Schedule C
Cohorta Service Terms
This schedule applies to Cohorta, the synthetic survey service formerly published as Seoul-Survey. It adds to the main body and prevails over it for that Service.
Cohorta draws a panel of synthetic respondents from a persona corpus, rakes it onto a published population frame, and has that panel answer a questionnaire you supply. In this schedule, “Panel” means the drawn sample, “Study” means one questionnaire and its fieldwork, and “Response Set” means the answers the Panel produced.
A synthetic respondent is generated. It is not a real person, it was not recruited, and it did not answer a questionnaire. A Response Set is a model-based estimate of how a described population might answer, and it is not a survey, a poll, an opinion measurement or a statistical inference about any real population.
You must not present a Response Set, a segment split or any figure derived from them as the result of a survey of real people. Where you publish, circulate or supply such a figure to a third party, you must state, in a location a reader will see with the figure, that the respondents were synthetic and generated by Cohorta.
Cohorta publishes a validation grid measuring the Panel against a real survey, topic by topic and segment by segment, including the cells where it performs poorly and the cells with too little reference data to grade. Where a validation grade exists for the topic and segment you are quoting, you must not describe the figure with a confidence that grade does not support, and you must not quote a figure from an ungraded cell as validated.
A Response Set must not be used as, or presented as, a substitute for:
- official statistics, or a survey required by law, regulation, funding condition or research-ethics approval;
- an electoral poll, a referendum poll or a projection of a public vote;
- a clinical, epidemiological or safety study; or
- evidence of what an identified group of real people believes.
A population frame used by Cohorta is built from published statistics. Its source and reference period are stated in the Service, and a frame is a description of the population as published rather than as it stands today.
A questionnaire, audience definition, screening rule or experiment configuration you supply is Customer Material. You represent that you have the right to supply it, and that supplying it does not breach an obligation owed to a client, a funder or a research-ethics body.
Cohorta does not need, and you must not submit, the personal data of real respondents. Do not upload response-level microdata, contact lists, panel rosters or any file identifying individuals who took part in a real survey.
A risk flag, bias warning or question-quality label produced by the Service is an automated indication about the wording of a question. It is not an ethics review, a methodological approval or a guarantee that the instrument is fit for field use.
Where a Plan issues an API key for programmatic access, that key is credentials within the meaning of Registration and Account, and you are responsible for calls made with it.