia self-report. The instrument asks people who have difficulty identifying and describing internal states to identify and describe internal states. Interoceptive differences are also less settled than the popular account suggests: a meta-analysis of 31 studies found no significant group difference on comparable adult cardiac interoceptive accuracy tasks (pooled estimate −0.21, p = 0.06).¹⁵

What a clinician does with it

Use it as a semi-structured interview schedule and discard the total. The 37 items are a reasonable prompt list covering ground the TAS-20 does not, and the four extra domains are clinically interesting precisely because they are contested. Where a client presents an OAQ-G2 result, follow it with the TAS-20 or the PAQ if a defensible figure is needed. Never enter the number in a formulation, a report or a referral letter.

Validity tier: 4 — honest and useful, not normed, no published validation. Tier 4 usually marks an instrument New Path built in-house; this one is not ours, but it earns the same tier for the same reason, and the underlying construct is well evidenced even though this particular questionnaire is not.

Diagram — E · One high total, many roads. This is the caveat that does the most work. Depression, exhaustion, trauma, being autistic, being asexual and being single will each push an OAQ-G2 total upward through different doors. The single number that comes out the other end looks identical whichever door you came through.

STRENGTHS LENS

You went looking for a longer version of the question. That instinct is the right one.

Most people take one questionnaire, get a band, and stop. You read three subscales and thought: that is not the whole shape of it. It is not the whole shape of it. The original clinical description of alexithymia had four components and the standard instrument measures three, and specialists are still publishing papers at each other about the missing one. Your sense that something had been left out was accurate before anybody told you.

Keep the same instinct when you read your result. A longer questionnaire feels more thorough, and the extra length is exactly why it deserves more scepticism rather than less — more items means more places for an untested question to add a point it should not have. Holding "this is useful to think about" and "this is not a measurement" in the same hand is a genuinely difficult skill. You are practising it right now.

F. What helps

The practical work for alexithymia itself is in Module 9, and this module does not repeat it. What follows is what to do with this particular questionnaire, which is a different job.

1. Read the items, not the total.

Go back through the 37 statements and mark the six or seven that made you stop and think. Those are the output. The number at the end is an unweighted sum of answers to questions of very different quality, and nothing in the literature tells you how to interpret it, so it is the least informative thing the questionnaire produced.

2. For every high answer, write the second explanation.

If you agreed that you have few close friendships, write down the alexithymia reading and then write the other one: fewer friends by choice, a smaller social appetite, an autistic social style, three house moves in five years. Do the same for the imagination and the sexual items. You are doing by hand the discrimination the questionnaire cannot do.

3. Take the TAS-20 as well, and treat that one as the number.

The TAS-20 in Module 9 is not a better set of questions than these, but it is a vastly better characterised one, with its weak points mapped in detail across 25 countries and tens of thousands of people. If you want a figure you can hand to a clinician, use that one, and use the OAQ-G2 for the content it raised.

4. Watch what happens to the four extra domains under stress.

Imagination, interest in sex and interest in other people all collapse under exhaustion, depression and autistic burnout, and all of them recover. If your high answers cluster there rather than in identifying and describing feelings, the question worth asking is not whether you are alexithymic but what the last two years have cost you.

5. Bring the questions to the conversation, not the score.

"I scored 128 on an alexithymia test" is a sentence a clinician can do nothing with, and a careful one will discount it. "Three of these questions describe me exactly and I have never had language for it before" is a sentence that starts an assessment. The workbook below is built to help you produce the second sentence.

Step 2 — Take the screener

Thirty-seven questions, seven to ten minutes, free and confidential. You get a total score with a band, drawn from seven areas of emotional life rather than the three the TAS-20 covers. One honest steer before you start: this questionnaire has never been validated in any published study, so treat the items as good questions to think about and treat the total as a rough marker rather than a measurement.

Take the OAQ-G2

Before you start

The OAQ-G2 is a free self-report questionnaire about alexithymia — difficulty identifying and describing your own feelings — that was written in 2005 and revised in 2007, and it has no published validation of any kind: no reliability data, no factor analysis, no normative sample and no testing against clinical assessment. Four of its seven areas sit outside the consensus definition of alexithymia, and three of them can be raised by things that have nothing to do with naming emotions, including being autistic, being asexual and being single. It is a screen and not a diagnosis, and alexithymia is not a diagnosable condition in any case: it is a trait that runs on a dimension. If you want a figure with evidence behind it, use the TAS-20 in Module 9 instead. Only a clinician can diagnose.

