Skip to main content
. 2014 Aug 25;39(2):83–103. doi: 10.1177/0146621614544195

Figure 1.

Figure 1.

Balanced condition: 15%, 30%, and 45% DIF items with no systematic advantage for one group; sample size varies from (250; 250) up to (1,500; 1,500); top row: false alarm rates; bottom row: hit rates in the balanced condition.

Note. DIF = differential item functioning; AO = all-other selection; SA = single-anchor selection; NC = next candidate selection.