Where's Waldo, Ohio? Using Cognitive Models to Improve the Aggregation of Spatial Knowledge.

Area of Science:

• Cognitive science
• Human-computer interaction
• Geographic information science

Background:

• The wisdom of the crowd phenomenon suggests group judgments can be more accurate than individual ones.
• Traditional aggregation methods like averaging may not fully leverage individual expertise or account for task variability.

Purpose of the Study:

• To improve the accuracy of collective intelligence in spatial knowledge tasks using cognitive modeling.
• To develop and evaluate models that incorporate individual expertise and task difficulty.

Main Methods:

• Participants estimated locations and confidence radii for US cities.
• Cognitive models were developed to infer group estimates, considering individual expertise and city difficulty.
• Models were applied to US city data and replicated with European city data.

Main Results:

• Statistical averages (simple and radius-weighted) outperformed individual majority judgments.
• Model-based estimates generally surpassed statistical averages in accuracy.
• Models allowing for varying individual expertise across cities showed the highest accuracy.

Conclusions:

• Cognitive modeling offers a superior approach to aggregating judgments in spatial tasks compared to simple statistical methods.
• Accounting for individual differences in expertise and task difficulty is crucial for maximizing collective intelligence.
• The developed models show promise for enhancing wisdom of the crowd applications in various domains.