- Email: [email protected]
Bias or equality? Unconscious thought equally integrates temporally scattered information
Consciousness and Cognition 25 (2014) 77–87
Contents lists available at ScienceDirect
Consciousness and Cognition journal homepage: www.elsevier.com/locate/concog
Bias or equality? Unconscious thought equally integrates temporally scattered information Jiansheng Li a,b, Qiyang Gao a, Jifan Zhou a, Xinyu Li a, Meng Zhang a, Mowei Shen a,⇑ a b
Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou 310028, China Department of Psychology, Northwest Normal University, Lanzhou 730070, China
a r t i c l e
i n f o
Article history: Received 1 May 2013 Available online 25 February 2014 Keywords: Unconscious thought Decision making Integration information
a b s t r a c t In previous experiments on unconscious thought, information was presented to participants in one continuous session; however, in daily life, information is delivered in a temporally partitioned way. We examined whether unconscious thought could equally integrate temporally scattered information when making overall evaluations. When presenting participants with information in two temporally partitioned sessions, participants’ overall evaluation was based on neither the information in the first session (Experiment 1) nor that in the second session (Experiment 2); instead, information in both sessions were equally integrated to reach a final judgment. Conscious thought, however, overemphasized information in the second session. Experiments 3 and 4 further ruled out possible influencing factors including differences in the distributions of positive/negative attributes in the first and second sessions and on-line judgment. These findings suggested that unconscious thought can integrate information from a wider range of periods during an evaluation, while conscious thought cannot. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction According to traditional wisdom, when confronted with a complex decision, people are accustomed to solving problems with deliberation. Nevertheless, a growing number of studies have suggested the opposite, that applying unconscious thought leads to a better decision than conscious thought (e.g., Bos, Dijksterhuis, & van Baaren, 2011, 2012; Dijksterhuis, 2004; Dijksterhuis, Bos, Nordgren, & Van Baaren, 2006; Ham, Van den Bos, & Van Doorn, 2009; Hasford, 2014). Unconscious thought is defined as ‘‘object-relevant or task-relevant cognitive or affective thought processes that occur while conscious attention is directed elsewhere’’ (Dijksterhuis & Nordgren, 2006, p. 96). A typical example was demonstrated by Dijksterhuis et al. (2006), who required participants to choose the best car from four alternatives that were each assigned positive or negative attributes. Compared with participants who made decisions after a period of conscious thinking, those who did not have the opportunity to deliberate (i.e., distracted by a 2-back task; Jonides et al., 1997) actually made relatively advantageous decisions. The researchers reasoned that unconscious thought has a higher processing capacity, allowing people to simultaneously process more information and make better decisions, while conscious thought focuses on limited dimensions (see also Dijksterhuis & Nordgren, 2006) and is less effective for dealing with complex problems.
⇑ Corresponding author. Address: Department of Psychology and Behavioral Sciences, Xixi Campus, Zhejiang University, Hangzhou 310028, China. E-mail address: [email protected] (M. Shen). http://dx.doi.org/10.1016/j.concog.2014.01.012 1053-8100/Ó 2014 Elsevier Inc. All rights reserved.
78
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
The effects of unconscious thought have been observed across many contexts, including solving creative problems (Yang, Chattopadhyay, Zhang, & Dahl, 2012; Zhong, Dijksterhuis, & Galinsky, 2008), predicting soccer matches (Dijksterhuis, Bos, Van der Leij, & Van Baaren, 2009; González-Vallejo & Phillips, 2010), making moral judgments (Ham & Van den Bos, 2010), selecting job applicants (Messner, Wänke, & Weibel, 2011), and doing simple arithmetic (Ric & Muller, 2012). Creswell and colleagues provided evidence from cognitive-neuroscience that brain regions activated when encoding information for decision making continue to be active while the brain responds to other, irrelevant, tasks (Creswell, Bursley, & Satpute, in press). In previous studies on unconscious thought, information for decision-making was presented to participants in one continuous session. However, in daily life, information frequently is delivered in a temporally partitioned way. For instance, while making a decision to purchase a mobile phone, a consumer may first ask for opinions from close friends and/or consult online assessments and then visit stores to check various models before deciding which phone to buy. Thus, when people are making a decision, time elapses between periods of information gathering. This raises the question of how unconscious thought integrates information for making an evaluation when information is presented in a temporally partitioned manner. Studies on conscious thought have demonstrated an emphasis on the most recently presented information when making a decision, otherwise known as the recency effect (e.g., Crano, 1977; Lichtenstein & Srull, 1987). For instance, evidence presented near the end of a prosecution’s defense was found to have the greatest impact on juror verdicts (Costabile & Klein, 2005). The same have been found for clinical judgments; the most recently received information has a relatively more powerful influence upon diagnoses given by junior, and even more experienced, physicians (Chapman, Bergus, & Elstein, 1996). Montgomery and Unnava (2009) explained that information presented more recently is much more easily retrieved from memory, and hence, will be regarded as the main source of information for making an overall evaluation, during which the recency effect inevitably occurs. More interestingly, previous research on conscious thought has demonstrated that whether or not information about an object is presented in a temporally partitioned way has distinct influence on people’s evaluation of the object. For instance, Ariely and Zauberman (2000, 2003) showed that people preferred sequentially presented information that was improving in intensity to that of declining intensity. However, people’s preferences for increasing sequences would be less favorable if information was partitioned. In addition, another experiment indicated that pleasant experiences would be more enjoyable and that unpleasant experiences would be more irritating, if pleasant and unpleasant experiences were partitioned (Nelson & Meyvis, 2008). Given the above findings regarding conscious thought, it is intriguing to ask whether information presented in a temporally partitioned manner also would profoundly influence unconscious evaluations. In other words, will unconscious thought place particular emphasis on information presented on a given temporal session (e.g., the most recent session) when people make evaluations? Alternatively, will unconscious thought integrate all partitioned information equally regardless of the order of presentation? Usher, Russo, Weyers, Brauner, and Zakay (2011, Experiment 3) designed an experiment asking people to choose from three hypothetical individuals as a flat mate. Each individual was characterized by 12 attributes. The percentage of positive attributes for each of the potential flat mates was 66%, 50%, and 33%. Information regarding these attributes was presented to participants in three sessions, with four attributes presented within each one. In addition, each mobile phones’ attributes were shown in a different color to help participants distinguish them from each other. Before the presentation, conscious thinkers were informed that ‘‘research has shown that the best decisions are the ones made using logic and rational thought,’’ and were asked to deliberate for 1 min after each session. Unconscious thinkers, on the other hand, were told that ‘‘research has shown that the best decisions are the ones made using intuition,’’ and were required to complete a 1-min interfering task after each session. At the end of the experiment, participants were asked to score each potential flat mate and choose the one that they most preferred. Results showed that accuracy and the discrimination rates of the unconscious thought group were more reasonable than those of the conscious thought group (Usher et al., 2011). However, this study had several limitations: First, experimental instructions differed between the two groups, which may lead to confounding influence regarding the final conclusion. It was firmly asserted that experimental instructions had a profound impact on the unconscious thought effect (Lassiter, Lindberg, González-Vallejo, Bellezza, & Phillips, 2009). Second, without a valid controlled experiment in which all information was presented in a single session, it is possible that the results were due to the different font colors rather than the partitioning of information. Third, researchers could not exclude the on-line judgment account, which might be formed due to the limited amount of information presented in each session (12 pieces per session). Finally, although the three flat mates had obvious general differences, there was no strict control of the relative merits of an individual flat mate (i.e., there was no objective data that A was better than B) in each session. This made the attributes in each session ambiguous, which may weaken each session affecting the overall evaluation. Based on these limitations, the present study intended to further examine the question of whether unconscious thought can integrate partitioned information without bias. We used four hypothetical mobile phones as experimental materials. Information about the phones was successively presented in two temporally partitioned sessions. We manipulated the information presented in the two sessions so that participants’ final evaluations would differ depending on whether information about the phones was integrated equally across two sessions or information from either session was particularly emphasized. 2. Experiment 1 Experiment 1 was designed to make an overall evaluation of unconscious thought by directly investigating whether it could equally integrate all temporally scattered information, or would attach particular importance to one certain piece
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
79
of information. In this experiment, information (attributes) of four hypothetical mobile phones was displayed in two separate sessions, although our study was only concerned with two of the phones. In the first session, one phone was clearly better than the other was because it possessed the highest number of positive attributes and the lowest number of negative attributes, while the other possessed the lowest number of positive and highest number of negative attributes. However, in the second session, the two phones were the same, possessing the same number of positive and negative attributes. Therefore, overall, one phone was clearly better than the other was. Our hypothesis stated that if unconscious thought equally integrated information from both sessions, participants would be able to distinguish between the superior and inferior phones. 2.1. Method 2.1.1. Participants and design Thirty-six Chinese undergraduate students (16 women and 20 men) from Northwest Normal University were equally but randomly assigned into either the unconscious-thought condition (7 men and 11 women) or the conscious-thought condition (9 men and 9 women). A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design was employed. In this and each subsequent experiment, the thought condition served as a between-subject variable and phone type was the within-subject variable. The scores participants assigned to the phones at the end of the experiment served as dependent variables. Participants all received financial compensation for their participation. 