Page 206 - Demo

204AppendicesTable 4. Overview of differences in duration of emotional expressions between videos in which the daters were attracted to their partner or not in9-second videos.)Inconspicuous ConspicuousBehavior M SD 95% CrIs M SD 95% CrIs BF10Coyness cheek raised 345.46 740.481 17.14, 673.77 420.00 436.654 107.64, 732.36 0.37Coyness 872.73 864.20 489.56, 1255.89 470.00 437.29 157.18, 782.82 0.72Genuine smile 604.545 1035.318 145.51, 1063.58 1540.00 1200.19 681.44, 2398.56 2.20Polite smile 2177.27 1213.783 1639.11, 2715.43 1040.00 915.545 385.06, 1694.94 4.09attracted to their partner. We conducted three Bayesian Generalized linear mixed models with participant response (yes/no) as dependent variableand Attraction to Partner as a fixed effect. All models included a randomintercept per participant (nested in Group ID for Experiment 1).The results show that participants were indeed more likely to generallyrespond yes than no (Exp 1: β = 0.37, [0.27, 0.48], p+ = 100%; Exp 2: β= 0.23, [0.13, 0.33], p+ = 100%; Exp 3: β = 0.18, [0.10, 0.26], p+ = 100%).General response propensity was not influenced by Attraction to Partner(Exp 1: β = -0.06, [-0.12, 0.01], p− = 96.28%; Exp 2: β = -0.01, [-0.09,0.08], p− = 57.26%; Exp 3: β = 0.01, [-0.04, 0.07], p+ = 68.91%).Effect of gender congruence on the detection of attractionTo examine whether gender congruence (i.e., a match between the genderof the observer and the person observed) facilitates attraction detection, weincluded the fixed effects of Age Group (Experiments 1 and 2) or VideoCondition (Experiment 3), respectively, Shuffled, and Gender Congruence,as well as their interaction. The analysis was conducted separately for eachexperiment (see Table 4). We found no substantial evidence that gendercongruence facilitated attraction detection.Differences in sample characteristics between Experiments 1,2, and 3.A Bayesian independent samples t-test showed no differences in age betweenchildren in Experiment 1 and Experiment 2 (BF01 = 3.13). A Bayesian chisquare test showed that there were no differences in gender distributionbetween Experiment 1 and Experiment 2 (BF01 = 3.39).A Bayesian one-way Analysis of Variance (ANOVA) showed that therewere differences in age between adults (BF10 > 10). Specifically, the agemean in Experiment 2 was higher than Experiment 1 (BF10 > 10) andExperiment 3 (BF10 > 10). There were no differences in age between Experiment 1 and Experiment 3 (BF01 = 0.23). Bayesian chi-square testsTable 4. Overview of differences in duration of emotional expressions between videos in which the daters were attracted to their partner or not in9-second videos.)Inconspicuous ConspicuousBehavior M SD 95% CrIs M SD 95% CrIs BF10Coyness cheek raised 345.46 740.481 17.14, 673.77 420.00 436.654 107.64, 732.36 0.37Coyness 872.73 864.20 489.56, 1255.89 470.00 437.29 157.18, 782.82 0.72Genuine smile 604.545 1035.318 145.51, 1063.58 1540.00 1200.19 681.44, 2398.56 2.20Polite smile 2177.27 1213.783 1639.11, 2715.43 1040.00 915.545 385.06, 1694.94 4.09attracted to their partner. We conducted three Bayesian Generalized linear mixed models with participant response (yes/no) as dependent variableand Attraction to Partner as a fixed effect. All models included a randomintercept per participant (nested in Group ID for Experiment 1).The results show that participants were indeed more likely to generallyrespond yes than no (Exp 1: β = 0.37, [0.27, 0.48], p+ = 100%; Exp 2: β= 0.23, [0.13, 0.33], p+ = 100%; Exp 3: β = 0.18, [0.10, 0.26], p+ = 100%).General response propensity was not