Responses are stored privately while collection continues.
Census data can estimate how rare a demographic combination is. It cannot reveal what men or women expect. This study collects that missing preference layer without pretending a voluntary website sample represents every American.
Build a result and contribute →How do measurable partner filters, dating-spend expectations, and perceived realism differ across adult users who voluntarily contribute a calculator result?
The primary comparison will be descriptive. It will not claim that gender causes a preference or that a smaller calculated pool makes a standard wrong.
An age band, a broad gender category, the intended partner pool, five calculator filters, a monthly spend band, a payment expectation, and a one-to-five realism rating.
The research record excludes the generated headcount because it can be reproduced from the versioned filter snapshot and source model.
Participation is optional, restricted to adults, and requires affirmative consent. Each response is stored as a separate object in private U.S.-region storage. The research record does not store a name, email, free text, exact birth date, IP address, or user agent.
A first-party HTTP-only cookie records that the browser already participated. Hosting and security providers may still process ordinary request metadata as described in the Privacy Policy.
Automated validation rejects impossible filter values and missing consent. A hidden anti-bot field and a one-response browser cookie reduce casual duplication, but neither proves a respondent is unique.
Before publication, records will be checked for malformed, duplicate, automated, or internally inconsistent patterns. Every exclusion rule and resulting sample count will be disclosed.
No subgroup table will appear below 50 unweighted responses. Small categories will be combined or suppressed. Aggregate tables will report the denominator, missing-answer treatment, field dates, and unweighted sample size.
A CSV distribution and codebook can be released after review. Raw response-level files will remain private because combinations of demographic answers can increase re-identification risk.
This is a voluntary convenience sample of site visitors. It may overrepresent people interested in dating calculators, viral discussions, or unusually specific standards. It is not nationally representative without an independently recruited probability or quota sample and defensible weighting.
Results will therefore be labeled “Pool Report participants,” never “all men,” “all women,” or “Americans” unless a later representative study supports that language.