Page 89 - Demo

87Replication of Maner et al. (2005)4Experiment 2Preliminary analysesTo examine whether the mood induction was successful in inducing fear, weconducted an ordinal model on fear ratings with Condition (control vs. fear)as a fixed effect. The model included a random intercept per participant.The results showed that participants that watched the fearful video indicatedthat they felt more fear than people who watched the control video (OR =2.73[0.63], 95% HDI [1.72, 4.31], pd = 100%, d = 0.55[0.13]). This effectsuggests that the manipulation was successful in eliciting fear.Main analysesRegarding anger ratings (see Table 4), the model showed that contrary tothe findings of Maner et al. (2005), participants who watched the fearfulvideo did not rate Black men as angrier than Black women, White men, andWhite women compared to participants that watched the control video (b= 0.06, 95% HDI [-0.07, 0.19], pd = 81%). The interaction between TargetSkin Color × Target Gender was robust (b = 0.32, 95% HDI [0.19, 0.45], d= 100%, see Figure 2). Pairwise comparisons showed that participants ratedBlack female targets (M = 3.39, SD = 2.28) and White male targets (M =4.06, SD = 2.60) as angrier than White female targets (M = 3.00, SD =2.10; HPD: [-0.78, -0.05], HPD: [0.73, 1.46], respectively). Contrary to thefindings of Maner et al. (2005), participants rated White male targets asangrier than Black male targets (M = 3.22, SD = 2.12; HPD: [0.48, 1.21]).Participants also rated White male targets as angrier than Black femaletargets (HPD: [0.31, 1.04]). There was no difference between mean angerresponses for Black female and Black male targets (HPD: [-0.54, 0.19]), andWhite female and male Black targets (HPD: [-0.61, 0.11]).Female targets were rated as less angry than male targets (b = -0.23,95% HDI [-0.36, -0.10], pd = 100%), and participants who watched thecontrol video rated targets as angrier than participants who watched thefear-inducing video, (b = 0.27, 95% HDI [0.02, 0.52], pd = 98%). All othermain effects and interactions were not robust.The ordinal model supported the findings of the metric model (see Table5); indicating that the findings are more likely to be indeed reflective of atrue effect on the population.Experiment 2Preliminary analysesTo examine whether the mood induction was successful in inducing fear, weconducted an ordinal model on fear ratings with Condition (control vs. fear)as a fixed effect. The model included a random intercept per participant.The results showed that participants that watched the fearful video indicatedthat they felt more fear than people who watched the control video (OR =2.73[0.63], 95% HDI [1.72, 4.31], pd = 100%, d = 0.55[0.13]). This effectsuggests that the manipulation was successful in eliciting fear.Main analysesRegarding anger ratings (see Table 4), the model showed that contrary tothe findings of Maner et al. (2005), participants who watched the fearfulvideo did not rate Black men as angrier than Black women, White men, andWhite women compared to participants that watched the control video (b= 0.06, 95% HDI [-0.07, 0.19], pd = 81%). The interaction between TargetSkin Color × Target Gender was robust (b = 0.32, 95% HDI [0.19, 0.45], d= 100%, see Figure 2). Pairwise comparisons showed that participants ratedBlack female targets (M = 3.39, SD = 2.28) and White male targets (M =4.06, SD = 2.60) as angrier than White female targets (M = 3.00, SD =2.10; HPD: [-0.78, -0.05], HPD: [0.73, 1.46], respectively). Contrary to thefindings of Maner et al. (2005), participants rated White male targets asangrier than Black male targets (M = 3.22, SD = 2.12; HPD: [0.48, 1.21]).Participants also rated White male targets as angrier than Black femaletargets (HPD: [0.31, 1.04]). There was no difference between mean angerresponses for Black female and Black male targets (HPD: [-0.54, 0.19]), andWhite female and male Black targets (HPD: [-0.61, 0.11]).Female targets were rated as less angry than male targets (b = -0.23,95% HDI [-0.36, -0.10], pd = 100%), and participants who watched thecontrol video rated targets as angrier than participants who watched thefear-inducing video, (b = 0.27, 95% HDI [0.02, 0.52], pd = 98%). All othermain effects and interactions were not robust.The ordinal model supported the findings of the metric model (see Table5); indicating that the findings are more likely to be indeed reflective of atrue effect on the population.Table 4. Overview of the Gaussian multilevel model predicting anger ratings as a function of Target Gender, Target Skin Color, and Condition.Predictors Median b 95% HDI pdIntercept 3.41 3.15 – 3.66 100%Condition [Control] 0.27 0.02 – 0.52 98%Skin Color [Black] -0.11 -0.24 – 0.02 95%Target Gender [Female] -0.23 -0.36 – -0.10 100%Condition [Fear] × Skin Color [Black] 0.02 -0.11 – 0.15 60%Condition [Fear] × Target Gender [Female] 0.00 -0.13 – 0.13 51%Skin Color [Black] × Target Gender [Female] 0.32 0.19 – 0.45 100%Condition [Control] × Skin Color [Black] × Target Gender [Female] 0.06 -0.07 – 0.19 81%Random EffectsSD(Participant) 0.85Figure 2. Estimated marginal means of the posterior distribution for angerratings as a function of Target Sex (Female vs. Male) and Skin Color (Black vs.White). The points reflect the median and the error bars the 66% to 95%Highest Density Interval (HDI).DiscussionIn the present study, we examined the functional projection hypothesis byconceptually replicating the work by Maner et al. (2005). In contrast toManer et al. (2005), we found that men rate highly attractive female targetsTable 4. Overview of the Gaussian multilevel model predicting anger ratings as a function of Target Gender, Target Skin Color, and Condition.Predictors Median b 95% HDI pdIntercept 3.41 3.15 – 3.66 100%Condition [Control] 0.27 0.02 – 0.52 98%Skin Color [Black] -0.11 -0.24 – 0.02 95%Target Gender [Female] -0.23 -0.36 – -0.10 100%Condition [Fear] × Skin Color [Black] 0.02 -0.11 – 0.15 60%Condition [Fear] × Target Gender [Female] 0.00 -0.13 – 0.13 51%Skin Color [Black] × Target Gender [Female] 0.32 0.19 – 0.45 100%Condition [Control] × Skin Color [Black] × Target Gender [Female] 0.06 -0.07 – 0.19 81%Random EffectsSD(Participant) 0.85Figure 2. Estimated marginal means of the posterior distribution for angerratings as a function of Target Sex (Female vs. Male) and Skin Color (Black vs.White). The points reflect the median and the error bars the 66% to 95%Highest Density Interval (HDI).DiscussionIn the present study, we examined the functional projection hypothesis byconceptually replicating the work by Maner et al. (2005). In contrast toManer et al. (2005), we found that men rate highly attractive female targetsIliana Samara 17x24.indd 87 08-04-2024 16:35