2.1.2. Materials and procedure Four hypothetical phones, respectively labeled as Best–Avg, Worst–Avg, Avg–Avg1 and Avg–Avg2, were used as research materials (labels were not accessible to participants). Each phone was assigned 16 attributes.1 The phone labeled as Best–Avg possessed 10 positive and 6 negative attributes, the Worst–Avg had 6 positive and 10 negative attributes, and both Avg–Avg1 and Avg–Avg2 had 8 positive and 8 negative attributes. In other words, Best–Avg was considered the superior choice, Worst– Avg was the inferior choice, and Avg–Avg1 and Avg–Avg2 were considered to be the same. Phone attributes were presented sequentially in two temporally partitioned sessions. In each session, 8 attributes of each phone were presented, with a total of 32 attributes presented. The order of presentation was counterbalanced across participants. In the first session, 6 positive and 2 negative attributes of Best–Avg were presented along with 2 positive and 6 negative attributes of Worst–Avg. In the second session, the remaining 6 positive and 6 negative attributes of either phone were presented. In other words, although the Best–Avg phone was clearly superior to Worst–Avg, overall, the second session of information presentation by itself would suggest Best–Avg and Worst–Avg were equally preferable. Four positive and 4 negative attributes of Avg–Avg1 and Avg–Avg2 also were presented in each session. In preparation for the task, participants were told they would read information about four phones on the screen and the presentation of information was divided into two sessions, one after another. Sentences describing the attributes of the four phones were randomly presented for 5 s each. At the end of each session, the conscious thinkers were asked to think about their impression of the phones for 3 min. Meanwhile, unconscious thinkers were required to complete a 2-back task, which also lasted for 3 min. At the end of the experiment, all participants indicated their attitude toward each phone on a scale of 0–100, with a higher score indicating a greater preference for the phone. 2.2. Results and discussion In the current experiment, Avg–Avg1 and Avg–Avg2 performed the function of adding complexity to the decision-making, therefore, only the scores of Best–Avg and Worst–Avg were subjected to data analysis. A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed-design ANOVA was conducted on participant scores. Results did not show a significant main effect of thought condition, F(1, 34) = .99, p > .05; however, the main effect of phone type was significant, F(1, 34) = 7.25, p < .05. Scores for the Best–Avg phone (M = 75.25, SD = 9.80) were significantly higher than for the Worst–Avg phone (M = 69.83, SD = 11.37). The interaction of thought condition and phone type also was significant, F(1, 34) = 16.14, p < .001. As shown in Fig. 1, conscious thinkers rated the Best–Avg (M = 72.50, SD = 9.36) and Worst–Avg (M = 75.17, SD = 11.45) phones similarly, implying that they relied prominently on information from the second session. In contrast, unconscious thinkers rated the Best–Avg phone (M = 78.00, SD = 9.70) as significantly better than the Worst–Avg phone (M = 64.50, SD = 9.39), F(1, 34) = 22.50, p < .001, indicating the evaluation was made based information presented in both sessions. However, as the Best–Avg phone possessed more positive attributes than the Worst–Avg phone in the first information session, one could argue that unconscious thinkers were more greatly influenced by information presented in the first session instead of integrating all of the presented information. Therefore, we conducted Experiment 2 to more fully examine this possibility. 1 To avoid the influence of some ultra attributes on evaluations, attributes of moderate importance were assessed prior to the study and chosen as research materials. During the assessment, 42 attributes of phone were evaluated by 40 participants, who were not included in the present experiment. They assessed the attributes on a seven-point scale (1 = not important, 7 = very important). According to the result, ultra attributes such as price and pattern, were eliminated. Thus 16 attributes with moderate importance were selected as research materials for the experiments (M = 4.28, SD = 0.66).
80
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
Fig. 1. Mean evaluations for the two phones under different thought conditions in Experiment 1. Errorbars represent standard errors of the mean.
3. Experiment 2 Experiment 2 aimed to eliminate the possibility that, when integrating temporally partitioned information, unconscious thought attached particular importance to information received earlier. To examine this issue, this experiment manipulated materials so that, overall, the four phones were equally preferable in terms of positive and negative attributes presented in the two sessions. Specifically, the experiment was designed so that one phone possessed the highest number of positive and lowest number of negative attributes in the first session, and accordingly, possessed the lowest number of positive and highest number of negative attributes in the second session. Likewise, another phone possessed the lowest number of positive and highest number of negative attributes in the first session, but had the highest number of positive and lowest number of negative attributes in the second session. Therefore, if unconscious thought integrated the partitioned information equally, an impartial evaluation would be made to indicate that the two phones were equally preferable. However, if unconscious thought attached particular importance to information received earlier, the phone with the highest number of positive and lowest number of negative attributes in the first session would be preferred. 3.1. Method 3.1.1. Participants and design Thirty-eight Chinese undergraduate students (20 women and 18 men) from Northwest Normal University were equally but randomly assigned into one of two conditions: an unconscious-thought condition (8 men and 11 women) vs. a conscious-thought condition (10 men and 9 women). A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 3.1.2. Materials and procedure Four hypothetical phones, labeled as Best–Worst, Worst–Best, Avg–Avg1, and Avg–Avg2, were used as research materials in Experiment 2 (as in Experiment 1, these labels were not accessible to participants). As in Experiment 1, the interval between sessions was 3 min. Other procedures were the same, with the following exceptions: (1) all four phones were equally preferable with 8 positive and 8 negative attributes; and (2) in the first information session, 6 positive and 2 negative attributes of Best–Worst were presented while 2 positive and 6 negative attributes of Worst–Best were presented. In the second session, the balance of positive and negative attributes for the two phones was reversed (2 positive and 6 negative attributes for Best–Worst and 6 positive and 2 negative attributes for Worst–Best). We hypothesized that participants would decide that Best–Worst was superior and Worst–Best was inferior if they made their evaluation under the influence of information presented in the first session. On the other hand, if their evaluations were made by integrating information from both sessions, they would conclude that Best–Worst and Worst–Best were equally preferable phones. 3.2. Results and discussion A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed ANOVA was conducted on participant scores. Results demonstrated no significant main effect of thought condition, F(1, 36) = .16, p > .05. There was a significant main effect of phone type, F(1, 36) = 11.34, p < .01, with the Worst–Best phone (M = 72.74, SD = 11.40) obtaining a higher score than the Best–Worst phone (M = 66.08, SD = 13.39). Furthermore, the inter-
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
81
Fig. 2. Mean evaluations for the two phones under different thought conditions in Experiment 2. Errorbars represent standard errors of the mean.
action between thought condition and phone type also was significant, F(1, 36) = 11.70, p < .001. As shown in Fig. 2, conscious thinkers assigned significantly higher scores to the Worst–Best phone (M = 76.79, SD = 11.63) than to the Best–Worst phone (M = 63.37, SD = 13.14), F(1, 36) = 23.04, p < .001, indicating that, once again conscious thought attached particular importance to the most recently received information. On the other hand, unconscious thinkers gave almost the same scores to the Best–Worst (M = 68.79, SD = 13.42) and Worst–Best (M = 68.68, SD = 9.85) phones, F(1, 36) = .01, p > .05. Although Best–Worst was preferable to Worst–Best based on the first information session, they possessed the same number of positive and negative attributes when information from the two sessions were integrated. Thus, the result of Experiment 2 ruled out the possibility that people in the unconscious thought group attached particular importance to information received early in their evaluation. 4. Experiments 3a and 3b As mentioned in the Introduction, compared with classical unconscious thought studies, the paradigm used in Experiments 1 and 2 not only presented information in a different temporal manner, but also elaborately arranged different distributions of positive/negative attributes in the first and second sessions. For example, the Best–Worst phone was better than all other alternatives in the first session, but appeared to be the worst choice based on information presented in the second session. Thus, the results of Experiments 1 and 2 could be attributed to differing distributions of positive/negative attributes in the first and second sessions. Therefore, Experiment 3 was designed to eliminate this confounding variable. Experiments 3a and 3b were controlled versions of Experiments 1 and 2, respectively. All information was presented in a single session in Experiments 3a and 3b, however, the presenting order of phone information was exactly the same as in Experiments 1 and 2. If the results of Experiments 1 and 2 were due to the different distributions of the positive/negative attributes in the first and second sessions, we would expect the results of Experiments 3a and 3b to be approximately the same as those of the first two experiments for the unconscious thought and conscious thought conditions. 4.1. Method 4.1.1. Participants and design Experiment 3a included 46 undergraduate students (19 women and 27 men) from Northwest Normal University who were equally but randomly assigned to one of two conditions: unconscious-thought (14 men and 9 women) or conscious-thought (13 men and 10 women). We used a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design. Experiment 3b included 38 undergraduate students (22 women and 16 men) from Northwest Normal University who were equally but randomly assigned to either the unconscious-thought (7 men and 12 women) or the conscious-thought (9 men and 10 women) conditions. A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 4.1.2. Materials and procedure The materials used in Experiments 3a and 3b were the same as those in Experiments 1 and 2. However, the experimental procedures differed from the first two experiments in the following ways: (1) there was no interval between the two information sessions and (2) both groups had 6 min to prepare for the final evaluation after all information was presented.
82
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
4.2. Results and discussion Experiment 3a used a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed ANOVA, which indicated no significant main effect of thought condition, F(1, 44) = .44, p > .05. However, we did find a significant main effect of phone type, F(1, 44) = 4.12, p < .05, with the Best–Avg phone (M = 71.35, SD = 13.59) receiving a significantly higher score than the Worst–Avg phone (M = 66.48, SD = 15.31). There also was a significant interaction between thought condition and phone type, F(1, 44) = 6.91, p < .05 (see Fig. 3a). The scores given to the Best–Avg phone (M = 73.35, SD = 15.39) by the unconscious thought group in Experiment 3a was significantly higher than the one given to the Worst–Avg phone (M = 62.17, SD = 15.87), F(1, 44) = 10.85, p < .001. Results for the unconscious thought group were approximately the same as those in Experiments 1. Hence, we can deduce that integration by unconscious thought is not impacted by whether or not information is partitioned, which further supported the Unconscious Thought Theory (Dijksterhuis & Nordgren, 2006). However, the scores given by the conscious thought group to the Best–Avg (M = 69.35, SD = 11.51) and Worst–Avg (M = 70.78, SD = 13.74) phones were no-differed in Experiment 3a, F(1, 44) = 0.18, p > .05. Although it appears that the results were the same as those of Experiment 1 for the conscious thought condition, our interpretations of the reasons for these results are different. In Experiment 1, the conscious thought group relied on information from the second session and evaluated the Best–Avg and Worst–Avg phones as equally preferable. Due to the elimination of the interval between the two sessions in Experiment 3a, the positive attributes of the Best–Avg and Worst–Avg phones were no longer as salient as those in Experiment 1. In addition, the amount of information in the temporally continuing session was doubled. We conclude that combining these two factors might skew the overall evaluation made by conscious thought, making it more difficult for participants to discriminate between a bad phone and a better one and resulting in the same scores being given to Best–Avg and Worst–Avg phones. Results of subsequent Experiment 3b also supported this assumption.
