As also a three-way HC-030031 interaction between Role, Cue Validity, and Hand Position, F(1,56) = 4.42, p = 0.04, 2 = 0.07 (see Figure 2). p A post hoc analysis comparing the cueing effects (difference in RT for valid and invalid cues) for all four conditions (Owner On Hands, Owner No Hands, Observer On Hands, Observer No Hands) revealed that the cost of orienting attention was greater when Owner responded to stimuli on their own hands compared to any other condition, all ts > 3.07, all ps < 0.003. No other comparison reached significance, all ts < 0.64, all ps > 0.525.FIGURE 2 | Response time plotted as a function of cue validity, hand position, and role (A). Participants were slower to respond to invalidly cued targets on their hands relative to off their hands, but only when stimuli appeared on their own hands and not when they appeared on another person’s hands. The cost 
to detecting an invalidly cued target is expressed as the difference between RTs for validly and invalidly cued targets (B). Error bars represent one within-subjects standard error of the mean.Owner DataTo assess how participants attended to stimuli presented on their own hands, we ran a 2 (Hand Position: On Hands vs. No Hands) ?2 (Cue Validity: Valid Cue vs. Invalid Cue) repeatedmeasures ANOVA with RTs from the Owner. Validly cued targetsreviewer raised the point that arbitrary RT cutoffs (e.g., 100?000 ms) would exclude more trials in the Owner than Observer conditions, because Owner responses were longer. To address this, we conducted an additional analysis by calculating the mean RT and SD for the Owner and Observer responses, and excluding trials that were <100 ms or >3 SDs for those means. This method would ensure that an equal number of trials were excluded in the Owner and Observer conditions. The results were similar to those reported in the main text. The critical three-way interaction remained significant, as well as the interaction between hand position and cue TG100 115 site validity for Owners, but not for Observers. The cueing effect for the Owner, On Hands condition remained longer than all other conditions. The means and SDs (in parentheses) follow. For the Owner condition: On Hands, Valid–393.88 (82.01); On Hands, Invalid–500.40 (84.98); No Hands, Valid–388.16 (64.80); No Hands, Invalid–479.29 (72.48). For the Observer condition: On Hands, Valid–334.90 (40.85); On Hands, Invalid–416.40 (37.88); No Hands, Valid–327.79 (38.49); No Hands, Invalid–411.59 (39.66).1Awere detected faster than invalidly cued targets, indicating a main effect of Cue Validity, F(1,56) = 620.50, p < 0.001, 2 = 0.92. In p addition, the position of the hands influenced RTs, indicated by a main effect of Hand Position, F(1,56) = 7.46, p = 0.008, 2 = 0.12. p Owners were slower to detect targets on their hands compared to off the hands. Critically, the strength of the cueing effect depended on the position of the hands; the interaction between Cue Validity and Hand Position, F(1,56) = 9.45, p = 0.003, 2 = 0.14 was p significant. A post hoc analysis comparing the cueing effect for the On Hands and No Hands conditions reveal that participants were especially slow to respond to targets appearing on their own hands relative to when their hands were far from the display, t(56) = 3.07, p = 0.003, d = 0.58.Observer DataTo assess whether attention orients slowly on someone else's hands, we conducted the same analyses described for the Owners with RTs from the Observers. Validly cued targets were detected.As also a three-way interaction between Role, Cue Validity, and Hand Position, F(1,56) = 4.42, p = 0.04, 2 = 0.07 (see Figure 2). p A post hoc analysis comparing the cueing effects (difference in RT for valid and invalid cues) for all four conditions (Owner On Hands, Owner No Hands, Observer On Hands, Observer No Hands) revealed that the cost of orienting attention was greater when Owner responded to stimuli on their own hands compared to any other condition, all ts > 3.07, all ps < 0.003. No other comparison reached significance, all ts < 0.64, all ps > 0.525.FIGURE 2 | Response time plotted as a function of cue validity, hand position, and role (A). Participants were slower to respond to invalidly cued targets on their hands relative to off their hands, but only when stimuli appeared on their own hands and not when they appeared on another person’s hands. The cost to detecting an invalidly cued target is expressed as the difference between RTs for validly and invalidly cued targets (B). Error bars represent one within-subjects standard error of the mean.Owner DataTo assess how participants attended to stimuli presented on their own hands, we ran a 2 (Hand Position: On Hands vs. No Hands) ?2 (Cue Validity: Valid Cue vs. Invalid Cue) repeatedmeasures ANOVA with RTs from the Owner. Validly cued targetsreviewer raised the point that arbitrary RT cutoffs (e.g., 100?000 ms) would exclude more trials in the Owner than Observer conditions, because Owner responses were longer. To address this, we conducted an additional analysis by calculating the mean RT and SD for the Owner and Observer responses, and excluding trials that were <100 ms or >3 SDs for those means. This method would ensure that an equal number of trials were excluded in the Owner and Observer conditions. The results were similar to those reported in the main text. The critical three-way interaction remained significant, as well as the interaction between hand position and cue validity for Owners, but not for Observers. The cueing effect for the Owner, On Hands condition remained longer than all other conditions. The means and SDs (in parentheses) follow. For the Owner condition: On Hands, Valid–393.88 (82.01); On Hands, Invalid–500.40 (84.98); No Hands, Valid–388.16 (64.80); No Hands, Invalid–479.29 (72.48). For the Observer condition: On Hands, 
Valid–334.90 (40.85); On Hands, Invalid–416.40 (37.88); No Hands, Valid–327.79 (38.49); No Hands, Invalid–411.59 (39.66).1Awere detected faster than invalidly cued targets, indicating a main effect of Cue Validity, F(1,56) = 620.50, p < 0.001, 2 = 0.92. In p addition, the position of the hands influenced RTs, indicated by a main effect of Hand Position, F(1,56) = 7.46, p = 0.008, 2 = 0.12. p Owners were slower to detect targets on their hands compared to off the hands. Critically, the strength of the cueing effect depended on the position of the hands; the interaction between Cue Validity and Hand Position, F(1,56) = 9.45, p = 0.003, 2 = 0.14 was p significant. A post hoc analysis comparing the cueing effect for the On Hands and No Hands conditions reveal that participants were especially slow to respond to targets appearing on their own hands relative to when their hands were far from the display, t(56) = 3.07, p = 0.003, d = 0.58.Observer DataTo assess whether attention orients slowly on someone else's hands, we conducted the same analyses described for the Owners with RTs from the Observers. Validly cued targets were detected.