Emotional vocabulary
Alexithymia test reported as three sub-scales, not one number
Answer 21 original statements on a five-point scale and 8 scenarios on two sliders each, and get back the three factors separately — trouble knowing what you feel, trouble putting it into words, and attention that points outward by default. Free, about eight minutes, no account, nothing held back. There is no total and no threshold anywhere on the page, because the published twenty-item scale owns its cut scores and these are not its items.
- 100% free
- No signup
- 21 items + 8 scenarios
- Three sub-scales
- Not a TAS-20 score
Two halves, about eight minutes
21 statements first, one at a time, answered on a five-point agreement scale with the number keys or the buttons. Then 8 short scenarios with two sliders each — how much there was to feel, and how readily a word for it arrived. The statements are grouped by the three factors the research describes and reported as three separate numbers; the scenarios are reported as the distance between two of your own ratings.
Difficulty identifying feelings
Knowing that something is happening in you, and knowing which thing it is — including telling a feeling apart from the bodily state it arrives with.
Difficulty describing feelings
Getting a feeling you have already identified into words that another person can use. A separate skill, and the two come apart in both directions.
Externally oriented thinking
Where attention goes by default. A style that points at facts, events and mechanics rather than at inner states — which is a preference, not a deficit, and the most argued-about of the three.
Every statement here was written for this page. The twenty-item scale this query usually means is copyrighted, its cut scores belong to it, and a threshold from an instrument that was never administered would be a number with nothing behind it. Nothing below produces a total, a band or a label. The construct itself is described in Bagby, Parker & Taylor (1994), cited in full beside the sub-scales at the end.
How to work through both halves
Statements first, then the two ratings that usually get collapsed into one.
Answer 21 statements, one at a time
Each one is a situation rather than a verdict about yourself — what happens when somebody asks what is wrong, what you reach for when the accurate word will not come — because a statement about an event is easier to answer honestly than a statement about a trait. Press 1 to 5 or tap, and Backspace steps back if you want to change one. Six of the twenty-one run the other way, so agreeing with everything does not push any sub-scale up.
Rate 8 scenarios twice each
The first slider asks how much there is to feel, whatever the feeling is. The second asks how readily the right word would arrive — not whether you would say it out loud, which is a question about privacy rather than vocabulary. Both sliders have to be moved on all eight: an untouched slider is left as unset rather than counted as the midpoint, and the page will not continue until every one of the sixteen has been placed.
Read the three bars and the one distance
The three sub-scales come back as raw scores out of 35 with the proportion of their range beside them, and they are never added together. The scenarios come back as your average size-of-reaction, your average ease-of-naming, and the distance between the two — a comparison of your own answers with each other, which is the only comparison this page can make honestly. The widest single gap is named so you can see which scenario produced it.
Technical specifications
| Items | 21 statements written for this page, 7 on each factor, answered 1 to 5 from strongly disagree to strongly agree. 6 are reverse-keyed — two per factor — and scored as 6 minus the answer |
|---|---|
| Sub-scale range | 7 to 35 each, since 7 items each carry 1 to 5. Reported as the raw score and as the percentage of that 28-point span, never as a percentile and never added into a fourth number |
| Second half | 8 scenarios, 2 sliders each on 0 to 10: size of the reaction and readiness of the word. Reported as two means and their difference, with the largest single-scenario gap named |
| Unanswered statements | Counted as 3, the middle point, which pulls a sub-scale toward its center. The results panel says how many of the 21 were answered so the effect is visible rather than silent |
| Cut score | None, on purpose. This page writes its own statements because the twenty published ones are copyrighted, and a threshold derived on those items does not carry across to these. The construct is described in Bagby, Parker & Taylor (1994) |
| Prevalence figure | Not printed. Every published estimate is a count of people above a threshold on a continuous scale, so the figure moves with the threshold rather than with the population, and quoting one would mean adopting one study's cutoff on items that study never saw |
| Input | Number keys 1 to 5 on the row or the keypad, on-screen buttons, and Backspace to step back one statement. Nothing on the page is timed, so the number of statements answered by tapping is recorded and reported but changes no figure |
| What leaves the page | Nothing. The 21 answers and the 16 slider positions are summed in this tab and never sent, which is also why a reload loses them and why the copy button exists |
Frequently asked questions
Is this the TAS-20?
No, and it is not a version of it either. The twenty-item Toronto Alexithymia Scale is a copyrighted instrument, its items are not reproduced anywhere on this page, and none of the statements here is a rewording of one. What this page borrows is the part that is publishable knowledge — the three-factor structure that scale's authors identified and validated, which is a finding about how difficulty with feelings is organized rather than a piece of the questionnaire. That is why the sub-scale names will look familiar and the statements will not.
Why is there no total score?
Because a total here would be a number shaped like the published one, sitting next to sub-scale names taken from the same research, and somebody would compare it with a real one they got from a clinician. That comparison would be meaningless in both directions: different items, a different number of them, a different scale length and no standardization sample. There is a second reason that has nothing to do with copyright. The three factors do not have to move together — plenty of people know precisely what they feel and cannot describe it, which is a high score on one and a low score on another — and a total averages that structure away just as it becomes interesting.
The third factor does not feel like a problem to me. Is it supposed to be?
It is not, and it is the most argued-about part of the construct for exactly that reason. Externally oriented thinking describes where attention goes by default — toward mechanics, events and what happened rather than toward inner states — and that is a cognitive style with a long list of situations where it is the right one. It earns its place in the construct because someone who never turns attention inward gets little practice at identifying or naming what is there, so it is better read as a maintaining condition than as a symptom. This page keeps it separate rather than folded into a total precisely so a high score there does not contaminate the other two.
