Data and methodology.

What we store

From researchers: email, password encrypted via hash, studies and service requests. From participants: selected options, response date, consent and a pseudo-anonymous identifier derived from the browser session. We do not request name or email to respond.

Your selections are saved during the survey even if you do not finish. The researcher may download these incomplete participations, identified separately; they do not count as completed responses and are not included in statistics.

To resume progress and prevent duplicate participation, we use a protected connection identifier specific to each survey (public IPv4 address or IPv6 network prefix). We do not store the IP in the response record. Private browsing, another browser or signing in on the same connection resumes the same participation. People sharing that connection can resume its questions and see saved selections; after completion, no further participation is accepted from that connection. A browser cookie also recognizes an earlier participation if the connection changes.

Who can see the data

Visitors can see aggregate statistics as soon as the survey is published. Technical administrators may access the database to operate the platform. Hypotheses remain private until the researcher decides to reveal them in public results. Do not publish identifying information in titles, hypotheses, or questions.

Anyone can view aggregate interim statistics during the study. The researcher can view and download responses to their study.

If the researcher enables it, people who finish the survey can also download an Excel file with completed responses and statistics. It excludes incomplete responses, private study details, and network or device identifiers.

AI and infrastructure

The application and database are hosted on a private server. Cloudflare protects access. The assistant uses Kimi (Moonshot AI): it receives the hypothesis, design, saved questions and conversation; for results analysis it receives aggregated statistics, never individual records or email addresses. If you use external prompts, you choose the provider. Infrastructure providers may process IP addresses in technical logs.

Participation and limits

Adults only. No interventions, diagnoses, or medical recommendations are provided. You can leave before submitting. A voluntary survey does not establish causes, guarantee unique participants, or necessarily represent a population. The platform does not perform mandatory prior review of studies. Publishers must obtain any authorizations or assessments required for their research.

Conservation and contact

Study data is retained while published and the account remains active. Access, correction, or deletion requests: [email protected]. If you participated without identifying yourself, it may not be possible to locate your individual response. Backups have limited retention.

Initial version · October 2026.