Contribution to the handbook
Supporting · Assessed 2026-10-27
The question helps assess the reliability and limits of a study measure used to analyze positions, but it is a narrow methodological contribution rather than a governance design or direct test of citizen outcomes.
Does this challenge the goal?
In-scope gap
Checking whether text-derived positions correspond to stated ratings is a measurement-validation detail relevant to testing deliberative systems within the defined goal.
The supplied records report a relationship within the study but do not establish validity across issues, groups, or other applications.
Why this category?
- Addresses whether a text-based measure reflects participants’ stated positions.
- Preserves the qualification that correspondence was not exact and may not transfer to other issues or groups.
- Useful for interpreting study results, but does not establish a system capability or governance safeguard.
Both supporting records interpret the same study source. The packet does not provide validation across other settings or establish that the measure captures every meaning in participants’ written opinions.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
Tests, stop conditions and unresolved work →
Read the exploration goalNew evidence is being assessed. Any answer below is provisional; parts whose evidence changed or was withdrawn are withheld.
State whether text-derived positions corresponded to participants’ stated ratings and qualify how closely they matched.
What we can say so far
Yes. The reported relationship between participants’ written opinions and their ratings was strong, and the text-derived scores were correlated with earlier ratings. The match was not exact, so this supports strong qualitative correspondence rather than one-to-one agreement.
What remains open
- The supplied evidence gives no numerical correlation or other exact measure of the strength of agreement, so its magnitude cannot be quantified more precisely.
Conditions and limits
- The comparison concerns participants’ written opinions and ratings of the same positions in this study.
- The reported association does not establish that the measure captures every meaning or nuance in the opinions, or that it generalizes to other issues or groups.
Objections to consider
- The reported correlation was not exact, and the supplied evidence does not quantify the unexplained variation or establish a more precise degree of agreement.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
Text matched stated positions Participants’ written opinions lined up strongly with their ratings of the same positions, suggesting that the AI’s text-based measure captured meaningful differences in viewpoint.
Reports a strong relationship between participants’ ratings and the position measure derived from their written opinions; limits the claim to this study.
AI can help humans find common ground in democratic deliberation
Read full source: AI can help humans find common ground in democratic deliberation (opens in a new tab) ↗Read source discussion: AI can help humans find common ground in democratic deliberation →
qualifies
Text scores used a question-specific scale The researchers encoded each written opinion as a 768-number text vector, then rotated those vectors onto a question-specific line running from agreement to disagreement. They checked the resulting position scores against participants’ earlier ratings; the scores were correlated, but the match was not exact.