Quantified Cognition privacy policy
Quantified Cognition analyzes your Anki history on your computer. Connecting a health account and contributing to research are choices you make separately. Research packages are sent only after you review and explicitly submit them.
Updated 4 November 2026. Health connections, research enrollment and research uploads are not yet available.
Your data and choices
JollyJoyJourney, operated by Christian Kleineidam, is responsible for the research service and its processing of personal data. Our address is Schweidnitzer Straße 6, 10709 Berlin, Germany. Contact contact@jollyjoyjourney.com.
You can use the add-on and export your results without participating in research. Participation is for adults aged 18 or over contributing their own data. Connecting a provider, joining research and sending a package are separate steps. Saying Yes in the initial sharing prompt sends nothing. Each upload requires a preview and an explicit Send; uploads are never automatic. The research information explains participation, benefits and risks.
Information used by the add-on
Local profile files hold calculated scores, health measurements, personal insights and sharing history. Sleep sessions and review times help match main sleep and naps to studying. Analysis can use temporary local Anki snapshots that contain card text. These files stay outside research uploads. Personal CSVs are saved to locations you choose.
Oura and Google Health connections import sleep durations, sleep stages and daily steps directly from the provider. Google's activity permission also covers other activity data; we fetch daily steps. Available history and missing measurements depend on the provider. You sign in on the provider's website. Our connection service exchanges and renews access tokens without storing them in the research database. The add-on saves credentials in your operating system's protected local store.
Research packages contain daily measurements: recall and overall, short-term and long-term memory formation scores; predicted and observed recall; evaluated, cohort, pending and excluded counts; follow-up coverage, uncertainty and quality indicators. Available health data include total, light, deep, REM and awake sleep durations, Google's restless time, sleep-period counts and daily steps. Main sleep before and after studying, nap counts, sleep-stage and duration summaries, gaps to studying, between-review nap counts and matching status are included when available.
Packages also contain dates, sources, method and Anki versions, an FSRS settings fingerprint, timezone, Anki day boundary, calculation settings, generation and sync times, provider date conventions, and relevant sleep types and algorithm versions. A random profile participant ID links contributions. Sharing choices, notice versions and the timing of sharing preferences, consent and first results help us understand participation patterns. Unknown earlier exposure is recorded as unknown; the full local event log is excluded.
Excluded from research packages: card questions and answers; card and deck names and IDs; individual reviews and review pairs; exact sleep or review times; raw health records and record IDs; heart rate; location; name and email; provider account or device IDs; passwords and tokens. Personal insight charts, cached correlations and return-visit comparisons are also excluded.
Eligibility and consent records are kept separately from measurements. They record the information agreed to, its version, the date, eligibility and a typed signature using a pseudonym you choose. Analysis exports can include enrollment version and consent timing; signatures and eligibility answers stay separate. Contact correspondence contains the sender address, message, attachments and delivery information.
Research and publication
We study how learning and recall vary over time, how they relate to sleep, physical activity and study patterns, and how to measure these relationships reliably. This includes validating the measurements, comparing sleep stages and naps, examining changes within and between participants, and assessing the reliability of findings.
New questions and methods can develop within this research area. Material changes to the purposes, data collected or access require updated information and any necessary fresh consent before the changed use.
Findings may appear in scientific publications, reports, presentations and public summaries. Public material uses anonymous summaries or synthetic examples. Individual daily histories remain under controlled access.
We follow Google's User Data and Health Research Policy, including its Limited Use requirements, when processing Google Health API data.
Access and safeguards
Authorized JollyJoyJourney researchers and approved academic, nonprofit or commercial collaborators can access data needed for this programme. Collaborator agreements limit purpose, access, security, retention and deletion, subject to provider permission. Partners and roles are identified in the research information before access. A partner responsible for its own processing must be identified separately, with additional consent where required.
We do not sell pseudonymous participant data or provide it for advertising, marketing or individual employment, insurance or credit decisions. The add-on does not send research measurements to AI services.
The research database is hosted by Supabase in Frankfurt, Germany. Hosting, connection services, support and subprocessors can involve processing outside that database region. Supabase's Data Processing Addendum describes its role and transfer safeguards, including Standard Contractual Clauses where applicable.
Our contact address forwards through Cloudflare Email Routing to Gmail. Cloudflare delivers messages; Google stores them. The Cloudflare and Google policies describe this processing, which can involve countries outside the European Economic Area. Correspondence is kept separately from research measurements. Research access outside the EEA requires an applicable adequacy decision or appropriate transfer safeguards; contact us for details or a copy of those safeguards.
We protect data with restricted access, encrypted connections, protected credentials and controlled research workspaces. Authorized hosting personnel can also process data to provide and secure the service. Research records are pseudonymous: dates, timezone and patterns in daily histories can still identify someone. These safeguards reduce, but cannot eliminate, the risks of identification or unauthorized access.
Data handling
Research relies on consent under Article 6(1)(a) GDPR, with explicit consent under Article 9(2)(a) for health-related data. Connecting a personal health account is a separate consent-based operation. Service security, handling enquiries and proportionate compliance evidence rely on legitimate interests under Article 6(1)(f), or a legal obligation where applicable.
We keep research measurements while needed for the programme's research, method validation or verification of findings. We regularly review the continuing need and safeguards, and delete or, where permitted, irreversibly anonymize records that are no longer needed. You can withdraw and request deletion. Applicable legal or provider retention and deletion limits take precedence.
Routine application security and service records are retained for up to 30 days. They can include IP addresses, request times, requested service, response status and abuse-control counters. Application logging is designed to exclude measurements, package bodies and credentials. Incident records may be kept longer for an investigation or legal obligation.
Routine correspondence is normally deleted within two years after resolution. Minimal consent, withdrawal and rights-request evidence is kept while needed to demonstrate lawful processing and normally for three years after the associated research data is deleted. This evidence contains no deleted measurements. A legal obligation or claim may require a different period, limited to the necessary records.
A random deletion marker containing no measurements may be retained longer where needed to prevent old clients or restored copies from reintroducing deleted data. Recovery copies are isolated from research use and retained for no more than 30 days. Deletions are reapplied before restored data can be used; shorter provider or legal deadlines also apply. Anonymous published summaries and synthetic data can remain available.
Your rights and contact
You can withdraw through Research sharing or request deletion through Delete uploaded data. Changing sharing to Not now or No blocks new uploads, removes unsent packages and queues deletion of earlier submissions from that profile. Once we receive withdrawal, we stop research use of the affected data and remove it from active storage and research extracts, including controlled collaborator copies. Withdrawal does not change the lawfulness of earlier processing.
Keep your profile until deletion is acknowledged, or contact us for help. Offline requests need to reach us before we can act on them. A participant ID or submission reference helps us locate records. We ask only for information needed to locate the data and verify your request.
Disconnecting a provider removes its local measurements and requests revocation of access. You can also revoke access in the provider's account settings. Research submissions and files you have exported have separate deletion controls; exported files remain under your control.
Subject to the GDPR's conditions, you can request access, correction, erasure, restriction and a portable copy of your data, and object to legitimate-interest processing. We normally respond within one month and explain any lawful extension. You can complain to a data-protection authority, including the Berlin Commissioner for Data Protection and Freedom of Information.
Policy and research-information updates carry version dates. They do not retrospectively expand consent. For questions or help with your data, contact contact@jollyjoyjourney.com.