Most tests ask how empathetic you think you are. This one gives you situations where being empathetic costs something, and looks at what you actually choose.
You get three scores — whether you registered the people involved, whether you explained your reasoning, and whether you reached the right decision. Then you can see how the same scenarios were answered by the AI models everyone is now asking for advice.
Start the assessment Free. No account, no sign-up.
Whether the person in the situation is someone you answer, or a case you process. The scenarios are built so that the considerate response is the inconvenient one.
Whether you say why. This is the gap the assessment finds most often — the decision is sound, but nobody is told the reason for it, which is what makes a good call unpersuasive.
Whether you act on the thing the situation actually turns on. In several scenarios the urgent detail is deliberately buried in procedural noise.
The questions change depending on what you do, because a clinician and a student are not making the same decisions. You pick your track at the start.
Everyday situations — grief, conflict, family and community. No domain knowledge needed.
Workplace decisions — grievances, redundancy, pay, promotion and the people policies land on. No HR qualification needed.
Triage, consent, end-of-life and AI-assisted clinical decisions.
This is not a validated psychometric test and does not pretend to be one. There is no clinical instrument behind it. It is a scenario-judgement exercise scored against a written reference for each situation — useful for thinking about your own decisions, and unsuitable for hiring, diagnosis, or assessing anybody else.
The AI side of your report comes from ai_empathy, an open-source Drupal module. It puts the same scenarios to several AI models and grades their answers on the same rubric yours are graded on, so the comparison is like for like rather than a score against a number somebody picked.
You read 7 short scenarios and say what you would do. Most give you four responses to choose between; two ask you to write a few sentences. It takes about fifteen minutes. At the end you get your three scores and an interpretation of your strongest dimension, free and immediately.
No, and it is worth being clear about that. There is no validated instrument behind it — this is not the Jefferson Scale, the IRI or the TEQ. It is a scenario-judgement exercise: your answers are compared against a written reference for each scenario. That makes it useful for reflection and unsuitable for selection, diagnosis, or assessing anyone other than yourself.
Your scenario choices carry scores written into each option in advance. Your written answers are scored by a language model against the same reference used for everything else, on three dimensions: whether you registered the people involved, whether you explained your reasoning, and whether you reached the right decision. The report names the model that graded each written answer, so you can see what produced your score.
Every scenario you answered, with why your answer scores where it does and what the fullest answer to it adds. Interpretation of all three dimensions rather than just your strongest. The specific patterns in how you answered — including whether you tend to reach the right decision without saying why, which is the single most common gap this finds. And your written answers set beside the AI models answering the same scenarios on the same rubric.
The assessment and your scores are free. The full report is ₹199 or ₹299, paid once, with no subscription and no account.
No. There is nothing to sign up for and no password. You give an email address at the end so the report can be sent to you, and that is the only thing asked of you.
The link in your email works for 14 days and can be opened on any device, so keep that email. The link shown on screen straight after payment is short-lived and tied to that browser — it is there so you can read the report immediately, not as your permanent copy. A PDF is attached to the email as well, which never expires.
Your answers are used to produce your report and nothing else. They are not sold, not shared, and not published. Your written answers are sent to the model that scores them, and nothing is stored against your name — there is no name to store. Ask and they will be deleted.
Not unless you tick the box, and not even then yet. There is an optional consent at the end, off by default. Research use requires ethics committee approval, which is not in place, and until it is nothing that has been consented to will be analysed or published. Consenting today gives permission for later; it does not put you in a study now.
Yes, and the scenarios stay the same, so a second attempt tells you more about what you remember than about how you have changed. If you want to see movement, leave real time between attempts and answer from what you would do rather than from what you scored well on last time.
Payment is handled by Razorpay and card details never reach this site. If you are charged and the report does not appear, it is still on its way to your email — wait a few minutes and check there before paying again. If it never arrives, get in touch with the email address you used and it will be sent or refunded.
Start the assessment About fifteen minutes.