influenced by Attraction to Partner(Exp 1: β = -0.06, [-0.12, 0.01], p− = 96.28%; Exp 2: β = -0.01, [-0.09,0.08], p− = 57.26%; Exp 3: β = 0.01, [-0.04, 0.07], p+ = 68.91%).Effect of gender congruence on the detection of attractionTo examine whether gender congruence (i.e., a match between the genderof the observer and the person observed) facilitates attraction detection, weincluded the fixed effects of Age Group (Experiments 1 and 2) or VideoCondition (Experiment 3), respectively, Shuffled, and Gender Congruence,as well as their interaction. The analysis was conducted separately for eachexperiment (see Table 4). We found no substantial evidence that gendercongruence facilitated attraction detection.Differences in sample characteristics between Experiments 1,2, and 3.A Bayesian independent samples t-test showed no differences in age betweenchildren in Experiment 1 and Experiment 2 (BF01 = 3.13). A Bayesian chisquare test showed that there were no differences in gender distributionbetween Experiment 1 and Experiment 2 (BF01 = 3.39).A Bayesian one-way Analysis of Variance (ANOVA) showed that therewere differences in age between adults (BF10 > 10). Specifically, the agemean in Experiment 2 was higher than Experiment 1 (BF10 > 10) andExperiment 3 (BF10 > 10). There were no differences in age between Experiment 1 and Experiment 3 (BF01 = 0.23). Bayesian chi-square testsTable 4. Overview of differences in duration of emotional expressions between videos in which the daters were attracted to their partner or not in9-second videos.)Inconspicuous ConspicuousBehavior M SD 95% CrIs M SD 95% CrIs BF10Coyness cheek raised 345.46 740.481 17.14, 673.77 420.00 436.654 107.64, 732.36 0.37Coyness 872.73 864.20 489.56, 1255.89 470.00 437.29 157.18, 782.82 0.72Genuine smile 604.545 1035.318 145.51, 1063.58 1540.00 1200.19 681.44, 2398.56 2.20Polite smile 2177.27 1213.783 1639.11, 2715.43 1040.00 915.545 385.06, 1694.94 4.09attracted to their partner. We conducted three Bayesian Generalized linear mixed models with participant response (yes/no) as dependent variableand Attraction to Partner as a fixed effect. All models included a randomintercept per participant (nested in Group ID for Experiment 1).The results show that participants were indeed more likely to generallyrespond yes than no (Exp 1: β = 0.37, [0.27, 0.48], p+ = 100%; Exp 2: β= 0.23, [0.13, 0.33], p+ = 100%; Exp 3: β = 0.18, [0.10, 0.26], p+ = 100%).General response propensity was not influenced by Attraction to Partner(Exp 1: β = -0.06, [-0.12, 0.01], p− = 96.28%; Exp 2: β = -0.01, [-0.09,0.08], p− = 57.26%; Exp 3: β = 0.01, [-0.04, 0.07], p+ = 68.91%).Effect of gender congruence on the detection of attractionTo examine whether gender congruence (i.e., a match between the genderof the observer and the person observed) facilitates attraction detection, weincluded the fixed effects of Age Group (Experiments 1 and 2) or VideoCondition (Experiment 3), respectively, Shuffled, and Gender Congruence,as well as their interaction. The analysis was conducted separately for eachexperiment (see Table 4). We found no substantial evidence that gendercongruence facilitated attraction detection.Differences in sample characteristics between Experiments 1,2, and 3.A Bayesian independent samples t-test showed no differences in age betweenchildren in Experiment 1 and Experiment 2 (BF01 = 3.13). A Bayesian chisquare test showed that there were no differences in gender distributionbetween Experiment 1 and Experiment 2 (BF01 = 3.39).A Bayesian one-way Analysis of Variance (ANOVA) showed that therewere differences