Fig. 3a. Mean evaluations for the two phones under different thought conditions in Experiment 3a. Errorbars represent standard errors of the mean.
Fig. 3b. Mean evaluations for the two phones under different thought conditions in Experiment 3b. Errorbars represent standard errors of the mean.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
83
In Experiment 3b, a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed ANOVA was conducted on the scores from Experiment 3a and revealed no significant main effects or interactions, all Fs < 1 (see Fig. 3b). The scores given to Best–Worst (M = 69.16, SD = 15.41) and Worst–Best (M = 68.89, SD = 16.75) phones by people in the unconscious thought group in Experiment 3b approximately were the same. The results for the unconscious thought group were approximately identical to those in Experiments 2. Likewise, as the scores given to the Best–Worst (M = 70.58, SD = 13.88) and Worst–Best (M = 70.05, SD = 13.37) phones by the conscious thought group were not significantly different in Experiment 3b. The results for the conscious thought group were different from those of Experiment 2 as well. In the second half of session, the Worst–Best phone was better than the Best–Worst one, but the conscious group evaluated them as equally preferable. These results further support our hypothesis that the complexity of information influenced participants’ evaluations in the conscious thought condition in Experiment 3a. Based on the above analyses, we could conclude that it was the partitioning of the information, rather than the distribution of the positive/negative attributes in the first and second sessions, that was responsible for the results in Experiments 1 and 2. Our finding also suggests that partitioning information impacts information integration made by conscious, but not unconscious, thought.
5. Experiments 4a and 4b Previous research showed that participants can make on-line judgments of four phones during the two periods of information acquisition (Hastie & Park, 1986; Lassiter et al., 2009; Waroquier, Marchiori, Klein, & Cleeremans, 2010). Therefore, the results of partitioned information integrated by participants in the unconscious thought condition in Experiments 1 and 2 merely may be due to them recalling a judgment that was formed on-line. We designed Experiment 4 to further assess, and hopefully exclude, this possibility. Experiments 4a and 4b were, respectively, controlled experiments of Experiments 1 and 2. In Experiments 4a and 4b, all participants had a period of delay (mere distraction) between sessions 1 and 2. After session 2, participants were divided into three groups: conscious thought, unconscious thought, and immediate decision. Previous research showed that unconscious thought was goal-dependent (Bos, Dijksterhuis, & van Baaren, 2008). In the study by Bos et al. (2008), participants were given information about some cars and divided into two groups. One group was required to think consciously about their impression of the cars for several minutes. The other group was asked to perform a distraction task within that same time. However, within the distraction group, some participants were told they would be asked their opinion of the cars after the distraction task (goal-dependent), while others were not given this information (mere distraction). Results showed that only the goal-dependent participants in the unconscious thought condition performed better than did those in the conscious thought condition. Thus, if the explanation of the on-line judgment account is correct, the immediate decision and unconscious thought groups would equally integrate information from both sessions to reach a final judgment. 5.1. Method 5.1.1. Participants and design Experiment 4a included 60 undergraduate students (34 women and 26 men) from Northwest Normal University who were equally but randomly assigned into one of three conditions: unconscious-thought (7 men and 13 women), conscious-thought (8 men and 12 women), or immediate decision (11 men and 9 women). A 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design was applied. Experiment 4b included 63 undergraduate students (33 women and 30 men) from Northwest Normal University who were equally but randomly assigned to one of three conditions: unconscious-thought (9 men and 12 women), consciousthought (10 men and 11 women), and immediate decision (11 men and 10 women). A 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 5.1.2. Materials and procedure The materials utilized in Experiments 4a and 4b were exactly the same as those in Experiments 1 and 2, respectively. As in Experiments 1 and 2, the between-session interval was 3 min. The other procedures were identical to those in Experiments 1 and 2, with the following exceptions: (1) participants were not informed that the information would be divided into two sessions, (2) after the first presentation of information, all participants were required to complete a 2-back task for 3 min but were not told that they would have to perform another task after the distraction, and (3) an immediate decision group was added. After the second presentation of information, participants in the conscious thought group were asked to think about phones for 3 min, those in the unconscious thought group were required to complete a 2-back task for 3 min, but were informed in advance that they would have to give their opinions about the phones. The immediate decision group was asked to immediately evaluate the four phones.
84
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
5.2. Results and discussion Experiments 4a and 4b both applied a 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Avg vs. Worst–Avg) mixed-design ANOVA. Results of Experiment 4a did not show a significant main effect of thought condition, F(2, 57) = .01, p > .05; however, there was a significant main effect of phone type, F(1, 57) = 4.36, p < .05. Specifically, scores of the Best–Avg phone (M = 70.71, SD = 13.01) were significantly higher than those for the Worst–Avg phone (M = 67.10, SD = 12.67). The interaction between thought condition and phone type also was significant, F(2, 57) = 7.58, p < .01 (see Fig. 4a). If the on-line judgment account was credible, the scores for Best–Avg and Worst–Avg phones given by the immediate decision and unconscious groups should be the same. However, in Experiment 4a, the immediate decision group gave approximately the same scores to the Best–Avg (M = 69.35, SD = 12.36) and Worst–Avg (M = 68.30, SD = 11.06) phones, F(1, 57) = 0.12, p > .05, while the scores given by the unconscious thought group to the Best–Avg phone (M = 75.60, SD = 12.89) were significantly higher than for the Worst–Avg phone (M = 62.75, SD = 12.30), F(1, 57) = 18.36, p < .001. These results suggest that the immediate decision group depended on information presented in the second session when making their evaluation. Therefore, results do not support the on-line judgment account. Furthermore, these results replicated our findings from Experiment 1 regarding unconscious thought. The scores given to the Best–Avg (M = 67.20, SD = 12.92) and Worst–Avg (M = 70.25, SD = 13.91) phones by the conscious thought group were appropriately the same, F(1, 57) = 1.03, p > .05. These results also replicated those regarding the role of conscious thought from Experiment 1. Results of Experiment 4b showed that the main effect of thought condition was not significant, F(2, 60) = .75, p > .05; however the main effect of phone type was significant, F(1, 60) = 4.36, p < .05. Specifically, scores for the Best–Avg phone (M = 67.17, SD = 12.91) were lower than those for the Worst–Avg phone (M = 72.95, SD = 12.29). The interaction between thought condition and phone type also was significant, F(2, 60) = 3.95, p < .05 (see Fig. 4b). In Experiment 4b, scores given by the immediate decision group to the Worst–Best phone (M = 73.43, SD = 13.05) were significantly higher than for the Best–Worst phone (M = 62.67, SD = 12.76), F(1, 60) = 9.30, p < .01, while scores given to Worst–Best (M = 70.29, SD = 10.11) and Best–worst (M = 72.52, SD = 12.08) phones by the unconscious thought group were approximately the same, F(1, 60) = .40, p > .05. These results were the same as those of Experiment 4a, namely, the evaluations of two groups were different. The score for the Worst–Best phone (M = 75.14, SD = 13.52) was significantly higher than for the Best–Worst phone (M = 66.33, SD = 12.46), F(1, 60) = 6.23, p < .05. Similarly, the results replicated those of Experiment 2 in conscious thought condition. Combining the results of Experiments 4a and 4b revealed that the on-line judgment account did not influence the integration of information made by the unconscious group, but possibly impacted the conscious group. Overall, Experiment 4 demonstrated that unconscious thought mainly relies on off-line judgment (Strick, Dijksterhuis, & van Baaren, 2010).
Fig. 4a. Mean evaluations for the two phones under different thought conditions in Experiment 4a. Errorbars represent standard errors of the mean.
