Study privacy notice
Vela Pen · Data protection
How study data are processed and how to exercise your rights. Notice version: asva-study-v5-us-180d-2026-11-03.
Data, hosting and the AI provider
Your questionnaire answers, product-page ratings, consent version and time, conversation transcript and necessary technical data are linked by a random participation identifier. We do not ask for your name, email address, student number or payment details. A pseudonym does not make your voice or speech anonymous.
The study application and its questionnaire database are hosted on a server in Germany, operated by Luke Boelling. When the call connects, the app sends your selected devices, essential pen requirements, consultation topics and optional question to ElevenLabs and the voice agent so the adviser can use them as background for this conversation. Only the first 600 characters of the optional question are sent. Other questionnaire answers, including willingness to pay, initial product interest, personal characteristics and attitude scales, are not sent to the agent. Only the study lead and authorised research collaborators may access the study records.
Saved call recordings are disabled in ElevenLabs, and this application does not save audio. Audio is still transmitted and processed to recognise speech, generate replies and produce the AI voice. Please avoid names, contact details, health information and information about other people in both the call and the optional question. The researcher reviews transcripts for identifying details before analysis; this is not automatic anonymisation.
Storage, deletion and backups
- ElevenLabs: conversation history, including the transcript and session background, is configured for 30 days. This is not a guarantee that every provider log, backup or legally required record is erased after 30 days.
- Active study content: answers, product-page responses, session background, transcripts and technical metadata expire 180 days after your participation record is created. A daily deletion run removes expired content within the next 24 hours. The provider’s 30-day setting is separate.
- Minimal records: the participation code, consent and withdrawal evidence and provider conversation links are kept only to document consent, handle deletion and prevent re-import. They expire 210 days after the same creation date and are removed by the daily deletion run.
- Backups: the policy requires encrypted, access-restricted backups with a rolling 30-day limit. Deleted content may remain in backup copies for up to 30 further days after deletion: approximately 210 days for study content and 240 days for minimal records, plus the daily cleanup interval. A restore must apply current deletion and expiry records before data are used.
- Research archive: only genuinely anonymised data may be archived for ten years, counted from creation of the final anonymous archive, for scientific documentation. The researcher must first review identifying content and remove the participation-code link. Removing names alone is insufficient; expiry does not automatically anonymise data. Anonymous archive records cannot be found or deleted by your code.
- Browser: this tab holds your participation identifier and unfinished answers in session storage. The completion and exit flows clear them after you save your code. There are no advertising or analytics trackers in the study application.
Selecting “End participation” stops the study and removes submitted answers, product-page responses, session background, device details and transcript content from the active study database. A minimal record of consent, withdrawal and the conversation link remains so the researcher can handle provider deletion and prevent withdrawn content from being imported again. Ending only the conversation does not withdraw your study data.
Provider deletion and deletion from previous exports require a separate follow-up by the researcher. Pseudonymous exports follow the same 180-day limit as active study content. Backups expire on their rolling 30-day schedule, and withdrawal records prevent deleted content from returning to use after a restore. Keep your participation code and send it to luke.boelling@fh-muenster.de to request deletion before anonymisation.
Consent, responsibility and your rights
Research processing relies on your consent under Article 6(1)(a) GDPR. The separate US-transfer confirmation records your explicit choice; it does not replace the required processor contract or verification of a lawful international-transfer arrangement.
The controller responsible for this study’s personal data is M. Sc. Luke Bölling, Corrensstr. 25, 48149 Münster, Germany: luke.boelling@fh-muenster.de. The study runs on the researcher’s German server. FH Münster is the researcher’s academic affiliation; it is not named as the controller. Affiliation does not state institutional ethics or data-protection approval.
Additional academic contact, FH Münster data-protection office: datenschutz@fh-muenster.de. This is not a data-protection officer appointed by the researcher; direct study data requests to the controller.
You can withdraw consent at any time without a reason or disadvantage. This does not affect the lawfulness of processing before withdrawal. You may request access, correction, erasure, restriction and, where applicable, data portability or object to processing. You may complain to a supervisory authority, including LDI Nordrhein-Westfalen. Please use your participation code for requests so no additional identifying information is needed.
The ElevenLabs workspace’s model-improvement setting is disabled for new data from 5 October 2026, 19:53 UTC. This does not retroactively erase earlier data or cover every processing operation by the provider and its service providers. Audio saving, retention and model improvement are separate settings. See the provider’s privacy policy, data processing addendum and data-residency information. Study contact details are also on the imprint.