in age between adults (BF10 > 10). Specifically, the agemean in Experiment 2 was higher than Experiment 1 (BF10 > 10) andExperiment 3 (BF10 > 10). There were no differences in age between Experiment 1 and Experiment 3 (BF01 = 0.23). Bayesian chi-square testsTable 4. Overview of differences in duration of emotional expressions between videos in which the daters were attracted to their partner or not in9-second videos.)Inconspicuous ConspicuousBehavior M SD 95% CrIs M SD 95% CrIs BF10Coyness cheek raised 345.46 740.481 17.14, 673.77 420.00 436.654 107.64, 732.36 0.37Coyness 872.73 864.20 489.56, 1255.89 470.00 437.29 157.18, 782.82 0.72Genuine smile 604.545 1035.318 145.51, 1063.58 1540.00 1200.19 681.44, 2398.56 2.20Polite smile 2177.27 1213.783 1639.11, 2715.43 1040.00 915.545 385.06, 1694.94 4.09attracted to their partner. We conducted three Bayesian Generalized linear mixed models with participant response (yes/no) as dependent variableand Attraction to Partner as a fixed effect. All models included a randomintercept per participant (nested in Group ID for Experiment 1).The results show that participants were indeed more likely to generallyrespond yes than no (Exp 1: β = 0.37, [0.27, 0.48], p+ = 100%; Exp 2: β= 0.23, [0.13, 0.33], p+ = 100%; Exp 3: β = 0.18, [0.10, 0.26], p+ = 100%).General response propensity was not influenced by Attraction to Partner(Exp 1: β = -0.06, [-0.12, 0.01], p− = 96.28%; Exp 2: β = -0.01, [-0.09,0.08], p− = 57.26%; Exp 3: β = 0.01, [-0.04, 0.07], p+ = 68.91%).Effect of gender congruence on the detection of attractionTo examine whether gender congruence (i.e., a match between the genderof the observer and the person observed) facilitates attraction detection, weincluded the fixed effects of Age Group (Experiments 1 and 2) or VideoCondition (Experiment 3), respectively, Shuffled, and Gender Congruence,as well as their interaction. The analysis was conducted separately for eachexperiment (see Table 4). We found no substantial evidence that gendercongruence facilitated attraction detection.Differences in sample characteristics between Experiments 1,2, and 3.A Bayesian independent samples t-test showed no differences in age betweenchildren in Experiment 1 and Experiment 2 (BF01 = 3.13). A Bayesian chisquare test showed that there were no differences in gender distributionbetween Experiment 1 and Experiment 2 (BF01 = 3.39).A Bayesian one-way Analysis of Variance (ANOVA) showed that therewere differences in age between adults (BF10 > 10). Specifically, the agemean in Experiment 2 was higher than Experiment 1 (BF10 > 10) andExperiment 3 (BF10 > 10). There were no differences in age between Experiment 1 and Experiment 3 (BF01 = 0.23). Bayesian chi-square testsTable 4. Overview of differences in duration of emotional expressions between videos in which the daters were attracted to their partner or not in9-second videos.)Inconspicuous ConspicuousBehavior M SD 95% CrIs M SD 95% CrIs BF10Coyness cheek raised 345.46 740.481 17.14, 673.77 420.00 436.654 107.64, 732.36 0.37Coyness 872.73 864.20 489.56, 1255.89 470.00 437.29 157.18, 782.82 0.72Genuine smile 604.545 1035.318 145.51, 1063.58 1540.00 1200.19 681.44, 2398.56 2.20Polite smile 2177.27 1213.783 1639.11, 2715.43 1040.00 915.545 385.06, 1694.94 4.09attracted to their partner. We conducted three Bayesian Generalized linear mixed models with participant response (yes/no) as dependent variableand Attraction to Partner as a fixed effect. All models included a randomintercept per participant (nested in Group ID for Experiment 1).The results show that