Fig. 4b. Mean evaluations for the two phones under different thought conditions in Experiment 4b. Errorbars represent standard errors of the mean.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
85
6. General discussion The present study demonstrated that when information was presented in two partitioned sessions, conscious thinkers tended to emphasize information in the second session (i.e., the non-different evaluation in Experiment 1 and the significantly different evaluation in Experiment 2), while unconscious thinkers emphasized neither the information received in the second session (Experiment 1), nor the information received in the first session (Experiment 2). In Experiment 3, all information was presented to participants in one temporally continuous session, but the order in which information was presented was kept the same as when information was partitioned (Experiments 1 & 2), excluding the possibility that the results from the first two studies were due to differences in the distributions of the positive/negative attributes in the first and second sessions. In Experiment 4, an immediate decision group was added, which further eliminated the on-line judgment account. These results support the hypothesis that unconscious thought can equally integrate temporally scattered information, while conscious thought tends to attach particular importance to the most recently received information. The current study explored the mechanism of unconscious thought in integrating temporally scattered complex information. The temporal factor has been believed to be essential for complex consumer decision-making (Peters & Büchel, 2011; Slovic, 2012). Previous research has suggested that unconscious thought leads to an unbiased evaluation by integrating relevant information. The current study examined this hypothesis by presenting complex information in a partitioned way and revealed that unconscious thought integrated the separated information without bias as well. Our results provide further support for the Unconscious Thought Theory and did so with a higher ecological adaptability. The present study provides a new perspective on long-term decision-making. In our daily lives, we frequently are involved in decision-making that lasts for a prolonged period. Under these circumstances, related information is not obtained in just one session, but rather, is presented in several successive sessions. Conscious thought mainly depends on the most recently received information, which can affect decision-making by leading to a biased evaluation. On the other hand, as supported by results of the current study, unconscious thought equally integrates information received in different sessions and leads to an impartial evaluation. Future research should examine the generalizability of these results for making decisions in authentic situations, particularly those that involve high risk. We have argued here that the greater processing capacity of unconscious thought may explain the absence of cognitive bias in integrating temporally partitioned information. With a greater capacity (see also Dijksterhuis, 2004; Miller, 1956; Wilson, 2002), unconscious thought is able to integrate information from a wider range of time periods during the evaluation, in contrast with conscious thought, which particularly relies on recently received information (i.e., the recency effect; Anderson, 1981; Sasaki, 2009) due to its limited processing capacity. When information needed for decision-making is displayed in a temporally separated manner (as in the current study), the interval between information presented first and last is significantly larger, resulting in an unavoidable recency effect. In other words, the overall evaluation is entirely dependent upon information presented in the second session in conscious thought due to the inability to hold information from the earlier session. However, in unconscious thought, information from both the earlier and later sessions can be retrieved for the decision-making process, eventually leading to an unbiased evaluation. Our results contradict those of previous studies on unconscious thought and decision-making (González-Vallejo et al., 2013; Newell, Wong, Cheung, & Rakow, 2009). In the Newell et al. study, participants were presented with attributes of two cars. One car’s attributes were ordered in an ascending pattern, while the other’s attributes were ordered in a descending pattern. The results showed that the car with the ascending pattern was rated as the top car by the unconscious thought group, providing evidence for a recency effect. Conversely, González-Vallejo et al. found a primacy effect. In this study, González-Vallejo and colleagues presented participants with the attributes of four cars in two ways. For the car with the greatest number of positive attributes, they presented the positive attributes first. In the other group, information on the positive attributes of the car with the least number of positive attributes was presented first. The results demonstrated that the unconscious group preferred the car with the greatest number of positive attributes only in the condition where positive attributes were presented at the beginning of the sequence. We believe task complexity is the main reason for the differences between our results and the results of these previous studies. According to Unconscious Thought Theory, unconscious thought is preferable for processing complex tasks (Dijksterhuis et al., 2006). Using meta-analysis, Strick, Dijksterhuis, Bos, van Sjoerdsma, and van Baaren (2011) found that the unconscious thought effect was larger when a problem was sufficiently complex. Compared with Newell et al. and González-Vallejo et al., the tasks in our study were more complex in several ways. First, we presented a larger amount of information. There were 4 options and 64 messages in the present study, while Newell et al.’s study merely had 2 options and 20 messages. While there were 4 options in the study by González-Vallejo et al., there were only a total of 48 messages. Secondly, the present study had longer periods between presentations of information. In contrast to the previously mentioned studies, our study inserted time intervals between information presented first and information last, which accounts for the longer presentation time. Future research should explore how unconscious thought equally integrates information temporally scattered across two sessions. We propose two possible approaches to evaluating this in more detail: first, unconscious thought evaluates each mobile phone by the addition of all positive attributes and that of negative ones across two sessions. Another possible approach is to evaluate the relative merits of each mobile phone based on its positive and negative attributes across two sessions. Subsequently, unconscious thought makes an overall judgment according to the integration of the two evaluations. For instance, in the present study, unconscious thought evaluates that the Best–Worst phone is superior to the Worst–Best one
86
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
in the first session, while the former is worse than the latter in the second session. Then, the results of these two sessions are integrated, giving participants an overall evaluation of mobile phones. Based on what we have mentioned above, the first approach asks people in the unconscious thought condition to follow strict rules. However, Unconscious Thought Theory states that unconscious thought cannot follow strict rules; it gives only rough estimates. Studies also showed that using unconscious thought to integrate information was a gist-based process (Abadie, Waroquier, & Terrier, 2013; Dijksterhuis, 2004; Strick et al., 2011). Specifically, Abadie et al. found that unconscious thought was dependent more on gist memory to process information than on verbatim memory. Therefore, we are inclined to support the second approach. However, this approach requires further confirmation through future experiments. In conclusion, the present study confirmed that recency effect is involved when conscious thought integrates the temporally scattered information, whereas unconscious thought is free from the influence of when information is presented and, thus, can equally integrate information scattered throughout partitioned sessions. Acknowledgments This research is supported by the National Natural Science Foundation of China (Nos. 31200786, 31170974), Key Project of Humanities and Social Sciences, Ministry of Education (No. 07JZD0029), and the Fundamental Research Funds for the Central Universities, Foundation for Young Teacher of Northwest Normal University (NWNU-LKQ N-11-34). References Abadie, M., Waroquier, L., & Terrier, P. (2013). Gist memory in the unconscious-thought effect. Psychological Science, 24, 1253–1259. Anderson, N. H. (1981). Foundations of information integration theory. New York, NY: Academic Press. Ariely, D., & Zauberman, G. (2000). On the making of an experience: The effects of breaking and combining experiences on their overall evaluation. Journal of Behavioral Decision Making, 13, 219–232. Ariely, D., & Zauberman, G. (2003). Differential partitioning of extended experiences. Organizational Behavior and Human Decision Processes, 91, 128–139. Bos, M. W., Dijksterhuis, A., & van Baaren, R. B. (2011). The benefits of ‘‘sleeping on things’’: Unconscious thought leads to automatic weighting. Journal of Consumer Psychology, 21, 4–8. Bos, M. W., Dijksterhuis, A., & van Baaren, R. B. (2008). On the goal dependency of unconscious thought. Journal of Experimental Social Psychology, 44, 1114–1120. Bos, M. W., Dijksterhuis, A., & Van Baaren, R. B. (2012). Food for thought? Trust your unconscious when energy is low. Journal of Neuroscience, Psychology, and Economics, 5, 124–130. Costabile, K. A., & Klein, S. B. (2005). Finishing strong: Recency effects in juror judgments. Basic and Applied Social Psychology, 27, 47–58. Chapman, G. B., Bergus, G. R., & Elstein, A. S. (1996). Order of information affects clinical judgment. Journal of Behavioral Decision Making, 9, 201–211. Crano, W. D. (1977). Primacy versus recency in retention of information and opinion change. The Journal of Social Psychology, 101, 87–96. Creswell, J. D., Bursley, J. K., & Satpute, A. B. (in press). Reactivation links unconscious thought to decision making performance. Social Cognitive and Affective Neuroscience.. Dijksterhuis, A. (2004). Think different: The merits of unconscious thought in preference development and decision making. Journal of Personality and Social Psychology, 87, 586–598. Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & Van Baaren, R. B. (2006). On making the right choice: The deliberation-without-attention effect. Science, 311, 1005–1007. Dijksterhuis, A., Bos, M. W., Van der Leij, A., & Van Baaren, R. B. (2009). Predicting soccer matches after unconscious and conscious thought as a function of expertise. Psychological Science, 20, 1381–1387. Dijksterhuis, A., & Nordgren, L. F. (2006). A theory of unconscious thought. Perspectives on Psychological Science, 1, 95–109. González-Vallejo, C., Cheng, J., Phillips, N., Chimeli, J., Bellezza, F., Harman, J., et al (2013). Early positive information impacts final evaluations: No deliberation-without-attention effect and a test of a dynamic judgment model. Journal of Behavioral Decision Making. http://dx.doi.org/10.1002/ bdm.1796. González-Vallejo, C., & Phillips, N. (2010). Predicting soccer matches: A reassessment of the benefit of unconscious thinking. Judgment and Decision Making, 5, 200–206. Ham, J., & Van den Bos, K. (2010). On unconscious morality: The effects of unconscious thinking on moral decision making. Social Cognition, 28, 74–83. Ham, J., Van den Bos, K., & Van Doorn, E. (2009). Lady justice thinks unconsciously: Unconscious thought can lead to more accurate justice judgments. Social Cognition, 27, 509–521. Hasford, J. (2014). Should I think carefully or sleep on it?: Investigating the moderating role of attribute learning. Journal of Experimental Social Psychology, 51, 51–55. Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the judgment is memory-based or on-line. Psychological Review, 93, 258–268. Lassiter, G. D., Lindberg, M. J., González-Vallejo, C., Bellezza, F. S., & Phillips, N. D. (2009). The deliberation-without-attention effect: Evidence for an artifactual interpretation. Psychological Science, 20, 671–675. Lichtenstein, M., & Srull, T. K. (1987). Processing objectives as a determinant of the relationship between recall and judgment. Journal of Experimental Social Psychology, 23, 93–118. Jonides, J., Schumacher, E. H., Smith, E. E., Lauber, E. J., Awh, E., Minoshima, S., et al (1997). Verbal working memory load affects regional brain activation as measured by PET. Journal of Cognitive Neuroscience, 9, 462–475. Messner, C., Wänke, M., & Weibel, C. (2011). Unconscious personnel selection. Social Cognition, 29, 699–710. Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81–97. Montgomery, N. V., & Unnava, H. R. (2009). Temporal sequence effects: A memory framework. Journal of Consumer Research, 36, 83–92. Nelson, L., & Meyvis, T. (2008). Interrupted consumption: Disrupting adaptation to hedonic experiences. Journal of Marketing Research, 45, 654–664. Newell, B. R., Wong, K. Y., Cheung, J. C. H., & Rakow, T. (2009). Think, blink or sleep on it? The impact of modes of thought on complex decision making. Quarterly Journal of Experimental Psychology, 62, 707–732. Peters, J., & Büchel, C. (2011). The neural mechanisms of inter-temporal decision-making: Understanding variability. Trends in Cognitive Sciences, 15, 227–239. Ric, F., & Muller, D. (2012). Unconscious addition: When we unconsciously initiate and follow arithmetic rules. Journal of Experimental Psychology: General, 141, 222–226. Sasaki, T. (2009). Individual differences in working memory capacity in recency effects: From the recall process. Psychological Reports, 104, 545–548. Slovic, P. (2012). Psychological study of human judgment: Implications for investment decision making. The Journal of Finance, 27, 779–799.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
87
Strick, M., Dijksterhuis, A., & van Baaren, R. B. (2010). Unconscious thought effects take place off-line, not on-line. Psychological Science, 21, 484–488. Strick, M., Dijksterhuis, A., Bos, M. W., van Sjoerdsma, A., & van Baaren, R. B. (2011). A meta-analysis on unconscious thought effects. Social Cognition, 29, 738–762. Usher, M., Russo, Z., Weyers, M., Brauner, R., & Zakay, D. (2011). The impact of the mode of thought in complex decisions: Intuitive decisions are better. Frontiers in Psychology, 2, 37. http://dx.doi.org/10.3389/fpsyg.2011.00037. Waroquier, L., Marchiori, D., Klein, O., & Cleeremans, A. (2010). Is it better to think unconsciously or to trust your first impression? A reassessment of unconscious thought theory. Social Psychological and Personality Science, 1, 111–118. Wilson, T. D. (2002). Strangers to ourselves: Discovering the adaptive unconscious. Cambridge, MA: Harvard University Press. Yang, H., Chattopadhyay, A., Zhang, K., & Dahl, D. W. (2012). Unconscious creativity: When can unconscious thought outperform conscious thought? Journal of Consumer Psychology, 22, 573–581. Zhong, C. B., Dijksterhuis, A., & Galinsky, A. D. (2008). The merits of unconscious thought in creativity. Psychological Science, 19, 912–918.
