Fig. 3.

Scattergrams showing the relationship of star cancellation omissions with DBE (top), and with EWB and EWS (bottom). Dotted lines show cut-offs for right and left neglect on DBE and EWB measures, and symbol shape indicates neglect status on star cancellation. Symbol colour indicates EWS performance in terms of z-score relative to controls, with negative (blue) scores representing low EWS and positive (red) scores high EWS (z scores are not distinguished below −2, although they ranged as low as −5.8). Panel (a) shows that the relationship of star cancellation omissions with DBE is relatively poor (r = 0.38). Panel (b) shows that the relationship with EWB is much stronger (r = 0.58), though imperfect. The grey line is the best fitting straight line. Patients above this fit line, who omit more targets than predicted from their EWB score, tend to have low EWS (blue symbols). Conversely, patients with high EWS (red symbols) tend to be below the fit line, omitting fewer targets than expected. This partial relationship is depicted in panel (c), which plots residual EWS against residual omissions (r = −0.49), after the influence of EWB has been removed. The multiple correlation of EWB and EWS with star omissions is r = 0.71 (see Table 2). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)