779,582 synthetic US personas, each one traceable.
A persona is not a character sketch. It is a set of variables drawn from public survey data, with the distribution it came from kept alongside it. When it answers an item, the answer can be read against that distribution.
What is in the panel
13 dimensions
What the General Social Survey actually carries: redistribution, interpersonal and institutional trust, financial situation, health, life satisfaction, religiosity, abortion, guns, racial attitudes, environment, religious affiliation and political orientation. Seven of the thirteen are calibrated draws rather than predictors, and the interface labels each one.
43 variables
The survey variables every persona holds a position on. A new questionnaire is mapped onto these variables rather than invented from scratch.
Abstention over invention
When the calibration data leaves a segment ambiguous, the persona abstains. An abstention is a usable result; a fabricated answer is not.
How an answer is produced
Your item is read for what it asks: the dimension, the direction, the scale. The persona answers on that dimension, using the distribution it was drawn from, and the run records the spread across personas rather than a single average.
A wide spread on an item is a finding. It usually means the item is ambiguous, not that the population is divided.
Limits, stated plainly
The panel reflects the public data it was built from, with the gaps of that data. It is not a probability sample of the United States, and it does not produce a margin of error. Segments that are thin in the source data are thin here too, and the run says so instead of smoothing it over.