Step 3 — Your workbook

Your answers save to this device only — we cannot see a word of what you write. This module's workbook is deliberately built around the items rather than the score: it helps you pull out the statements that landed, write the second explanation for each of the four untested areas, and turn all of it into one sentence you could actually say out loud.

Your OAQ-G2 result

Took the screener? Record it here if you want to. One steer before you do: this questionnaire has no published validation, so the number is a marker in your own notes rather than a measurement. Entirely optional — skip it if you would rather just read.

Score bands: 0–94 = No to Low Prevalence of Alexithymia Traits; 95–112 = Moderate Presence of Traits; 113–185 = High Presence of Traits

Fields: OAQ-G2 · Online Alexithymia Questionnaire, 37 items; Total score (0–185); My total (enter 0–185); The seven areas; The live screener names seven areas but publishes no score range or band for any of them, so there is nothing to enter. Note instead which areas the questions seemed to land in for you.

The items that stopped you

Section F, item 1. The number is the least informative thing this questionnaire produced. The statements that made you pause are the useful output.

Fields: The statements I read twice; The one that was most accurate; Had I ever had words for that before?: No, not once / Sort of, never clearly / Yes, but not in those words / Yes

The second explanation

Section F, item 2. For each high answer in the four extra areas, write the alexithymia reading and then write the other one. This is the discrimination the questionnaire cannot make for you.

Fields: Tick: Imagination items: is there another explanation?; Tick: Relationship items: is there another explanation?; Tick: Sexual items: is there another explanation?; The other explanations, written out

Naming vs saying vs everything else

Three of the seven areas are the alexithymia the research is about. Four are additions nobody has tested. Sorting your own answers into those two piles is worth more than the total.

Fields: Where did most of my high answers sit?: Identifying and describing feelings / The four extra areas / Spread evenly across both / I could not tell; What that split tells me, in my own words

What the last two years cost

Section F, item 4. Imagination, interest in other people and interest in sex all fall under exhaustion, depression and burnout, and all of them come back. Worth separating a trait from a state.

Fields: Has this been true for as long as I can remember?: Always, as far back as I can recall / Most of my adult life / Only the last few years / Only since something specific happened; What was going on when it changed

The sentence you would actually use

Section F, item 5. Not the score. One sentence built from the items, that a clinician or a partner could do something with.

Fields: Who I would say it to; The sentence

Appendix — Research companion

Peer-reviewed research

5. Bagby RM, Parker JD, Taylor GJ (1994). The twenty-item Toronto Alexithymia Scale--I. Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38(1), 23-32. DOI 10.1016/0022-3999(94)90005-1. View the paper The originating paper for the TAS-20, included here as the benchmark against which the OAQ-G2 is compared. Two studies produced a twenty-item scale with good internal consistency and test-retest reliability and a three-factor structure - difficulty identifying feelings, difficulty describing feelings and externally-oriented thinking - which was then replicated by confirmatory factor analysis in separate clinical and non-clinical samples. It shows what item selection and cross-validation look like when they are actually carried out and reported. Limitation: developed in Canadian student and clinical samples in the early 1990s, with no autistic participants and no analysis establishing the sensitivity or specificity of any cut-off score.

8. Best C, Arora S, Porter F, Doherty M (2015). The relationship between subthreshold autistic traits, ambiguous figure perception and divergent thinking. Journal of Autism and Developmental Disorders, 45(12), 4064-4073. DOI 10.1007/s10803-015-2518-2. View the paper Study of 312 people from a non-clinical population, relating self-reported autistic traits to ambiguous figure perception and to divergent thinking tasks. Higher autistic traits were associated with reduced fluency - fewer ideas generated - but with an unusually high number of atypical responses, leading the authors to conclude that generation of novel ideas may be an adaptive advantage associated with autistic traits. It undercuts the assumption that low imaginative output in autistic people is a deficit to be scored. Limitation: a non-clinical sample using self-reported subthreshold autistic traits rather than diagnosed autistic adults, so the findings describe a trait gradient in the general population and cannot be read directly onto a clinically diagnosed group.

