RESEARCH NOTES · ORIGINAL SOURCES
What research can—and can’t—say about how people see one another.
ViewsMeet’s games create prompts, not diagnoses. These notes trace the ideas behind each mode to original studies, state the sample and limits, and keep validated instruments separate from playful translations.
- Primary sources
- Visible limits
- No relationship verdicts
- Reviewed August 25, 2026
TRY THE IDEA · NO ACCOUNT
Turn a research question into five everyday choices.
For a one-to-one read, pick five answers and send a private WhatsApp link. For a group split, freeze five questions and reveal only the aggregate after three responses. Both are conversation games—not validated measures.
How well do friends really know your preferences?
Familiarity can reveal patterns while confidence about exact choices still outruns accuracy.
Read the evidence → PARTNER PREFERENCESWhich preferences do partners predict best?
Common, strongly expressed likes may be easier to learn than quieter, idiosyncratic ones.
Read the evidence → GROUP INDEPENDENCEWhy ask before showing the split?
A social-influence experiment—and the published dispute around how its results should be read.
Read the evidence → ROOMMATE AGREEMENTDo roommates read each other better over time?
One longitudinal college sample found increasing personality-rating agreement, with important limits.
Read the evidence → SELF × OTHER PERSONALITYWhy can you and a friend see different sides?
Different vantage points can carry different information. Neither is the complete truth about a person.
Read the evidence →EDITORIAL BOUNDARY
Research-informed is not the same as validated.
The exact Mini-IPIP 20 self-report has published validation evidence. ViewsMeet’s friendship, partner, roommate, and Group Picks modes are low-stakes conversation games. Their scores do not measure closeness, empathy, compatibility, relationship quality, or group wisdom.
Read the product methodology →