How it works
Each quiz shows your answers back to you as they are. This page explains how that works for the four friendship situations, then sets out how the other quizzes differ.
What the situations ask
Each situation shows an ordinary moment between friends and asks what you would do first. The question is not what you are like in general. It is what you would reach for in this particular moment.
Every situation spells out conditions such as timing, the relationship and how urgent things are. Left to the imagination, those details vary so much from reader to reader that they would swamp the differences between the options.
Four is not a validated number. It was an editorial choice: short enough for a first-timer to finish without effort, and varied enough to cover different kinds of moments.
How a result is put together
Every option has its result written in advance: a short name, a sentence describing the choice, and one thing that is still open afterward. Each situation also has one question you could ask a friend. The result screen shows the sentences for the option you chose, exactly as written.
That is the whole calculation. Nothing is added up, nothing is compared with other people’s answers, and no combination of answers is turned into a type. Choose the same option and you get the same sentences, whoever you are and whenever you take it. No AI runs when your result is shown.
Situations you skip are left out. An unanswered situation is never filled in with a guess.
What a result does not say
Your choice is a single decision in an imagined situation. It does not mean you would act that way in real life, or with every person you know.
It makes no judgment about your personality, your abilities or how good a friendship is. It is not a medical or psychological assessment.
None of the four options is more mature or better than the others. The result sentences describe the choice and what comes next, and are written to avoid sounding like praise or criticism.
Why each result ends with a question
Guessing what someone wants is different from asking them. In research published in 2018, Eyal, Steffel and Epley found that telling people to imagine another person’s perspective did not reliably make them better at knowing what that person thought, while hearing it from the person directly did.
There is no evidence yet that this exercise improves friendships, and it does not claim to. What it tries to do is hand you one question you could ask instead of guessing.
How the other quizzes differ
Everything above describes the friendship quiz. The groups-and-work quiz and the everyday-life-in-Korea quiz also show your choices as they are, without rating them or comparing them with anyone else’s. The Korea quiz does not judge how Korean anyone is, or anyone’s nationality.
Only the reading quiz has right answers. Each can be worked out from the text alone, and your total says something about those six questions and nothing more. The festival story picks an ending from your two choices; it is for fun and says nothing about your personality.
You can also take the friendship quiz with a friend on one phone. You answer in turn, and if you both agree, your answers are shown side by side. The page counts how often they match, but that is not a measure of how good the friendship is, and neither person’s answers are saved or sent anywhere.
If you also answer what the others expect in the groups-and-work quiz, your first move and what you think they expect are shown side by side. Nobody can know what other people actually expect without asking, so a difference is not a mistake. The same goes for the Korea quiz questions that ask you to guess what most people would pick. There is no survey data yet, so those guesses are not marked right or wrong.
The communication survey asks how often, over the past three months, you actually did each thing, on a scale from 1 to 5. For each area it shows the average of your own answers. That is not a ranking against other people and not a personality type. The statements are still being refined and have not been validated with real participants, and each area asks about different things, so averages cannot be compared across areas. Statements you marked as hard to answer, such as “Hasn’t happened,” are left out of the average.
In the full 120-statement version, the last two statements in each area are worded the other way around and are still under review, so you can answer them but they are left out of the average. An area gets no average if fewer than 5 of its statements were answered on the five-point scale, or fewer than 15 in the full version. Each average also comes with the answer it is closest to, such as “Sometimes.” The combined 58-question version is the 36 standard statements plus 16 questions from the group situations and the 6 reading-quiz questions.
How the questions were written
The situations, options and result sentences were written for this service. None of them is taken or adapted from an existing psychological test.
AI tools helped with the first drafts, and nothing is published until a person has read and reviewed it. The photographs on the site are AI-generated images and show no real person.
Where your answers go
Your answers and your result are put together in your browser and are never sent to a server. Nothing is stored, so they are gone when you close the page, and nobody running this site can see what anyone chose. The privacy policy has the details.