Write three associations a day for a week and you end up with about twenty-one notes of two or three words each. On their own they don’t say much. Together they start to have a shape, and that shape is the only thing Naipa tries to show you.
For that shape to exist, things have to be grouped. And grouping means choosing the buckets, which is the most delicate decision in the whole project.
The seven
They are: fear, desire, loss, control, identity, hope and belonging.
They don’t come from any clinical manual, and that’s deliberate. They sit closer to basic psychological needs and to the “core themes” used in coaching than to any diagnostic category. They’re broad enough that almost any association fits into one or two, and distinct enough from each other that the count means something.
Each answer is filed under one or two categories, never more. Forcing a choice is what makes frequency informative: if everything landed in all seven, counting would tell you nothing.
What the app does with the count
Very little, on purpose.
At the end of the week it writes a short paragraph in an exploratory register — and it writes it from your answers, not from the tally. The model is handed both, and the prompt makes it quote at least two of your own phrases word for word. So it sounds like “‘the door was already shut’ came back on Thursday, the same words as Monday”. Never like “you have a control problem”.
The difference between those sentences isn’t stylistic. The first points at words you wrote and can go back and read; the second asserts something about you that nobody measured. Naipa is only entitled to the first.
There’s a third sentence it isn’t allowed either, and this one is the one we had to write into the prompt by hand: “this week the theme of control keeps coming up”. It’s true, it’s harmless, and it’s useless — the counts are printed right next to the paragraph. A reading that recites the categories back at you has said nothing you weren’t already looking at.
And there’s one more thing it does with the count: it hides it until the week is over. Across loose days you see no patterns at all, and the reading only arrives once there is a whole week behind it. The reason isn’t commercial: if we tell you on day four that “fear comes up a lot”, on day five you write with one eye on that sentence. The data spoils the moment you show it too early. (That a subscription is needed from the third reading on is a separate conversation, and it lives in the terms — it isn’t what decides when a pattern gets shown.)
Why this isn’t a test
Worth spelling out, because the format resembles things that are:
- It measures nothing. No score, no scale, no comparison to anyone. It’s a frequency count over what you wrote.
- It isn’t validated. No study says these seven buckets track real psychological constructs, and we don’t claim they do.
- It isn’t Rorschach. The cards are our own illustrations and words, not plates from a psychometric test, and there’s nothing like a norm table behind them.
- It replaces no one. If something you see worries you, that conversation belongs with a professional, not with an app.
What it is: a notebook that reads itself back to you. You leave the notes, it tells you which ones resemble each other. Interpreting them is still your job, and that part we have no plans to automate.
The interesting bit isn’t the category, it’s the change
A single week’s snapshot is the least useful part of all this. “Loss” showing up a lot this week could be a house move, a film, or nothing.
What starts to say something is the series: six straight weeks where “belonging” never appeared, and then suddenly it appears three times. That change is yours, you wrote it without meaning to, and no app suggested it to you. It’s the only kind of data this game can genuinely hand you.
That’s why Naipa asks for sixty seconds and not an hour. The quality of any one note matters little; having many of them, spread over time, is everything.