whether immigrants are generally good for America's economy) as opposed to utilizing a full‐range bar chart to depict the data would result, on average, in a 12.67% 4 difference in perception of issue‐specific ideological polarization, 24.13% difference in estimates of intergroup opinion overlap, a 20.98% difference in perceptions of group discreteness and a 21.05% difference in estimates of compromise likelihood—all of which represent highly significant statistical differences.
← all excerpts
How the manner in which data is visualized affects and corrects (mis)perceptions of political polarization.
1
—
—