Page 64 - Demo

62Chapter 3Figure 1. Conditional effect plot showing associations between Pre-dateattractiveness rating and Reaction Time (RT) separate per Gender. The blackline represents the median effect, while the grey ribbon represents the 95%credible interval.Date outcomeSecond, we investigated the association between Date outcome and ReactionTime using a Bayesian mixed model with a Gaussian distribution (Descriptives: Table S6-8; Model Table: Table S9). We found a robust effect ofDate again distractor picture on Reaction time: participants were slower by4.41 ms to respond to the probe if the distractor image depicted someonethey later indicated as a successful date compared to when the distractorimage depicted someone that they did not consider a successful date duringtheir speed-dates (bno−yes = -4.41 [1.96], 89% CrI [-7.51, -1.29], pd− = .99),and this effect did not substantially differ per Gender (b women−men = -2.48[3.93], 89% CrI [-10.30, 5.26], pd− = .74; see Figure 2 top panel).When investigating the effect o f D ate a gain p robe p icture o n ReactionTime, we did not find a r obust o verall e ffect (b no−yes = 1. 68 [1 .94], 89%CrI [-1.49, 4.70], pd+ = .81). However, we did find a r obust interactionwith Gender (bwomen−men = -9.33 [3.88], 89% CrI [-16.80, -1.60], pd− =.99). Therefore, we explored the effect o f D ate a gain p robe p icture withineach level of Gender. For women, we found no robust effect ( bwomen = -2.97[3.06], 89% CrI [-8.96, 3.12], pd− = 0.84). For men, on the other hand, wefound that they responded faster to the probe by 6.33 ms when it replacedFigure 1. Conditional effect plot showing associations between Pre-dateattractiveness rating and Reaction Time (RT) separate per Gender. The blackline represents the median effect, while the grey ribbon represents the 95%credible interval.Date outcomeSecond, we investigated the association between Date outcome and ReactionTime using a Bayesian mixed model with a Gaussian distribution (Descriptives: Table S6-8; Model Table: Table S9). We found a robust effect ofDate again distractor picture on Reaction time: participants were slower by4.41 ms to respond to the probe if the distractor image depicted someonethey later indicated as a successful date compared to when the distractorimage depicted someone that they did not consider a successful date duringtheir speed-dates (bno−yes = -4.41 [1.96], 89% CrI [-7.51, -1.29], pd− = .99),and this effect did not substantially differ per Gender (b women−men = -2.48[3.93], 89% CrI [-10.30, 5.26], pd− = .74; see Figure 2 top panel).When investigating the effect o f D ate a gain p robe p icture o n ReactionTime, we did not find a r obust o verall e ffect (b no−yes = 1. 68 [1 .94], 89%CrI [-1.49, 4.70], pd+ = .81). However, we did find a r obust interactionwith Gender (bwomen−men = -9.33 [3.88], 89% CrI [-16.80, -1.60], pd− =.99). Therefore, we explored the effect o f D ate a gain p robe p icture withineach level of Gender. For women, we found no robust effect ( bwomen = -2.97[3.06], 89% CrI [-8.96, 3.12], pd− = 0.84). For men, on the other hand, wefound that they responded faster to the probe by 6.33 ms when it replacedFigure 2. Conditional effect plot showing the effect of Date outcome onReaction Time (RT) separate per Gender. Values are conditioned on the otherpredictor set to “No”. Error bars represent 95% Credible Intervals.a picture of someone whom they later considered a successful date duringtheir speed-dates (bmen = 6.33 [2.36], 89% CrI [1.75, 11.00], pd+ = 1.00; seeFigure 2 lower panel).Voluntary attention (eye-tracking)Pre-date attractiveness ratingsWe first explored the association between Pre-date attractiveness rating andLooking time bias, using Bayesian zero-one inflated beta regression (Descriptives: Table S10-12; Model Table: Table S13). We found that attractivenessratings had a robust effect o n v oluntary a ttention. M ore s pecifically, participant’s attractiveness ratings of the left picture correlated positively withproportion of time spent looking at the left picture (b = 0.087 [.0050], 89%CrI [0.079, 0.095], pd+ = 1.00), while we found the opposite effect f or theattractiveness rating of the right picture (b = -0.098 [.0041], 89% CrI [-0.106,-0.091], pd− = 1.00). The results were similar for other values of Pre-dateattractiveness rating: increased attractiveness ratings of the left picture wereassociated with an increased probability of looking at the left picture, whilethe opposite was true for the right picture (Table S14).To see whether the effect was m odulated by G ender, we c ompared theslopes for men and women. However, we found no robust interaction between Gender and Pre-date attractiveness rating for both the left picture(bwomen−men = -0.001 [.010], 89% CrI [-0.026; 0.007], pd− = 0.83) and theIliana Samara 17x24.indd 62 08-04-2024 16:35