Figure 2From: Does the design of correspondence studies influence the measurement of discrimination? Preferences and group differences in the means of unobserved variables. Back to article page