2. Bird G, Cook R (2013). Mixed emotions: the contribution of alexithymia to the emotional symptoms of autism. Translational Psychiatry, 3(7), e285. DOI 10.1038/tp.2013.61. View the paper The paper that set out the alexithymia hypothesis: that impaired emotion recognition and reduced empathy in autistic people, where found, are attributable to co-occurring alexithymia rather than to autism itself. It reports alexithymia in roughly 10% of the general population against 40 to 65% of autistic adults, correlations above 0.60 between alexithymia and accuracy at recognising vocal affect with no independent contribution of autism severity, and anterior insula empathic responses that tracked alexithymia in autistic and non-autistic participants alike. It also documents elevated alexithymia across eating disorders, schizophrenia, substance use and several neurological conditions. Limitation: a theoretical review synthesising the authors' own earlier work rather than new data, with small experimental samples behind several of the effects it summarises.

1. Kinnaird E, Stewart C, Tchanturia K (2019). Investigating alexithymia in autism: a systematic review and meta-analysis. European Psychiatry, 55, 80-89. DOI 10.1016/j.eurpsy.2018.09.004. View the paper Systematic review and meta-analysis of 15 studies using the Toronto Alexithymia Scale to compare autistic and neurotypical groups, covering 366 autistic and 348 comparison participants. 49.93% of autistic participants met the alexithymia threshold against 4.89% of comparison participants, and autistic groups scored significantly higher on the total and on every subscale. The authors conclude that alexithymia is common rather than universal in autism and that autistic people with alexithymia may represent a distinct subgroup with distinct clinical needs. Limitation: the included studies recruited mainly cognitively able adults with normal to high IQ, so the pooled figure does not generalise across the whole autistic population, and every study relied on self-report of a difficulty that is itself about self-report.

15. Klein M, Witthoft M, Jungmann SM (2025). Interoception in individuals with autism spectrum disorder: a systematic literature review and meta-analysis. Frontiers in Psychiatry, 16, 1573263. DOI 10.3389/fpsyt.2025.1573263. View the paper Systematic review and meta-analysis of 31 interoception studies across the lifespan in autism, comprising 9 child or adolescent studies, 15 adult studies and 7 lifespan studies. Across the five adult studies using comparable cardiac interoceptive accuracy tasks there was no significant difference between autistic and non-autistic adults (pooled estimate -0.21, SE 0.11, p = 0.06), and findings for interoceptive sensibility and awareness were inconsistent in both directions. The review notes that roughly half of autistic people have co-occurring alexithymia and that interoceptive confusion is the mechanism usually proposed to link the two. Limitation: only five studies entered the adult accuracy meta-analysis, the tasks differed widely between studies, and alexithymia was discussed but never tested as a moderator.

11. Koppelberg P, Kersting A, Suslow T (2023). Alexithymia and interpersonal problems in healthy young individuals. BMC Psychiatry, 23, 688. DOI 10.1186/s12888-023-05191-z. View the paper Cross-sectional study of 200 healthy young adults, 100 women and 100 men, mean age 23.97 years, relating alexithymia dimensions to the circumplex model of interpersonal problems. All alexithymia dimensions were positively associated with general interpersonal distress, and after adjusting for negative affect, difficulty identifying feelings remained the principal predictor, while difficulty describing feelings specifically predicted lower agency and lower communion. Women scored higher on difficulty identifying feelings and men on externally oriented thinking. Limitation: a small non-clinical sample of young university-age adults, cross-sectional and self-reported throughout, so it establishes association rather than direction and does not generalise to older or clinical populations.

