How to select appropriate hue ranges for sequential color schemes on choropleth maps? A quantitative evaluation using map reading experiments

We propose map reading experiments to quantitatively evaluate the
selection of hue ranges for sequential color schemes on choropleth maps.
In these experiments, 60 sequential color schemes with six base hues and
ten hue ranges were employed as experimental color schemes, and a total of
414 college students were invited to complete identification, comparison,
and ranking tasks. Both controlled and real-map experiments were
performed, each involving a web-based survey and an eye-tracking
experiment. In the controlled experiments, the shapes of the map objects
were relatively regular, and attribute data were randomized. In contrast,
the shapes were complex in real-map experiments, and real data were
employed. Our findings show that widely used color schemes with a hue
range of 0º yield poor performance in all tasks; 15º hue ranges yield good
performance in the comparison and ranking tasks but poor performance in
the identification task. For large hue ranges of 120-360º, participants
showed good performance in the identification task but poor performance in
the comparison and ranking tasks. For 30-60º hue ranges, participants
achieved excellent performance in the comparison and ranking tasks and
acceptable performance in the identification task. We also found that the
ratings of 0-60º ranges were high.