O effect of 2883-98-9 custom synthesis condition [F(1,28) = 2.07, n.s., 2 = 0.04]. There was a significant interaction between these factors [F(1,28) = 7.00, p < 0.05, 2 = 0.07]. Under the non-interactive condition, the difference was larger for non-devalued images than for devalued images [F(1,28) = 19.38, p < 0.001], and of the non-devalued images, the difference under the non-interactive condition was greater than that under the interactive condition [F(1,56) = 7.60, p < 0.01]. Furthermore, the conversation data recorded during the experiment was analyzed. Two raters who were na e about the purpose of this study individually heard the audio data and rated how often negative statements were purchase Digitoxin delivered by either member of the PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19910816 pair; for each pair, this was done for the first half (15 s) and last half (15 s) of 
the 30-s session, using a score of 0?0 (i.e., 0 indicating that negative statements were not delivered at all; 10 indicating that all statements were negative). As the ratings between the two raters showed a high positive correlationFIGURE 2 | Mean likeability ratings for each image under each condition in Experiment 1. Error bars indicate SE.www.frontiersin.orgJanuary 2015 | Volume 5 | Article 1558 |ArigaSocial-devaluation effectFIGURE 3 | Mean difference (absolute values) in the ratings given by pairs under each condition and for each image type in Experiment 1. Error bars indicate SE.(r = 0.83, p < 0.01), I considered the ratings as reliable, and took their averages as negativity scores. The mean negativity scores for devalued and non-devalued images were calculated and then averaged across pairs (Table 1). A within-subject 2 (image type: non-devalued and devalued) ?2 (period: first half and last half) two-way ANOVA revealed no significant main effects of image type [F(1,14) = 3.89, n.s., 2 = 0.13] or period [F(1,14) = 0.19, n.s., 2 = 0.00]. There was a significant interaction between these factors [F(1,14) = 6.00, p < 0.05, 2 = 0.08]. For devalued images, negative information was delivered significantly more often by either member of the pairs in the last half than in the first half [F(1,28) = 4.20, p < 0.01], though there was no significant difference for non-devalued images [F(1,28) = 2.09, n.s.].DISCUSSIONAlthough participants evaluated identical images under both conditions, likeability was generally lower under the interactive condition than under the non-interactive condition. The interactive evaluation influenced the likeability ratings of objects even though preferences were individually recorded without discussion or agreement between partners. Furthermore, the decrease in likeability under the interactive condition was especially pronounced for images that elicited relatively consistent evaluations across individuals. This is consistent with my prediction derived from the previous studies (Brodbeck et al., 2002; Schulz-Hardt et al., 2006), which indicated that the effect of shared information was stronger when group members held the same initial preferences than whenTable 1 | Mean negativity scores for non-devalued and devalued images in the first and last half of the sessions in Experiment 1. First half Non-devalued Devalued 5.06 5.17 Last half 4.77 5.they held different preferences. In fact, the analysis of the contents of the discussion revealed that for devalued images, negative information was increasingly delivered (or shared) by either member of the pairs as the discussion progressed, though this was not the case for non-devalued images.O effect of condition [F(1,28) = 2.07, n.s., 2 = 0.04]. There was a significant interaction between these factors [F(1,28) = 7.00, p < 0.05, 2 = 0.07]. Under the non-interactive condition, the difference was larger for non-devalued images than for devalued images [F(1,28) = 19.38, p < 0.001], and of the non-devalued images, the difference under the non-interactive condition was greater than that under the interactive condition [F(1,56) = 7.60, p < 0.01]. Furthermore, the conversation data recorded during the experiment was analyzed. Two raters who were na e about the purpose of this study individually heard the audio data and rated how often negative statements were delivered by either member of the PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19910816 pair; for each pair, this was done for the first half (15 s) and last half (15 s) of the 30-s session, using a score of 0?0 (i.e., 0 indicating that negative statements were not delivered at all; 10 indicating that all statements were negative). As the ratings between the two raters showed a high 
positive correlationFIGURE 2 | Mean likeability ratings for each image under each condition in Experiment 1. Error bars indicate SE.www.frontiersin.orgJanuary 2015 | Volume 5 | Article 1558 |ArigaSocial-devaluation effectFIGURE 3 | Mean difference (absolute values) in the ratings given by pairs under each condition and for each image type in Experiment 1. Error bars indicate SE.(r = 0.83, p < 0.01), I considered the ratings as reliable, and took their averages as negativity scores. The mean negativity scores for devalued and non-devalued images were calculated and then averaged across pairs (Table 1). A within-subject 2 (image type: non-devalued and devalued) ?2 (period: first half and last half) two-way ANOVA revealed no significant main effects of image type [F(1,14) = 3.89, n.s., 2 = 0.13] or period [F(1,14) = 0.19, n.s., 2 = 0.00]. There was a significant interaction between these factors [F(1,14) = 6.00, p < 0.05, 2 = 0.08]. For devalued images, negative information was delivered significantly more often by either member of the pairs in the last half than in the first half [F(1,28) = 4.20, p < 0.01], though there was no significant difference for non-devalued images [F(1,28) = 2.09, n.s.].DISCUSSIONAlthough participants evaluated identical images under both conditions, likeability was generally lower under the interactive condition than under the non-interactive condition. The interactive evaluation influenced the likeability ratings of objects even though preferences were individually recorded without discussion or agreement between partners. Furthermore, the decrease in likeability under the interactive condition was especially pronounced for images that elicited relatively consistent evaluations across individuals. This is consistent with my prediction derived from the previous studies (Brodbeck et al., 2002; Schulz-Hardt et al., 2006), which indicated that the effect of shared information was stronger when group members held the same initial preferences than whenTable 1 | Mean negativity scores for non-devalued and devalued images in the first and last half of the sessions in Experiment 1. First half Non-devalued Devalued 5.06 5.17 Last half 4.77 5.they held different preferences. In fact, the analysis of the contents of the discussion revealed that for devalued images, negative information was increasingly delivered (or shared) by either member of the pairs as the discussion progressed, though this was not the case for non-devalued images.