Contents lists available at ScienceDirect
Consciousness and Cognition journal homepage: www.elsevier.com/locate/concog
Bias or equality? Unconscious thought equally integrates temporally scattered information Jiansheng Li a,b, Qiyang Gao a, Jifan Zhou a, Xinyu Li a, Meng Zhang a, Mowei Shen a,⇑ a b
Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou 310028, China Department of Psychology, Northwest Normal University, Lanzhou 730070, China
a r t i c l e
i n f o
Article history: Received 1 May 2013 Available online 25 February 2014 Keywords: Unconscious thought Decision making Integration information
a b s t r a c t In previous experiments on unconscious thought, information was presented to participants in one continuous session; however, in daily life, information is delivered in a temporally partitioned way. We examined whether unconscious thought could equally integrate temporally scattered information when making overall evaluations. When presenting participants with information in two temporally partitioned sessions, participants’ overall evaluation was based on neither the information in the first session (Experiment 1) nor that in the second session (Experiment 2); instead, information in both sessions were equally integrated to reach a final judgment. Conscious thought, however, overemphasized information in the second session. Experiments 3 and 4 further ruled out possible influencing factors including differences in the distributions of positive/negative attributes in the first and second sessions and on-line judgment. These findings suggested that unconscious thought can integrate information from a wider range of periods during an evaluation, while conscious thought cannot. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction According to traditional wisdom, when confronted with a complex decision, people are accustomed to solving problems with deliberation. Nevertheless, a growing number of studies have suggested the opposite, that applying unconscious thought leads to a better decision than conscious thought (e.g., Bos, Dijksterhuis, & van Baaren, 2011, 2012; Dijksterhuis, 2004; Dijksterhuis, Bos, Nordgren, & Van Baaren, 2006; Ham, Van den Bos, & Van Doorn, 2009; Hasford, 2014). Unconscious thought is defined as ‘‘object-relevant or task-relevant cognitive or affective thought processes that occur while conscious attention is directed elsewhere’’ (Dijksterhuis & Nordgren, 2006, p. 96). A typical example was demonstrated by Dijksterhuis et al. (2006), who required participants to choose the best car from four alternatives that were each assigned positive or negative attributes. Compared with participants who made decisions after a period of conscious thinking, those who did not have the opportunity to deliberate (i.e., distracted by a 2-back task; Jonides et al., 1997) actually made relatively advantageous decisions. The researchers reasoned that unconscious thought has a higher processing capacity, allowing people to simultaneously process more information and make better decisions, while conscious thought focuses on limited dimensions (see also Dijksterhuis & Nordgren, 2006) and is less effective for dealing with complex problems.
⇑ Corresponding author. Address: Department of Psychology and Behavioral Sciences, Xixi Campus, Zhejiang University, Hangzhou 310028, China. E-mail address: [email protected] (M. Shen). http://dx.doi.org/10.1016/j.concog.2014.01.012 1053-8100/Ó 2014 Elsevier Inc. All rights reserved.
78
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
The effects of unconscious thought have been observed across many contexts, including solving creative problems (Yang, Chattopadhyay, Zhang, & Dahl, 2012; Zhong, Dijksterhuis, & Galinsky, 2008), predicting soccer matches (Dijksterhuis, Bos, Van der Leij, & Van Baaren, 2009; González-Vallejo & Phillips, 2010), making moral judgments (Ham & Van den Bos, 2010), selecting job applicants (Messner, Wänke, & Weibel, 2011), and doing simple arithmetic (Ric & Muller, 2012). Creswell and colleagues provided evidence from cognitive-neuroscience that brain regions activated when encoding information for decision making continue to be active while the brain responds to other, irrelevant, tasks (Creswell, Bursley, & Satpute, in press). In previous studies on unconscious thought, information for decision-making was presented to participants in one continuous session. However, in daily life, information frequently is delivered in a temporally partitioned way. For instance, while making a decision to purchase a mobile phone, a consumer may first ask for opinions from close friends and/or consult online assessments and then visit stores to check various models before deciding which phone to buy. Thus, when people are making a decision, time elapses between periods of information gathering. This raises the question of how unconscious thought integrates information for making an evaluation when information is presented in a temporally partitioned manner. Studies on conscious thought have demonstrated an emphasis on the most recently presented information when making a decision, otherwise known as the recency effect (e.g., Crano, 1977; Lichtenstein & Srull, 1987). For instance, evidence presented near the end of a prosecution’s defense was found to have the greatest impact on juror verdicts (Costabile & Klein, 2005). The same have been found for clinical judgments; the most recently received information has a relatively more powerful influence upon diagnoses given by junior, and even more experienced, physicians (Chapman, Bergus, & Elstein, 1996). Montgomery and Unnava (2009) explained that information presented more recently is much more easily retrieved from memory, and hence, will be regarded as the main source of information for making an overall evaluation, during which the recency effect inevitably occurs. More interestingly, previous research on conscious thought has demonstrated that whether or not information about an object is presented in a temporally partitioned way has distinct influence on people’s evaluation of the object. For instance, Ariely and Zauberman (2000, 2003) showed that people preferred sequentially presented information that was improving in intensity to that of declining intensity. However, people’s preferences for increasing sequences would be less favorable if information was partitioned. In addition, another experiment indicated that pleasant experiences would be more enjoyable and that unpleasant experiences would be more irritating, if pleasant and unpleasant experiences were partitioned (Nelson & Meyvis, 2008). Given the above findings regarding conscious thought, it is intriguing to ask whether information presented in a temporally partitioned manner also would profoundly influence unconscious evaluations. In other words, will unconscious thought place particular emphasis on information presented on a given temporal session (e.g., the most recent session) when people make evaluations? Alternatively, will unconscious thought integrate all partitioned information equally regardless of the order of presentation? Usher, Russo, Weyers, Brauner, and Zakay (2011, Experiment 3) designed an experiment asking people to choose from three hypothetical individuals as a flat mate. Each individual was characterized by 12 attributes. The percentage of positive attributes for each of the potential flat mates was 66%, 50%, and 33%. Information regarding these attributes was presented to participants in three sessions, with four attributes presented within each one. In addition, each mobile phones’ attributes were shown in a different color to help participants distinguish them from each other. Before the presentation, conscious thinkers were informed that ‘‘research has shown that the best decisions are the ones made using logic and rational thought,’’ and were asked to deliberate for 1 min after each session. Unconscious thinkers, on the other hand, were told that ‘‘research has shown that the best decisions are the ones made using intuition,’’ and were required to complete a 1-min interfering task after each session. At the end of the experiment, participants were asked to score each potential flat mate and choose the one that they most preferred. Results showed that accuracy and the discrimination rates of the unconscious thought group were more reasonable than those of the conscious thought group (Usher et al., 2011). However, this study had several limitations: First, experimental instructions differed between the two groups, which may lead to confounding influence regarding the final conclusion. It was firmly asserted that experimental instructions had a profound impact on the unconscious thought effect (Lassiter, Lindberg, González-Vallejo, Bellezza, & Phillips, 2009). Second, without a valid controlled experiment in which all information was presented in a single session, it is possible that the results were due to the different font colors rather than the partitioning of information. Third, researchers could not exclude the on-line judgment account, which might be formed due to the limited amount of information presented in each session (12 pieces per session). Finally, although the three flat mates had obvious general differences, there was no strict control of the relative merits of an individual flat mate (i.e., there was no objective data that A was better than B) in each session. This made the attributes in each session ambiguous, which may weaken each session affecting the overall evaluation. Based on these limitations, the present study intended to further examine the question of whether unconscious thought can integrate partitioned information without bias. We used four hypothetical mobile phones as experimental materials. Information about the phones was successively presented in two temporally partitioned sessions. We manipulated the information presented in the two sessions so that participants’ final evaluations would differ depending on whether information about the phones was integrated equally across two sessions or information from either session was particularly emphasized. 2. Experiment 1 Experiment 1 was designed to make an overall evaluation of unconscious thought by directly investigating whether it could equally integrate all temporally scattered information, or would attach particular importance to one certain piece
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
79
