2 Pictorial stimuliAs in Mirandola et al. (2014), a set of episodes (also named ‘‘scripts’’) with varying emotional valence were used as encoding stimuli. The material for each episode consisted of a sequence of 15 colour pictures depicting an ordinary event (e.g., going grocery shopping, waking up in the morning, etc.); out of them, 11 were actually shown, in chronological order, during the encoding phase, while the remaining 4 were used as distractors in the recognition test. The Appendix shows how one episode was presented during the encoding phase and which pictures were later used at thememory test. Other photographs were created to depict three different outcomes for each episode. Each outcome was depicted by a sequence of two pictures and was either neutral, positive, or negative (positive and negative being associated with high arousal). Valence of the outcomes was counterbalanced across participants,such that each of the three alternative outcomes of each episodes was presented to one third of the participants. The nine episodes were presented one after the other without interruptions. In order to avoid primacy and recency effects, 5 script-inconsistent pictures were shown at the very beginning of the presentation, and other 5 at the very end. Each episode had different characters and settings.Four old (i.e. ‘‘targets’’) and four new (i.e. ‘‘distractors’’) photographs were presented for each episode during the recognition test; one of the four distractors was named causal antecedent, as it depicted the causal antecedent whose outcome had been presented during the encoding phase, while the other three were named gap-filling pictures. Overall, 45 old and 45 new photographs were presented during the recognition test in a randomized order.Negative, positive, and neutral outcomes of each episode significantly differed in valence (ps <.001), and all the negative and positive consequences were significantly more arousing than the neutral ones (ps <.001), as indicated by previous ratings provided by 11 independent judges using the SAM (Self-Assessment Manikin) scale (Lang, 1980).2.3. Procedure2.3.1. Encoding phase and retention intervalParticipants were tested individually in a quiet room in their own school and were told that they would see a series ofphotographs depicting young people doing different daily activities,and that their task would be to understand what happened in the different situations. The entire encoding phase consisted of a series of 127 photographs, each of which was presented for 2 s and was followed by a black screen lasting 2 s as interstimuli. The stimuli were presented on a computer screen placed about 50 cm from the head of the participant. During the presentation, the room was dim. A 15-min retention interval followed the encoding phase.2.3.2. Recognition phaseMemory was tested with a self-paced recognition task. Participants had to respond ‘‘yes’’ or ‘‘no’’ to each photograph,depending on whether they thought they had seen it during the encoding phase or not. A confidence rating (with three options,i.e. ‘‘uncertain’’ scored as 1, ‘‘pretty confident’’ scored as 2, or ‘‘very confident’’ scored as 3) was also collected for each response.3. ResultsTable 1 reports the mean proportions (with standard errors) of hits (i.e. correct recognitions) and causal errors in the three types of contexts (i.e., positive, negative, and neutral episodes). We then calculated the d’ scores, a measure of recognition memory performance that combines hit and false alarm rates in an overall index of ability to discriminate new vs. old stimuli; the d’ score was calculated by subtracting z(false alarm rate) from z(hit rate) (Macmillan & Kaplan, 1985). A 2 (group: Low sensation seekers vs. High sensation seekers) Â 3 (emotionality of the episode: positive vs. negative vs.neutral) mixed ANOVA was performed on the mean d’ scores. A significant interaction between group and emotionality emerged, F(2,90) = 3.87, p < .05, g2 = .08, such that d’ scores for negative episodes were significantly higher in low sensation seekers than in high sensation seekers (p < .01, at post hoc analysis, with Bonferroni correction). Further, in low sensation seekers the d’ scores for negative material were significantly higher than the d’ scores for positive material (p < .05), and also higher than the d’ scores for neutral material although not significantly (p = .10). Response bias C = À0.5[z(hit rate) + z(false alarm rate)] was also calculated, but the difference between high and low sensation seekers was not significant, nor it was the main effect of emotionality. A 2 (group: Low sensation seekers vs. High sensationseekers) Â 3 (emotionality of the episode: positive vs. negative vs. neutral) mixed ANOVA was also performed on the mean proportion of causal errors, with group varying between-participants and emotionality of consequence varying within-participants. No significant main effects or interaction emerged (ps >.10). Finally, two mixed ANOVAs similar to the above were performed on the mean confidence scores associated with hits and with false alarms. In either case no effect was significant (ps >.10).
การแปล กรุณารอสักครู่..
