Science and safety
A mirror, not a label.
Every Glim quiz is a reflection that helps you put words to how you are. It is not a diagnosis. Here is how quizzes are made, what happens when an answer points to something heavy, and how we use AI.
How quizzes are made
Published research scales
Some quizzes use well-known research questionnaires, word for word, with the original scoring and credit to their authors. We do not change their wording, and we only use scales whose licences allow it.
Glim originals
Other quizzes are written by Glim and draw on research ideas, such as burnout, belonging or attachment. They say so plainly, and their result ranges are Glim's own until a local study checks them.
Reviewed before they go live
Every new quiz and every change starts as a draft and needs approval from a second person before it is published. Quiz content is reviewed by a licensed psychologist.
Written in the Glim voice
Questions are in clear English so everyone answers accurately, with a full Filipino version where we have one. Results speak like the voice in your own head, with an English meaning beside any Filipino line.
Every quiz has an About this quiz note that names what it is built on. Tap the small i on any quiz card to read it.
When an answer points to something heavy
Glim is not an emergency service. It is built so that no one who needs help leaves without knowing where to find it.
Heavy answers always show help
Some questions are marked as signs of crisis. If you choose one, your reading shows crisis lines above everything else, whatever the rest of your answers say. A heavy sign can only make a reading more careful, never lighter.
Changed answers still count
If you pick a crisis answer and then change it before finishing, Glim keeps that as a quiet signal and still shows you support lines.
Help is never behind a payment
Readings that point to something heavy, and their support lines, are never locked, even on premium quizzes. They also cannot be shared.
A person reads the wall
Every note on the wall is read by someone on the Glim team before it shows. Notes with personal details are refused, and notes with signs of crisis are not published; the writer sees support lines instead.
Your diary stays yours
Tala entries are private. If an entry shows signs of crisis, Glim shows support lines and checks in with you the next day, and that entry is never sent to AI.
A way to ask for help at school or work
When you take a quiz through your school, workplace or LGU, your reading can show their support contact. You can ask them to reach out. Only your name and how to reach you are shared, never your answers.
If you or someone else is in danger right now
Call 911, or the NCMH Crisis Hotline at 1553. The Need help? button at the top of every page shows the lines for your country.
How Glim uses AI
- AI reflections in Tala are opt-in: Glim only writes one when you tap Ask Glim to reflect.
- Only that entry is sent to the AI provider, without your history. Your content is not used to train their models.
- Replies are short, warm and checked before you see them. A reply that diagnoses, gives medical or medication advice, or claims to be a person is replaced with one written by the Glim team.
- Each reply is marked as written by AI, not by a professional.
- AI never decides anything about you. It does not score quizzes and does not handle crisis signs.
Your data
Private by default
Your readings are yours. When you share one, people see the headline only, never your answers. Mental health readings can never be made public on your profile.
Groups, never individuals
Schools, workplaces and LGUs see group totals only. Small groups are combined so no one can be singled out, and written answers never appear in their reports.
Never sold
Glim does not sell personal information and does not use your mental health information for ads, under the Data Privacy Act of 2012.
Questions about how Glim works?
The FAQ answers the common ones. For anything else, write to us.