Skip to main content
. 2010 Jul 12;2010:984673. doi: 10.4061/2010/984673

Figure 2.

Figure 2

Distributions of the locations of people (upper histogram in each panel) and response category thresholds (lower histogram in each panel) on the common logit metric (positive values = more problems). Response category thresholds are the locations where there is a 50/50 probability of endorsing either of two adjacent categories and represent the quantitative “notches” on the latent ruler defined by a set of items. (a) Item set BFa (items 17, 18, 20–22); (b) Item set BFb (items 30, 32, 33, 37–39); (c) Item set APa (items 1, 2, 4–7, 11, 12); (d) Item set APb (items 13–16, 24); (e) Item set E (items 23, 25, 26, 28, 29, 36). Mean (SD) person locations are BFa (panel (a)), −1.115 (1.951); BFb (panel (b)), −1.363 (1.096); APa (panel (c)), −0.814 (2.326); APb (panel (d)), −0.671 (1.370); E (panel (e)), −1.967 (1.355).