Is alexithymia the same as being autistic?
No, and the relationship between them is one of the more interesting findings in this area. Difficulty recognizing and describing emotion is often listed as an autistic trait, but a line of research argues that where the two occur together it is the alexithymia doing the work: autistic people who do not also have high alexithymia perform much like anyone else on emotion-recognition tasks, which suggests the two are separable and frequently co-occurring rather than one being part of the other. Alexithymia also appears in people with no connection to autism at all, and nothing on this page can distinguish any of these, which is a large part of why it produces no verdict.
I am not sure I can judge my own feelings. Does that break the test?
It is the built-in weakness of every self-report scale in this area, and being aware of it is the reason for the second half. Each statement asks you to introspect on the very faculty in question, so someone who genuinely cannot tell what they feel is not well placed to report that they cannot — the scale needs the ability it is trying to measure. The scenarios are an attempt to get at the same thing sideways: instead of asking how good you are at naming feelings, they ask for two ratings and take the distance between them, which is a comparison you can make even when you do not trust either number on its own.
Why two sliders instead of one question about how I feel?
Because the ordinary question hides the distinction the whole word is about. Alexithymia is not an absence of feeling — the literal Greek is roughly no words for feelings, not no feelings — and people described by the term routinely report the physical side of an emotion at full strength while the name for it stays out of reach. Ask a single question and someone with a large unnamed reaction and someone with a small clearly-labeled one can give the same answer. Two sliders keep them apart, and the difference between the sliders is the only quantity on this half of the page that is worth reading.
Both of my numbers came out low. What does that mean?
It means two things at once and the page cannot separate them. A low rating on size-of-reaction is what you get from someone whose reaction really was small, and it is also what you get from someone who could not gauge how large their reaction was — the second is arguably the core of the construct, and it arrives as the same slider position as the first. No arrangement of these questions distinguishes the two, so the results panel states the ambiguity instead of resolving it. If that pattern is familiar and it matters to you, it is a conversation to have with someone who can watch you have a feeling rather than a number to collect from a page.
Where the word came from, and what the three factors separate
The word was assembled in 1973 out of three pieces of Greek — a, lexis, thymos, roughly no words for feelings — by a psychiatrist describing patients whose physical complaints arrived with no accompanying account of an inner life. It has traveled a long way from that clinic since, and the important step for anything you might do on a web page came two decades later, when Bagby, Parker & Taylor (1994) tested the structure and found that what had been treated as one trait behaves as three: trouble working out what is going on inside you, trouble getting it into words once you have worked it out, and a habit of attention that points at the world rather than at yourself. Those three are what this page reports, separately, and the separation is not a formality — the second is not the first with extra steps. Knowing precisely what you feel and being unable to say it is a common profile, and so is talking fluently about emotions in general while being unable to tell which one is currently happening. Full citation, since these things travel: Bagby, Parker & Taylor (1994), The twenty-item Toronto Alexithymia Scale — I. Item selection and cross-validation of the factor structure, Journal of Psychosomatic Research.
What no page can give you is a line to be on one side of. The twenty items are copyrighted and not ours to reproduce, so this page cannot administer the TAS-20 — and cut scores from an instrument the page did not administer are meaningless applied to items it wrote itself. The registry blurb refers to the scale's three-factor structure, which is publishable knowledge; the items and the thresholds are not. Beneath the copyright question sits a measurement one that would remain even if every item were public: alexithymia is a continuous trait spread through the general population rather than a category people fall into, so a threshold is a decision somebody made for the purposes of a study. That is also why no prevalence figure appears here. Every published estimate is a count of respondents above a cutoff, so the number moves when the cutoff moves, and the estimates disagree with each other for that reason rather than because anybody counted badly. A page that quoted one would be adopting a particular study’s threshold and applying it to statements that study has never seen.
The half of this page that is not a questionnaire exists because of a specific problem with questionnaires here. Every item asks the respondent to introspect on the faculty being measured, which is a strange thing to ask of someone whose difficulty is precisely with introspection, and no amount of careful wording fixes it. The scenarios sidestep it by asking for two ratings that everyday language collapses into one, and reading the distance between them — a within-person comparison that needs no norm and no cutoff, and that a respondent can make even while distrusting both numbers separately. The same reasoning shows up next door: the sensory processing test asks about signals from inside the body as one of its channels, and people who report difficulty naming emotions often report difficulty reading hunger, thirst and fatigue too — the bodily and the emotional halves of the same question. If you would rather have a task than a set of statements, the aphantasia test pairs its questionnaire with something behavioral, and the face blindness test is a measurement with no self-report in it at all.
One deliberate absence: nothing on this page is timed. How long you take over a statement about your own feelings is not a measurement of anything except how carefully you are reading, so no clock runs, no reaction time is collected, and none of the millisecond-level caveats that govern the speeded tasks elsewhere on this site apply to any figure here.
This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish alexithymia. Only a qualified professional, working with more than a browser, can make that judgment.
Where your 21 answers are added up
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Your answers, your reaction times and your score are never uploaded, logged or kept — which is also why the test carries on working after you disconnect from the network, and why nothing here can be held back behind an email address.
These are among the more personal answers anyone gives a web page, so it is worth being exact: the 21 statement answers and the 16 slider positions exist only as JavaScript objects in this tab, are read once to produce three sums and two averages, and are not written to storage or to a URL. Nothing on this page performs a network request after it has loaded.