of information. In this experiment, information (attributes) of four hypothetical mobile phones was displayed in two separate sessions, although our study was only concerned with two of the phones. In the first session, one phone was clearly better than the other was because it possessed the highest number of positive attributes and the lowest number of negative attributes, while the other possessed the lowest number of positive and highest number of negative attributes. However, in the second session, the two phones were the same, possessing the same number of positive and negative attributes. Therefore, overall, one phone was clearly better than the other was. Our hypothesis stated that if unconscious thought equally integrated information from both sessions, participants would be able to distinguish between the superior and inferior phones. 2.1. Method 2.1.1. Participants and design Thirty-six Chinese undergraduate students (16 women and 20 men) from Northwest Normal University were equally but randomly assigned into either the unconscious-thought condition (7 men and 11 women) or the conscious-thought condition (9 men and 9 women). A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design was employed. In this and each subsequent experiment, the thought condition served as a between-subject variable and phone type was the within-subject variable. The scores participants assigned to the phones at the end of the experiment served as dependent variables. Participants all received financial compensation for their participation. 2.1.2. Materials and procedure Four hypothetical phones, respectively labeled as Best–Avg, Worst–Avg, Avg–Avg1 and Avg–Avg2, were used as research materials (labels were not accessible to participants). Each phone was assigned 16 attributes.1 The phone labeled as Best–Avg possessed 10 positive and 6 negative attributes, the Worst–Avg had 6 positive and 10 negative attributes, and both Avg–Avg1 and Avg–Avg2 had 8 positive and 8 negative attributes. In other words, Best–Avg was considered the superior choice, Worst– Avg was the inferior choice, and Avg–Avg1 and Avg–Avg2 were considered to be the same. Phone attributes were presented sequentially in two temporally partitioned sessions. In each session, 8 attributes of each phone were presented, with a total of 32 attributes presented. The order of presentation was counterbalanced across participants. In the first session, 6 positive and 2 negative attributes of Best–Avg were presented along with 2 positive and 6 negative attributes of Worst–Avg. In the second session, the remaining 6 positive and 6 negative attributes of either phone were presented. In other words, although the Best–Avg phone was clearly superior to Worst–Avg, overall, the second session of information presentation by itself would suggest Best–Avg and Worst–Avg were equally preferable. Four positive and 4 negative attributes of Avg–Avg1 and Avg–Avg2 also were presented in each session. In preparation for the task, participants were told they would read information about four phones on the screen and the presentation of information was divided into two sessions, one after another. Sentences describing the attributes of the four phones were randomly presented for 5 s each. At the end of each session, the conscious thinkers were asked to think about their impression of the phones for 3 min. Meanwhile, unconscious thinkers were required to complete a 2-back task, which also lasted for 3 min. At the end of the experiment, all participants indicated their attitude toward each phone on a scale of 0–100, with a higher score indicating a greater preference for the phone. 2.2. Results and discussion In the current experiment, Avg–Avg1 and Avg–Avg2 performed the function of adding complexity to the decision-making, therefore, only the scores of Best–Avg and Worst–Avg were subjected to data analysis. A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed-design ANOVA was conducted on participant scores. Results did not show a significant main effect of thought condition, F(1, 34) = .99, p > .05; however, the main effect of phone type was significant, F(1, 34) = 7.25, p < .05. Scores for the Best–Avg phone (M = 75.25, SD = 9.80) were significantly higher than for the Worst–Avg phone (M = 69.83, SD = 11.37). The interaction of thought condition and phone type also was significant, F(1, 34) = 16.14, p < .001. As shown in Fig. 1, conscious thinkers rated the Best–Avg (M = 72.50, SD = 9.36) and Worst–Avg (M = 75.17, SD = 11.45) phones similarly, implying that they relied prominently on information from the second session. In contrast, unconscious thinkers rated the Best–Avg phone (M = 78.00, SD = 9.70) as significantly better than the Worst–Avg phone (M = 64.50, SD = 9.39), F(1, 34) = 22.50, p < .001, indicating the evaluation was made based information presented in both sessions. However, as the Best–Avg phone possessed more positive attributes than the Worst–Avg phone in the first information session, one could argue that unconscious thinkers were more greatly influenced by information presented in the first session instead of integrating all of the presented information. Therefore, we conducted Experiment 2 to more fully examine this possibility. 1 To avoid the influence of some ultra attributes on evaluations, attributes of moderate importance were assessed prior to the study and chosen as research materials. During the assessment, 42 attributes of phone were evaluated by 40 participants, who were not included in the present experiment. They assessed the attributes on a seven-point scale (1 = not important, 7 = very important). According to the result, ultra attributes such as price and pattern, were eliminated. Thus 16 attributes with moderate importance were selected as research materials for the experiments (M = 4.28, SD = 0.66).
80
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
Fig. 1. Mean evaluations for the two phones under different thought conditions in Experiment 1. Errorbars represent standard errors of the mean.
3. Experiment 2 Experiment 2 aimed to eliminate the possibility that, when integrating temporally partitioned information, unconscious thought attached particular importance to information received earlier. To examine this issue, this experiment manipulated materials so that, overall, the four phones were equally preferable in terms of positive and negative attributes presented in the two sessions. Specifically, the experiment was designed so that one phone possessed the highest number of positive and lowest number of negative attributes in the first session, and accordingly, possessed the lowest number of positive and highest number of negative attributes in the second session. Likewise, another phone possessed the lowest number of positive and highest number of negative attributes in the first session, but had the highest number of positive and lowest number of negative attributes in the second session. Therefore, if unconscious thought integrated the partitioned information equally, an impartial evaluation would be made to indicate that the two phones were equally preferable. However, if unconscious thought attached particular importance to information received earlier, the phone with the highest number of positive and lowest number of negative attributes in the first session would be preferred. 3.1. Method 3.1.1. Participants and design Thirty-eight Chinese undergraduate students (20 women and 18 men) from Northwest Normal University were equally but randomly assigned into one of two conditions: an unconscious-thought condition (8 men and 11 women) vs. a conscious-thought condition (10 men and 9 women). A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 3.1.2. Materials and procedure Four hypothetical phones, labeled as Best–Worst, Worst–Best, Avg–Avg1, and Avg–Avg2, were used as research materials in Experiment 2 (as in Experiment 1, these labels were not accessible to participants). As in Experiment 1, the interval between sessions was 3 min. Other procedures were the same, with the following exceptions: (1) all four phones were equally preferable with 8 positive and 8 negative attributes; and (2) in the first information session, 6 positive and 2 negative attributes of Best–Worst were presented while 2 positive and 6 negative attributes of Worst–Best were presented. In the second session, the balance of positive and negative attributes for the two phones was reversed (2 positive and 6 negative attributes for Best–Worst and 6 positive and 2 negative attributes for Worst–Best). We hypothesized that participants would decide that Best–Worst was superior and Worst–Best was inferior if they made their evaluation under the influence of information presented in the first session. On the other hand, if their evaluations were made by integrating information from both sessions, they would conclude that Best–Worst and Worst–Best were equally preferable phones. 3.2. Results and discussion A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed ANOVA was conducted on participant scores. Results demonstrated no significant main effect of thought condition, F(1, 36) = .16, p > .05. There was a significant main effect of phone type, F(1, 36) = 11.34, p < .01, with the Worst–Best phone (M = 72.74, SD = 11.40) obtaining a higher score than the Best–Worst phone (M = 66.08, SD = 13.39). Furthermore, the inter-
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
81
Fig. 2. Mean evaluations for the two phones under different thought conditions in Experiment 2. Errorbars represent standard errors of the mean.
action between thought condition and phone type also was significant, F(1, 36) = 11.70, p < .001. As shown in Fig. 2, conscious thinkers assigned significantly higher scores to the Worst–Best phone (M = 76.79, SD = 11.63) than to the Best–Worst phone (M = 63.37, SD = 13.14), F(1, 36) = 23.04, p < .001, indicating that, once again conscious thought attached particular importance to the most recently received information. On the other hand, unconscious thinkers gave almost the same scores to the Best–Worst (M = 68.79, SD = 13.42) and Worst–Best (M = 68.68, SD = 9.85) phones, F(1, 36) = .01, p > .05. Although Best–Worst was preferable to Worst–Best based on the first information session, they possessed the same number of positive and negative attributes when information from the two sessions were integrated. Thus, the result of Experiment 2 ruled out the possibility that people in the unconscious thought group attached particular importance to information received early in their evaluation. 4. Experiments 3a and 3b As mentioned in the Introduction, compared with classical unconscious thought studies, the paradigm used in Experiments 1 and 2 not only presented information in a different temporal manner, but also elaborately arranged different distributions of positive/negative attributes in the first and second sessions. For example, the Best–Worst phone was better than all other alternatives in the first session, but appeared to be the worst choice based on information presented in the second session. Thus, the results of Experiments 1 and 2 could be attributed to differing distributions of positive/negative attributes in the first and second sessions. Therefore, Experiment 3 was designed to eliminate this confounding variable. Experiments 3a and 3b were controlled versions of Experiments 1 and 2, respectively. All information was presented in a single session in Experiments 3a and 3b, however, the presenting order of phone information was exactly the same as in Experiments 1 and 2. If the results of Experiments 1 and 2 were due to the different distributions of the positive/negative attributes in the first and second sessions, we would expect the results of Experiments 3a and 3b to be approximately the same as those of the first two experiments for the unconscious thought and conscious thought conditions. 4.1. Method 4.1.1. Participants and design Experiment 3a included 46 undergraduate students (19 women and 27 men) from Northwest Normal University who were equally but randomly assigned to one of two conditions: unconscious-thought (14 men and 9 women) or conscious-thought (13 men and 10 women). We used a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design. Experiment 3b included 38 undergraduate students (22 women and 16 men) from Northwest Normal University who were equally but randomly assigned to either the unconscious-thought (7 men and 12 women) or the conscious-thought (9 men and 10 women) conditions. A 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 4.1.2. Materials and procedure The materials used in Experiments 3a and 3b were the same as those in Experiments 1 and 2. However, the experimental procedures differed from the first two experiments in the following ways: (1) there was no interval between the two information sessions and (2) both groups had 6 min to prepare for the final evaluation after all information was presented.