7. Preece D, Becerra R, Robinson K, Dandy J, Allan A (2018). The psychometric assessment of alexithymia: development and validation of the Perth Alexithymia Questionnaire. Personality and Individual Differences, 132, 32-44. DOI 10.1016/j.paid.2018.05.011. View the paper Development and validation of the 24-item Perth Alexithymia Questionnaire, built on an attention-appraisal model that separates difficulty attending to and difficulty appraising emotion across negative and positive valence, and that deliberately excludes fantasising on the grounds that it is a correlate of alexithymia rather than a component. Study 1 (N = 231) supported the theorised factor structure with high internal consistency, and study 2 (N = 748) replicated it and demonstrated concurrent and discriminant validity against psychopathology and emotion regulation measures. Limitation: both validation samples were general-community and student adults with no autistic participants and no clinical criterion measure, so its performance in autistic populations is not established by this paper.

13. Schroeders U, Kubera F, Gnambs T (2022). The structure of the Toronto Alexithymia Scale (TAS-20): a meta-analytic confirmatory factor analysis. Assessment, 29(8), 1806-1823. DOI 10.1177/10731911211033894. View the paper Meta-analytic structural equation modelling of 88 samples nested in 62 studies, totalling 69,722 participants from 25 countries in 16 languages, comparing nine competing measurement models. The original three-factor solution fitted best, but reliability differed sharply by subscale: omega was .84 for difficulty identifying feelings, .75 for difficulty describing feelings and .62 for externally-oriented thinking, and every negatively keyed externally-oriented thinking item loaded more strongly on a method factor than on its content factor. The authors recommend reporting subscale scores rather than a single total. It is included here as the benchmark for what a validated instrument's evidence base looks like. Limitation: about three quarters of the pooled samples were non-clinical and largely university students, no autistic samples were analysed separately, and a pooled correlation matrix cannot detect problems specific to any one population.

3. Sifneos PE (1973). The prevalence of 'alexithymic' characteristics in psychosomatic patients. Psychotherapy and Psychosomatics, 22(2-6), 255-262. DOI 10.1159/000286529. View the paper The paper that introduced the term alexithymia, based on a comparison of psychosomatic patients with control patients. Sifneos described a relative constriction in emotional functioning, poverty of fantasy life, and inability to find appropriate words to describe emotions, and reported these characteristics at roughly twice the rate in the psychosomatic group. It is the reference for the claim that impoverished fantasy was part of the construct from the beginning, which is the theoretical warrant for any questionnaire that includes an imagination domain. Limitation: a small clinical study from 1973 using unstandardised interview-based ratings by the author, with no formal instrument, no blinding and no replication design, so it establishes the origin of the concept rather than any measurable property of it.

10. Weir E, Allison C, Baron-Cohen S (2021). The sexual health, orientation, and activity of autistic adolescents and adults. Autism Research, 14(11). DOI 10.1002/aur.2604. View the paper Survey of 2,386 people aged 16 to 90, comprising 1,183 autistic and 1,203 non-autistic participants, covering sexual orientation, activity and health. Autistic participants were around eight times more likely to report asexuality or an 'other' sexuality and were less likely to report being sexually active - 70% of autistic males and 76% of autistic females against 89% of non-autistic participants - with no group difference in age at first sexual activity or in rates of sexually transmitted infection. The authors emphasise that the majority of autistic adults are interested in and engaged in sexual relationships. Limitation: a large but self-selected online convenience sample of predominantly cognitively able adults, cross-sectional and entirely self-reported, so the proportions describe the respondents rather than the autistic population.

9. West MJ, Somer E, Eigsti IM (2023). Immersive and maladaptive daydreaming and divergent thinking in autism spectrum disorders. Imagination, Cognition and Personality, 42(4), 372-398. DOI 10.1177/02762366221129819. View the paper Two online samples were compared: 544 adults recruited through maladaptive daydreaming communities and 223 adults with confirmed autism diagnoses recruited through the SPARK research cohort. 42% of the autistic sample scored above the maladaptive daydreaming cut-off, against 88% of the daydreaming community sample, and divergent thinking originality in the autistic sample was comparable to the community sample. Broader autism traits predicted higher maladaptive daydreaming in the community sample. It is direct evidence against the assumption that autistic imaginative life is impoverished. Limitation: online self-report with self-selected samples, no direct measure of non-pathological immersive daydreaming, and a single divergent thinking task type, so the 42% figure is not a population prevalence.

