Fit Matters: Format–Distance Alignment Improves Conversational Search

Paper Title

Fit Matters: Format–Distance Alignment Improves Conversational Search

Publication Info

• Topic area: Enhancing conversational search systems by aligning information presentation formats with users' psychological distance.
• Keywords: Conversational search, psychological distance, construal-level theory, information presentation formats, cognitive load, user experience, multimedia, adaptive systems, decision confidence, processing fluency.

Background and Problem

• Problem / challenge: Current conversational search systems fail to adapt response formats (e.g., granularity, media type) to users' psychological distance, which systematically shapes information-processing preferences.
• Significance: Misaligned formats can reduce decision confidence, increase perceived risk, and impair user experience, limiting the effectiveness of conversational AI in decision-making tasks.
• Motivation and related work: Prior research on construal-level theory (CLT) shows that psychological distance influences preferences for abstract vs. concrete information. However, conversational systems have not systematically applied CLT to optimize response formats, leaving a gap in user-centered design.

Solution

• Proposed approach: Format–distance alignment, which adapts information granularity (abstract vs. concrete) and media type (text vs. image-and-text) to users' psychological distance (near vs. far).
• Novelty:
  1. Empirically validates format–distance alignment as a design principle for conversational search systems.
  2. Demonstrates that cognitive load can be productive when aligned with task demands.
  3. Shows that multimedia benefits depend on complementarity between text and images, moderated by information granularity.
• Procedure and key techniques:
  • Conducted a 2 × 2 × 4 between-subjects experiment (N = 464) using travel-planning tasks.
  • Manipulated psychological distance (temporal/spatial × near/far) and information presentation formats (abstract text, detailed text, abstract image-and-text, detailed image-and-text).
  • Measured cognitive load, decision confidence, user perceptions (e.g., ease of use, usefulness, enjoyment, credibility), and behavioral markers in written travel plans.

Results

• Concrete findings:
  • Matched formats (e.g., concrete–near, abstract–far) improved decision confidence (5.42 vs. 4.88, d = 0.40), ease of use (5.59 vs. 5.01, d = 0.45), usefulness (4.99 vs. 4.32, d = 0.45), enjoyment (4.89 vs. 3.97, d = 0.63), and intention to use (3.39 vs. 2.81, d = 0.45).
  • Concrete formats imposed higher cognitive load (38.69 vs. 32.70, p < .001), but this load was productive when matched to near-distance tasks.
  • Images enhanced concrete text across multiple dimensions but provided no benefit for abstract text.
• Advantage over baselines:
  • Format–distance alignment consistently outperformed mismatched conditions across all user-experience metrics.
  • Multimedia formats (image-and-text) outperformed text-only formats for concrete information but not for abstract information.
• Experiments / evaluation:
  • Used factorial ANOVA to analyze interaction effects of psychological distance, granularity, and media type.
  • Validated manipulations through psychological distance and abstraction checks.
  • Analyzed written travel plans for exploratory and closure markers as behavioral evidence of cognitive engagement.
• Limitations and future work:
  • Generalizability: Tested only temporal and spatial distance; future work should include social and hypothetical dimensions.
  • Methodological constraints: Single-turn interactions; future studies should explore multi-turn dialogues.
  • Implementation challenges: Detecting psychological-distance cues in real-world systems requires further research.

Summary

This study introduces format–distance alignment as a novel design principle for conversational search systems, showing that aligning information granularity and media type with users' psychological distance improves decision confidence, perceived ease of use, usefulness, enjoyment, and intention to use. While concrete formats impose higher cognitive load, this load becomes productive when matched to near-distance tasks. Multimedia benefits depend on complementarity between text and images, with significant advantages for concrete but not abstract information. These findings provide actionable guidelines for designing adaptive conversational systems that dynamically tailor responses to users' cognitive states.