Contribution to the handbook
Supporting · Assessed 2026-10-28
The evidence gives useful context for a possible elicitation design: the authors found ordering 67 proposals challenging and cited earlier research about model difficulty with sets of three or more. This does not show that pairwise elicitation is preferable for civic decisions or resolve its practical tradeoffs, so it is supporting rather than central.
Does this challenge the goal?
In-scope gap
The question concerns a preference-elicitation method that may inform testing and participation design within the accepted goal; it does not challenge the goal.
The stated rationale is study-specific, and the evidence does not establish whether the cited model difficulty applies to other data or settings.
Why this category?
- useful background
- implementation context
- testing
The record reports the authors’ rationale and their reference to earlier research; it does not establish that larger-set difficulty holds for this setup or that pairwise choices suit other civic tasks.
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 Any answer below is provisional; parts whose evidence changed or was withdrawn are withheld.
Give the stated design rationale, including the difficulty of ordering the full set and the reason for preferring pairs over larger sets.
What we can say so far
The authors said ordering preferences across all 67 proposals was challenging. They focused on pairs rather than larger sets because earlier research indicated that language models struggle to predict choices when shown three or more proposals.
What remains open
- Whether the cited model difficulty applied in this study's specific setup is not established; the requested stated design rationale is nevertheless covered.
Conditions and limits
- The difficulty with sets of three or more proposals is attributed to earlier research, not presented as a new result of this study.
- The supplied material does not establish whether that model difficulty applied to this dataset or experimental setup.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
Why the experiment used pairs The authors divided the 67 proposals into sets of two or five because ordering preferences across all 67 was challenging. They focused on pairs, citing earlier research that language models struggle to predict choices when shown three or more proposals.
Provides the authors' stated reasons: ordering all 67 proposals was challenging, and earlier research suggested models struggle with three or more proposals.
Large Language Models (LLMs) as Agents for Augmented Democracy
Read full source: Large Language Models (LLMs) as Agents for Augmented Democracy (opens in a new tab) ↗Read source discussion: Large Language Models (LLMs) as Agents for Augmented Democracy →