82
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
4.2. Results and discussion Experiment 3a used a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Avg vs. Worst–Avg) mixed ANOVA, which indicated no significant main effect of thought condition, F(1, 44) = .44, p > .05. However, we did find a significant main effect of phone type, F(1, 44) = 4.12, p < .05, with the Best–Avg phone (M = 71.35, SD = 13.59) receiving a significantly higher score than the Worst–Avg phone (M = 66.48, SD = 15.31). There also was a significant interaction between thought condition and phone type, F(1, 44) = 6.91, p < .05 (see Fig. 3a). The scores given to the Best–Avg phone (M = 73.35, SD = 15.39) by the unconscious thought group in Experiment 3a was significantly higher than the one given to the Worst–Avg phone (M = 62.17, SD = 15.87), F(1, 44) = 10.85, p < .001. Results for the unconscious thought group were approximately the same as those in Experiments 1. Hence, we can deduce that integration by unconscious thought is not impacted by whether or not information is partitioned, which further supported the Unconscious Thought Theory (Dijksterhuis & Nordgren, 2006). However, the scores given by the conscious thought group to the Best–Avg (M = 69.35, SD = 11.51) and Worst–Avg (M = 70.78, SD = 13.74) phones were no-differed in Experiment 3a, F(1, 44) = 0.18, p > .05. Although it appears that the results were the same as those of Experiment 1 for the conscious thought condition, our interpretations of the reasons for these results are different. In Experiment 1, the conscious thought group relied on information from the second session and evaluated the Best–Avg and Worst–Avg phones as equally preferable. Due to the elimination of the interval between the two sessions in Experiment 3a, the positive attributes of the Best–Avg and Worst–Avg phones were no longer as salient as those in Experiment 1. In addition, the amount of information in the temporally continuing session was doubled. We conclude that combining these two factors might skew the overall evaluation made by conscious thought, making it more difficult for participants to discriminate between a bad phone and a better one and resulting in the same scores being given to Best–Avg and Worst–Avg phones. Results of subsequent Experiment 3b also supported this assumption.
Fig. 3a. Mean evaluations for the two phones under different thought conditions in Experiment 3a. Errorbars represent standard errors of the mean.
Fig. 3b. Mean evaluations for the two phones under different thought conditions in Experiment 3b. Errorbars represent standard errors of the mean.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
83
In Experiment 3b, a 2 (thought condition: conscious thought vs. unconscious thought) 2 (phone type: Best–Worst vs. Worst–Best) mixed ANOVA was conducted on the scores from Experiment 3a and revealed no significant main effects or interactions, all Fs < 1 (see Fig. 3b). The scores given to Best–Worst (M = 69.16, SD = 15.41) and Worst–Best (M = 68.89, SD = 16.75) phones by people in the unconscious thought group in Experiment 3b approximately were the same. The results for the unconscious thought group were approximately identical to those in Experiments 2. Likewise, as the scores given to the Best–Worst (M = 70.58, SD = 13.88) and Worst–Best (M = 70.05, SD = 13.37) phones by the conscious thought group were not significantly different in Experiment 3b. The results for the conscious thought group were different from those of Experiment 2 as well. In the second half of session, the Worst–Best phone was better than the Best–Worst one, but the conscious group evaluated them as equally preferable. These results further support our hypothesis that the complexity of information influenced participants’ evaluations in the conscious thought condition in Experiment 3a. Based on the above analyses, we could conclude that it was the partitioning of the information, rather than the distribution of the positive/negative attributes in the first and second sessions, that was responsible for the results in Experiments 1 and 2. Our finding also suggests that partitioning information impacts information integration made by conscious, but not unconscious, thought.
5. Experiments 4a and 4b Previous research showed that participants can make on-line judgments of four phones during the two periods of information acquisition (Hastie & Park, 1986; Lassiter et al., 2009; Waroquier, Marchiori, Klein, & Cleeremans, 2010). Therefore, the results of partitioned information integrated by participants in the unconscious thought condition in Experiments 1 and 2 merely may be due to them recalling a judgment that was formed on-line. We designed Experiment 4 to further assess, and hopefully exclude, this possibility. Experiments 4a and 4b were, respectively, controlled experiments of Experiments 1 and 2. In Experiments 4a and 4b, all participants had a period of delay (mere distraction) between sessions 1 and 2. After session 2, participants were divided into three groups: conscious thought, unconscious thought, and immediate decision. Previous research showed that unconscious thought was goal-dependent (Bos, Dijksterhuis, & van Baaren, 2008). In the study by Bos et al. (2008), participants were given information about some cars and divided into two groups. One group was required to think consciously about their impression of the cars for several minutes. The other group was asked to perform a distraction task within that same time. However, within the distraction group, some participants were told they would be asked their opinion of the cars after the distraction task (goal-dependent), while others were not given this information (mere distraction). Results showed that only the goal-dependent participants in the unconscious thought condition performed better than did those in the conscious thought condition. Thus, if the explanation of the on-line judgment account is correct, the immediate decision and unconscious thought groups would equally integrate information from both sessions to reach a final judgment. 5.1. Method 5.1.1. Participants and design Experiment 4a included 60 undergraduate students (34 women and 26 men) from Northwest Normal University who were equally but randomly assigned into one of three conditions: unconscious-thought (7 men and 13 women), conscious-thought (8 men and 12 women), or immediate decision (11 men and 9 women). A 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Avg vs. Worst–Avg) mixed design was applied. Experiment 4b included 63 undergraduate students (33 women and 30 men) from Northwest Normal University who were equally but randomly assigned to one of three conditions: unconscious-thought (9 men and 12 women), consciousthought (10 men and 11 women), and immediate decision (11 men and 10 women). A 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Worst vs. Worst–Best) mixed design was employed. 5.1.2. Materials and procedure The materials utilized in Experiments 4a and 4b were exactly the same as those in Experiments 1 and 2, respectively. As in Experiments 1 and 2, the between-session interval was 3 min. The other procedures were identical to those in Experiments 1 and 2, with the following exceptions: (1) participants were not informed that the information would be divided into two sessions, (2) after the first presentation of information, all participants were required to complete a 2-back task for 3 min but were not told that they would have to perform another task after the distraction, and (3) an immediate decision group was added. After the second presentation of information, participants in the conscious thought group were asked to think about phones for 3 min, those in the unconscious thought group were required to complete a 2-back task for 3 min, but were informed in advance that they would have to give their opinions about the phones. The immediate decision group was asked to immediately evaluate the four phones.
84
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
5.2. Results and discussion Experiments 4a and 4b both applied a 3 (thought condition: conscious thought vs. unconscious thought vs. immediate decision) 2 (phone type: Best–Avg vs. Worst–Avg) mixed-design ANOVA. Results of Experiment 4a did not show a significant main effect of thought condition, F(2, 57) = .01, p > .05; however, there was a significant main effect of phone type, F(1, 57) = 4.36, p < .05. Specifically, scores of the Best–Avg phone (M = 70.71, SD = 13.01) were significantly higher than those for the Worst–Avg phone (M = 67.10, SD = 12.67). The interaction between thought condition and phone type also was significant, F(2, 57) = 7.58, p < .01 (see Fig. 4a). If the on-line judgment account was credible, the scores for Best–Avg and Worst–Avg phones given by the immediate decision and unconscious groups should be the same. However, in Experiment 4a, the immediate decision group gave approximately the same scores to the Best–Avg (M = 69.35, SD = 12.36) and Worst–Avg (M = 68.30, SD = 11.06) phones, F(1, 57) = 0.12, p > .05, while the scores given by the unconscious thought group to the Best–Avg phone (M = 75.60, SD = 12.89) were significantly higher than for the Worst–Avg phone (M = 62.75, SD = 12.30), F(1, 57) = 18.36, p < .001. These results suggest that the immediate decision group depended on information presented in the second session when making their evaluation. Therefore, results do not support the on-line judgment account. Furthermore, these results replicated our findings from Experiment 1 regarding unconscious thought. The scores given to the Best–Avg (M = 67.20, SD = 12.92) and Worst–Avg (M = 70.25, SD = 13.91) phones by the conscious thought group were appropriately the same, F(1, 57) = 1.03, p > .05. These results also replicated those regarding the role of conscious thought from Experiment 1. Results of Experiment 4b showed that the main effect of thought condition was not significant, F(2, 60) = .75, p > .05; however the main effect of phone type was significant, F(1, 60) = 4.36, p < .05. Specifically, scores for the Best–Avg phone (M = 67.17, SD = 12.91) were lower than those for the Worst–Avg phone (M = 72.95, SD = 12.29). The interaction between thought condition and phone type also was significant, F(2, 60) = 3.95, p < .05 (see Fig. 4b). In Experiment 4b, scores given by the immediate decision group to the Worst–Best phone (M = 73.43, SD = 13.05) were significantly higher than for the Best–Worst phone (M = 62.67, SD = 12.76), F(1, 60) = 9.30, p < .01, while scores given to Worst–Best (M = 70.29, SD = 10.11) and Best–worst (M = 72.52, SD = 12.08) phones by the unconscious thought group were approximately the same, F(1, 60) = .40, p > .05. These results were the same as those of Experiment 4a, namely, the evaluations of two groups were different. The score for the Worst–Best phone (M = 75.14, SD = 13.52) was significantly higher than for the Best–Worst phone (M = 66.33, SD = 12.46), F(1, 60) = 6.23, p < .05. Similarly, the results replicated those of Experiment 2 in conscious thought condition. Combining the results of Experiments 4a and 4b revealed that the on-line judgment account did not influence the integration of information made by the unconscious group, but possibly impacted the conscious group. Overall, Experiment 4 demonstrated that unconscious thought mainly relies on off-line judgment (Strick, Dijksterhuis, & van Baaren, 2010).