participants were indeed more likely to generallyrespond yes than no (Exp 1: β = 0.37, [0.27, 0.48], p+ = 100%; Exp 2: β= 0.23, [0.13, 0.33], p+ = 100%; Exp 3: β = 0.18, [0.10, 0.26], p+ = 100%).General response propensity was not influenced by Attraction to Partner(Exp 1: β = -0.06, [-0.12, 0.01], p− = 96.28%; Exp 2: β = -0.01, [-0.09,0.08], p− = 57.26%; Exp 3: β = 0.01, [-0.04, 0.07], p+ = 68.91%).Effect of gender congruence on the detection of attractionTo examine whether gender congruence (i.e., a match between the genderof the observer and the person observed) facilitates attraction detection, weincluded the fixed effects of Age Group (Experiments 1 and 2) or VideoCondition (Experiment 3), respectively, Shuffled, and Gender Congruence,as well as their interaction. The analysis was conducted separately for eachexperiment (see Table 4). We found no substantial evidence that gendercongruence facilitated attraction detection.Differences in sample characteristics between Experiments 1,2, and 3.A Bayesian independent samples t-test showed no differences in age betweenchildren in Experiment 1 and Experiment 2 (BF01 = 3.13). A Bayesian chisquare test showed that there were no differences in gender distributionbetween Experiment 1 and Experiment 2 (BF01 = 3.39).A Bayesian one-way Analysis of Variance (ANOVA) showed that therewere differences in age between adults (BF10 > 10). Specifically, the agemean in Experiment 2 was higher than Experiment 1 (BF10 > 10) andExperiment 3 (BF10 > 10). There were no differences in age between Experiment 1 and Experiment 3 (BF01 = 0.23). Bayesian chi-square testsTable 5. Overview of all Gender Congruency models for Experiments 1-3.Predictors Accuracy (Median estimate of the coefficient with 95% HDIModel 1 Model 2 Model 3β (95% HDI) β (95% HDI) β (95% HDI)Intercept -0.06 -0.13, 0.01 -0.02 -0.11, 0.07 0.01 -0.04, 0.07Age Group -0.14 -0.21, -0.07 -0.05 -0.14, 0.04Shuffled -0.01 -0.07, 0.06Gender Congruence 0.02 -0.04, 0.08 -0.01 -0.10, 0.09 -0.02 -0.07, 0.04VI3 0.03 -0.07, 0.12VI6 0.04 -0.06, 0.13VI9 0.05 -0.04, 0.14Age Group × Shuffled 0.04 -0.02, 0.11Age Group × Gender Congruence -0.01 -0.07, 0.06 0.01 -0.07–0.10Shuffled × Gender Congruence 0.02 -0.04, 0.09Age Group × Shuffled × Gender Congruence 0.06 0.00, 0.12VI3 × Gender Congruence 0.01 -0.08, 0.11VI6 × Gender Congruence -0.05 -0.15, 0.04VI9 × Gender Congruence 0.07 -0.02, 0.17Random EffectsVar(Participant) 0.00 0.00 0.00Var(GroupID) 0.00showed no differences in gender distribution between experiments (Experiments 1-2: BF10 = 0.58; Experiments 1-3: BF10 = 0.50; Experiments 2-3:BF10 = 0.23).Stimuli employed in the Emotion Recognition TaskRegarding the Emotion Recognition Task (ERT), we used stimuli from theFacial Expressions and Emotion Database (FEEDTUM; Wallhoff, Schuller,Hawellek, & Rigoll, 2006). The FEEDTUM database consists of 18 individuals displaying 7 spontaneously elicited emotional facial expressions(happiness, disgust, anger, fear, sadness, surprise, and neutral). Here, weonly included 10 actors (5 female) and opted to not include the emotionof disgust. Therefore, the final stimulus set consisted of 60 videos (6 emotional expressions × 10 actors). To ensure potential luminance confounds,the background of all videos was standardized (r = 128, g = 128, b = 128;Akdag, 2020) ). All videos were 2000 ms in length, whereby the first 500 msconsisted of a neutral expression and 1500 ms of an emotional expression.Table 5. Overview of all Gender Congruency models for Experiments 1-3.Predictors Accuracy (Median estimate of the coefficient with 