Prefer candidate choropleth maps with stronger regionalized spatial patterns

When several mapped variables are already relevant to the article, keep the choropleth whose geographic distribution forms the strongest regional pattern. For example, compare candidate state or county maps with a spatial autocorrelation score such as Moran’s I and prefer the higher-scoring view over a low-saliency map with random-looking regions.

Readers can grasp a geographically structured pattern faster than a scattered one. A map with visible regionalization gives the viewer something to interpret at a glance and makes the display feel more informative.

Mechanism: Stronger spatial patterning increases visual interest during overview reading, so the chosen map is more likely to draw attention and communicate geographic structure.

Evidence: Maps selected with spatial-pattern saliency were rated more visually interesting than low-saliency versions built from randomized data, and Moran’s I showed a positive correlation with overall map ratings in the study (Gao et al., 2014).

• User Goal: Choose the most engaging geographic view among several relevant options.
• Data: Multiple relevant georeferenced variables are available over comparable regions.
• Chart Setting: One thematic map must be selected from several candidate choropleth views.
• Success Criterion: Readers judge the map as visually interesting and useful for explaining regional variation.

Break it when: The candidate maps are not all already relevant to the article. Why: The paper’s ranking strategy prioritizes article relevance before visual interestingness.

Sacrifice: You may drop a less patterned but still relevant alternative. Risk: Blind saliency optimization can favor an eye-catching map that explains the wrong measure. Mitigation: Apply spatial-pattern ranking only within the high-relevance candidate set.

Mistake: Maximizing spatial pattern before filtering by article match. Why it fails: A map can look interesting while still being off-topic.

Failure Sign: A lower-relevance map is chosen only because its pattern looks stronger. Quick Check: Confirm that each candidate first passes the relevance filter, then compare their spatial autocorrelation scores. Stronger Test: Show reviewers two equally relevant candidates and test whether the higher-saliency map is judged more interesting.

• Remove low-relevance map candidates before any saliency comparison.
• Compute a spatial autocorrelation score for the remaining choropleth candidates.
• Select the highest-scoring map only from that relevance-filtered set.
• Re-run the comparison if a more relevant candidate enters the shortlist.