Fig. 4a. Mean evaluations for the two phones under different thought conditions in Experiment 4a. Errorbars represent standard errors of the mean.
Fig. 4b. Mean evaluations for the two phones under different thought conditions in Experiment 4b. Errorbars represent standard errors of the mean.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
85
6. General discussion The present study demonstrated that when information was presented in two partitioned sessions, conscious thinkers tended to emphasize information in the second session (i.e., the non-different evaluation in Experiment 1 and the significantly different evaluation in Experiment 2), while unconscious thinkers emphasized neither the information received in the second session (Experiment 1), nor the information received in the first session (Experiment 2). In Experiment 3, all information was presented to participants in one temporally continuous session, but the order in which information was presented was kept the same as when information was partitioned (Experiments 1 & 2), excluding the possibility that the results from the first two studies were due to differences in the distributions of the positive/negative attributes in the first and second sessions. In Experiment 4, an immediate decision group was added, which further eliminated the on-line judgment account. These results support the hypothesis that unconscious thought can equally integrate temporally scattered information, while conscious thought tends to attach particular importance to the most recently received information. The current study explored the mechanism of unconscious thought in integrating temporally scattered complex information. The temporal factor has been believed to be essential for complex consumer decision-making (Peters & Büchel, 2011; Slovic, 2012). Previous research has suggested that unconscious thought leads to an unbiased evaluation by integrating relevant information. The current study examined this hypothesis by presenting complex information in a partitioned way and revealed that unconscious thought integrated the separated information without bias as well. Our results provide further support for the Unconscious Thought Theory and did so with a higher ecological adaptability. The present study provides a new perspective on long-term decision-making. In our daily lives, we frequently are involved in decision-making that lasts for a prolonged period. Under these circumstances, related information is not obtained in just one session, but rather, is presented in several successive sessions. Conscious thought mainly depends on the most recently received information, which can affect decision-making by leading to a biased evaluation. On the other hand, as supported by results of the current study, unconscious thought equally integrates information received in different sessions and leads to an impartial evaluation. Future research should examine the generalizability of these results for making decisions in authentic situations, particularly those that involve high risk. We have argued here that the greater processing capacity of unconscious thought may explain the absence of cognitive bias in integrating temporally partitioned information. With a greater capacity (see also Dijksterhuis, 2004; Miller, 1956; Wilson, 2002), unconscious thought is able to integrate information from a wider range of time periods during the evaluation, in contrast with conscious thought, which particularly relies on recently received information (i.e., the recency effect; Anderson, 1981; Sasaki, 2009) due to its limited processing capacity. When information needed for decision-making is displayed in a temporally separated manner (as in the current study), the interval between information presented first and last is significantly larger, resulting in an unavoidable recency effect. In other words, the overall evaluation is entirely dependent upon information presented in the second session in conscious thought due to the inability to hold information from the earlier session. However, in unconscious thought, information from both the earlier and later sessions can be retrieved for the decision-making process, eventually leading to an unbiased evaluation. Our results contradict those of previous studies on unconscious thought and decision-making (González-Vallejo et al., 2013; Newell, Wong, Cheung, & Rakow, 2009). In the Newell et al. study, participants were presented with attributes of two cars. One car’s attributes were ordered in an ascending pattern, while the other’s attributes were ordered in a descending pattern. The results showed that the car with the ascending pattern was rated as the top car by the unconscious thought group, providing evidence for a recency effect. Conversely, González-Vallejo et al. found a primacy effect. In this study, González-Vallejo and colleagues presented participants with the attributes of four cars in two ways. For the car with the greatest number of positive attributes, they presented the positive attributes first. In the other group, information on the positive attributes of the car with the least number of positive attributes was presented first. The results demonstrated that the unconscious group preferred the car with the greatest number of positive attributes only in the condition where positive attributes were presented at the beginning of the sequence. We believe task complexity is the main reason for the differences between our results and the results of these previous studies. According to Unconscious Thought Theory, unconscious thought is preferable for processing complex tasks (Dijksterhuis et al., 2006). Using meta-analysis, Strick, Dijksterhuis, Bos, van Sjoerdsma, and van Baaren (2011) found that the unconscious thought effect was larger when a problem was sufficiently complex. Compared with Newell et al. and González-Vallejo et al., the tasks in our study were more complex in several ways. First, we presented a larger amount of information. There were 4 options and 64 messages in the present study, while Newell et al.’s study merely had 2 options and 20 messages. While there were 4 options in the study by González-Vallejo et al., there were only a total of 48 messages. Secondly, the present study had longer periods between presentations of information. In contrast to the previously mentioned studies, our study inserted time intervals between information presented first and information last, which accounts for the longer presentation time. Future research should explore how unconscious thought equally integrates information temporally scattered across two sessions. We propose two possible approaches to evaluating this in more detail: first, unconscious thought evaluates each mobile phone by the addition of all positive attributes and that of negative ones across two sessions. Another possible approach is to evaluate the relative merits of each mobile phone based on its positive and negative attributes across two sessions. Subsequently, unconscious thought makes an overall judgment according to the integration of the two evaluations. For instance, in the present study, unconscious thought evaluates that the Best–Worst phone is superior to the Worst–Best one
86
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
in the first session, while the former is worse than the latter in the second session. Then, the results of these two sessions are integrated, giving participants an overall evaluation of mobile phones. Based on what we have mentioned above, the first approach asks people in the unconscious thought condition to follow strict rules. However, Unconscious Thought Theory states that unconscious thought cannot follow strict rules; it gives only rough estimates. Studies also showed that using unconscious thought to integrate information was a gist-based process (Abadie, Waroquier, & Terrier, 2013; Dijksterhuis, 2004; Strick et al., 2011). Specifically, Abadie et al. found that unconscious thought was dependent more on gist memory to process information than on verbatim memory. Therefore, we are inclined to support the second approach. However, this approach requires further confirmation through future experiments. In conclusion, the present study confirmed that recency effect is involved when conscious thought integrates the temporally scattered information, whereas unconscious thought is free from the influence of when information is presented and, thus, can equally integrate information scattered throughout partitioned sessions. Acknowledgments This research is supported by the National Natural Science Foundation of China (Nos. 31200786, 31170974), Key Project of Humanities and Social Sciences, Ministry of Education (No. 07JZD0029), and the Fundamental Research Funds for the Central Universities, Foundation for Young Teacher of Northwest Normal University (NWNU-LKQ N-11-34). References Abadie, M., Waroquier, L., & Terrier, P. (2013). Gist memory in the unconscious-thought effect. Psychological Science, 24, 1253–1259. Anderson, N. H. (1981). Foundations of information integration theory. New York, NY: Academic Press. Ariely, D., & Zauberman, G. (2000). On the making of an experience: The effects of breaking and combining experiences on their overall evaluation. Journal of Behavioral Decision Making, 13, 219–232. Ariely, D., & Zauberman, G. (2003). Differential partitioning of extended experiences. Organizational Behavior and Human Decision Processes, 91, 128–139. Bos, M. W., Dijksterhuis, A., & van Baaren, R. B. (2011). The benefits of ‘‘sleeping on things’’: Unconscious thought leads to automatic weighting. Journal of Consumer Psychology, 21, 4–8. Bos, M. W., Dijksterhuis, A., & van Baaren, R. B. (2008). On the goal dependency of unconscious thought. Journal of Experimental Social Psychology, 44, 1114–1120. Bos, M. W., Dijksterhuis, A., & Van Baaren, R. B. (2012). Food for thought? Trust your unconscious when energy is low. Journal of Neuroscience, Psychology, and Economics, 5, 124–130. Costabile, K. A., & Klein, S. B. (2005). Finishing strong: Recency effects in juror judgments. Basic and Applied Social Psychology, 27, 47–58. Chapman, G. B., Bergus, G. R., & Elstein, A. S. (1996). Order of information affects clinical judgment. Journal of Behavioral Decision Making, 9, 201–211. Crano, W. D. (1977). Primacy versus recency in retention of information and opinion change. The Journal of Social Psychology, 101, 87–96. Creswell, J. D., Bursley, J. K., & Satpute, A. B. (in press). Reactivation links unconscious thought to decision making performance. Social Cognitive and Affective Neuroscience.
J. Li et al. / Consciousness and Cognition 25 (2014) 77–87
87
Strick, M., Dijksterhuis, A., & van Baaren, R. B. (2010). Unconscious thought effects take place off-line, not on-line. Psychological Science, 21, 484–488. Strick, M., Dijksterhuis, A., Bos, M. W., van Sjoerdsma, A., & van Baaren, R. B. (2011). A meta-analysis on unconscious thought effects. Social Cognition, 29, 738–762. Usher, M., Russo, Z., Weyers, M., Brauner, R., & Zakay, D. (2011). The impact of the mode of thought in complex decisions: Intuitive decisions are better. Frontiers in Psychology, 2, 37. http://dx.doi.org/10.3389/fpsyg.2011.00037. Waroquier, L., Marchiori, D., Klein, O., & Cleeremans, A. (2010). Is it better to think unconsciously or to trust your first impression? A reassessment of unconscious thought theory. Social Psychological and Personality Science, 1, 111–118. Wilson, T. D. (2002). Strangers to ourselves: Discovering the adaptive unconscious. Cambridge, MA: Harvard University Press. Yang, H., Chattopadhyay, A., Zhang, K., & Dahl, D. W. (2012). Unconscious creativity: When can unconscious thought outperform conscious thought? Journal of Consumer Psychology, 22, 573–581. Zhong, C. B., Dijksterhuis, A., & Galinsky, A. D. (2008). The merits of unconscious thought in creativity. Psychological Science, 19, 912–918.