14. Williams ZJ, Gotham KO (2021). Improving the measurement of alexithymia in autistic adults: a psychometric investigation of the 20-item Toronto Alexithymia Scale and generation of a general alexithymia factor score using item response theory. Molecular Autism, 12, 56. DOI 10.1186/s13229-021-00463-5. View the paper Psychometric study comparing 743 autistic adults from the SPARK cohort with 721 general-population adults matched on mean age. The full TAS-20 fitted poorly in both groups; the externally-oriented thinking subscale had omega of 0.451 in the autistic sample with seven of its eight items loading poorly on the general alexithymia factor, and one item showed practically significant differential item functioning between groups. The authors derived an eight-item general alexithymia factor score correlating 0.910 with the full total and with negligible between-group differential functioning. It demonstrates the level of scrutiny that has never been applied to the OAQ-G2. Limitation: an online, largely white and highly educated self-report sample from a research register, and this article is a revised replacement for an earlier version retracted over copyright concerns.

Lived experience

12. Musings of an Aspie (2013). Taking the Alexithymia Questionnaire. Musings of an Aspie, an autistic-authored weblog, 5 February 2013. View the source A first-person account by an autistic writer of taking the Online Alexithymia Questionnaire and working through her results domain by domain. She scored 141, in the high band, and reported her domain scores openly, but identified the externally-oriented thinking items as conflating concrete thinking with lack of introspection, noting that she thinks in concrete terms and also spends a great deal of time examining her thoughts and feelings. She flagged the sexual difficulties items as not part of the formal definition of alexithymia and questioned whether relationship status skews them, concluding that the questionnaire is a reasonable amateur measure and a useful starting point but lacks scientific validation. Limitation: a single individual's account with no sample, no data and no peer review, illustrating how the instrument reads from the inside rather than evidencing any psychometric claim.

Emerging or contested

6. Taylor GJ, Porcelli P, Bagby RM (2024). Alexithymia: a defense of the original conceptualization of the construct and a critique of the attention-appraisal model. Clinical Neuropsychiatry, 21(5), 329-357. DOI 10.36131/cnfioritieditore20240501. View the paper A theoretical paper defending the original four-component definition of alexithymia - difficulty identifying feelings, difficulty describing feelings, constricted imaginal processes and externally-oriented thinking - against the three-component attention-appraisal model proposed by Preece and colleagues. The authors argue there is insufficient evidence to warrant removing the imaginal activity component, noting that the alternative model has had limited empirical testing. It is the clearest statement that the field has not settled what alexithymia includes, which is directly relevant to any instrument that adds domains. Limitation: a conceptual and polemical review by the developers of the TAS-20 defending their own construct, with no new data, and it is one side of a live disagreement rather than a resolution of it.

4. Thompson J (2007). Online Alexithymia Questionnaire - G2: scoring, factors and background. Self-published developer weblog for the OAQ-G2 (oaq.blogspot.com), February 2007. View the paper The developer's own account of the instrument and the only primary source for how it is built. It states that the OAQ was created in 2005 in response to perceived shortcomings of the TAS-20 and the Bermond-Vorst Alexithymia Questionnaire, revised in May 2007 with nine items reworded, and consists of 37 items on a five-point agreement scale with a maximum of 185, items 3, 7, 8, 23 and 32 reverse-scored, and cut-offs of 94 and below, 95 to 112, and 113 and above. It lists seven factors with item counts summing to 37, and states that clinical application and validation trials were then underway with results to be published in due course. Limitation: a self-published weblog by the instrument's author with no data, no sample, no psychometric coefficients and no peer review, and the validation results it promised in 2007 do not appear in the published literature.

Up next

Module 11 - ASIM-24 — Sex, Intimacy and Sensory Reality

All modules in Autistic Self-Discovery

The questions were good. The number was not. This course was built by clinicians who are part of the New Path Family. If a set of unvalidated questions has just put language to something you have carried for twenty years, that is still worth taking somewhere — the items are the useful part, and we are happy to start from those rather than from a score. Therapy for clients in California and coaching worldwide, all by telehealth, are offered by our sister company New Path Family of Therapy Centers, Inc. A conversation costs nothing and there is no pressure. Saving this for later counts too. Talk with the New Path team

NEW PATH FAMILY