95% HDIModel 1 Model 2 Model 3β (95% HDI) β (95% HDI) β (95% HDI)Intercept -0.06 -0.13, 0.01 -0.02 -0.11, 0.07 0.01 -0.04, 0.07Age Group -0.14 -0.21, -0.07 -0.05 -0.14, 0.04Shuffled -0.01 -0.07, 0.06Gender Congruence 0.02 -0.04, 0.08 -0.01 -0.10, 0.09 -0.02 -0.07, 0.04VI3 0.03 -0.07, 0.12VI6 0.04 -0.06, 0.13VI9 0.05 -0.04, 0.14Age Group × Shuffled 0.04 -0.02, 0.11Age Group × Gender Congruence -0.01 -0.07, 0.06 0.01 -0.07–0.10Shuffled × Gender Congruence 0.02 -0.04, 0.09Age Group × Shuffled × Gender Congruence 0.06 0.00, 0.12VI3 × Gender Congruence 0.01 -0.08, 0.11VI6 × Gender Congruence -0.05 -0.15, 0.04VI9 × Gender Congruence 0.07 -0.02, 0.17Random EffectsVar(Participant) 0.00 0.00 0.00Var(GroupID) 0.00showed no differences in gender distribution between experiments (Experiments 1-2: BF10 = 0.58; Experiments 1-3: BF10 = 0.50; Experiments 2-3:BF10 = 0.23).Stimuli employed in the Emotion Recognition TaskRegarding the Emotion Recognition Task (ERT), we used stimuli from theFacial Expressions and Emotion Database (FEEDTUM; Wallhoff, Schuller,Hawellek, & Rigoll, 2006). The FEEDTUM database consists of 18 individuals displaying 7 spontaneously elicited emotional facial expressions(happiness, disgust, anger, fear, sadness, surprise, and neutral). Here, weonly included 10 actors (5 female) and opted to not include the emotionof disgust. Therefore, the final stimulus set consisted of 60 videos (6 emotional expressions × 10 actors). To ensure potential luminance confounds,the background of all videos was standardized (r = 128, g = 128, b = 128;Akdag, 2020) ). All videos were 2000 ms in length, whereby the first 500 msconsisted of a neutral expression and 1500 ms of an emotional expression.Table 5. Overview of all Gender Congruency models for Experiments 1-3.Predictors Accuracy (Median estimate of the coefficient with 95% HDIModel 1 Model 2 Model 3β (95% HDI) β (95% HDI) β (95% HDI)Intercept -0.06 -0.13, 0.01 -0.02 -0.11, 0.07 0.01 -0.04, 0.07Age Group -0.14 -0.21, -0.07 -0.05 -0.14, 0.04Shuffled -0.01 -0.07, 0.06Gender Congruence 0.02 -0.04, 0.08 -0.01 -0.10, 0.09 -0.02 -0.07, 0.04VI3 0.03 -0.07, 0.12VI6 0.04 -0.06, 0.13VI9 0.05 -0.04, 0.14Age Group × Shuffled 0.04 -0.02, 0.11Age Group × Gender Congruence -0.01 -0.07, 0.06 0.01 -0.07–0.10Shuffled × Gender Congruence 0.02 -0.04, 0.09Age Group × Shuffled × Gender Congruence 0.06 0.00, 0.12VI3 × Gender Congruence 0.01 -0.08, 0.11VI6 × Gender Congruence -0.05 -0.15, 0.04VI9 × Gender Congruence 0.07 -0.02, 0.17Random EffectsVar(Participant) 0.00 0.00 0.00Var(GroupID) 0.00showed no differences in gender distribution between experiments (Experiments 1-2: BF10 = 0.58; Experiments 1-3: BF10 = 0.50; Experiments 2-3:BF10 = 0.23).Stimuli employed in the Emotion Recognition TaskRegarding the Emotion Recognition Task (ERT), we used stimuli from theFacial Expressions and Emotion Database (FEEDTUM; Wallhoff, Schuller,Hawellek, & Rigoll, 2006). The FEEDTUM database consists of 18 individuals displaying 7 spontaneously elicited emotional facial expressions(happiness, disgust, anger, fear, sadness, surprise, and neutral). Here, weonly included 10 actors (5 female) and opted to not include the emotionof disgust. Therefore, the final stimulus set consisted of 60 videos (6 emotional expressions × 10 actors). To ensure potential luminance confounds,the background of all videos was standardized (r = 128, g = 128, b = 128;Akdag, 2020) ). All videos were 2000 ms in length, whereby the first 500 msconsisted of a neutral expression and 1500 ms of an emotional expression.Iliana Samara 17x24.indd 204 08-04-2024 16:37