THE BIZARRENESS EFFECT AND INDIVIDUAL DIFFERENCES IN EMOTIONAL INTELLIGENCE

,

The bizarreness effect refers to a greater likelihood to recall bizarre sentences compared to common sentences (Wollen & Cox, 1981a, 1981b;Cornoldi, Cavedon, de Beni, & Pra Baldi, 1988;Imai & Richman, 1991;McDaniel, Einstein, DeLosh, May & Brady, 1995).The bizarreness effect is a phenomenon that is influenced by a variety of factors which have been examined in a number of studies.A summary of these factors has been described by Toyota (2002).First, the manipulation of bizarreness as a between-subjects variable (unmixedlist) or within-subject variables (mixed-list) impacts the likelihood of the effect.The effect occurred when bizarreness was manipulated as a within-participants variable (Merry, 1980;Wollen & Cox, 1981b;McDaniel & Einstein, 1986;Imai & Richman, 1991;McDaniel et al ., 1995), but it did not occur when bizarreness was manipulated as a between-participants variable (Wollen, Weber, & Lowry, 1972;Cox & Wollen, 1981).Second, the type of memory test that is used impacts whether or not the effect will be observed.The effect has been observed in free recall tests but not in cued recall measures (Wollen & Cox, 1981a, 1981b;McDaniel & Einstein, 1986, 1989;Toyota, 1987).A third factor that impacts the effect is the type of encoding task.The bizarreness effect occurs when participants are given instructions to create a mental image of a sentence or provided an orientation task to rate the vividness of the image in sentences.The effect was not present when the instructions or the orientation task were not provided (McDaniel & Einstein, 1986;Toyota, 1987).Finally, the type of bizarre sentence impacts the likelihood of the effect.Unusual sentences (e.g., a waiter serving a tie ) which were strange but possible, and truly bizarre sentences (e.g., a waiter driving a tie ), which were strange and impossible have been compared and findings indicated unusual sentences were less likely to initiate the bizarreness effect (Cornoldi et al ., 1988).
The dimensions of bizarreness (bizarre vs. common) and the dimensions of the vividness of images (vivid vs. dull) have also been examined as indicators of the bizarreness effect (Toyota, 1987).Four types of sentences have been compared which include, bizarre image sentences (e.g., Baby drinks beer ), bizarre non-image sentences (e.g.,Baby is old ), common image sentences (e.g., Baby drinks milk ), and common non-image sentences (e.g., Baby is young ).In two separate experiments, participants rated the congruity of each target to its sentence frame and rated the vividness of each image for the sentence frames.Even though the conditions (congruity vs. vividness) differed in the experiments, bizarre image sentences led to better free recall than the common image sentences.Moreover, differences were not observed between bizarre non-image sentences and common non-image sentences.Most notably, although participants were not given instructions to create a mental image of the referents, the bizarreness effect was found with image sentences (Toyota, 1987).The results suggest that the occurrence of the bizarreness effect was mediated by whether an image in each bizarre sentence was imagined.In other words, there was a greater likelihood of creating a mental image with a bizarre sentence which led to greater recall of these sentences.
To examine these findings, Toyota (2002) investigated whether the vividness of images could be determined by individual differences in imaging ability.Previous work by Robinson-Rielger & McDaniel (1994) had examined the relationship between individual differences in imaging ability and the bizarreness effect.Results indicated that the difference in magnitude of the bizarreness effect was not found between participants with a high imaging ability (good imagers) and those with a low ability (poor imagers).Interestingly, the bizarreness effect was not observed in participants exposed to complex sentences revised from the original sentence set by Kroll, Schepeler, & Angin (1986).Although the interaction of imaging ability and type of sentence (bizarre vs. common) was not significant, good imagers recalled targets in bizarre sentences more often than poor imagers (Kroll et al , 1986).This suggests that the interaction of imaging ability and type of sentence do not depend on the type of sentences.
Previous studies found a small bizarreness effect when the targets were embedded in complex sentences (McDaniel & Einstein, 1989;McDanielet al. , 1995).Additional research also suggests that the benefits of bizarre sentences compared to common sentences on increasing recall are non-existent, because the cues associated with the complexity of sentences are more effective for retrieval than focusing on sentence bizarreness (Robinson-Rielger & McDaniel, 1994).According to Wothen & Deschamps (2008), "Several researchers (Cornoldi et al. , 1988;McDaniel & Einstein, 1986, 1991) have suggested that mental imagery mediates the effects of bizarre elaboration effects and, as a result, the phrase'bizarre imagery effect'has been used to describe the facilitative effects of bizarre elaboration on memory.However, research has demonstrated that the facilitative effects of bizarreness are affected by neither a significant reduction in mental imagery formation (Weir & Richman, 1996;Worthen, 1997) nor differences in imaging ability (Anderson & Buyer, 1994;Zoller, Workman, & Kroll, 1989)."(p.462).However, Toyota (2002) suggested that using simple sentences may be more appropriate to examine the relationship between imaging ability and the bizarreness effect.The bizarreness effect has been found in good imagers, whereas the effect was not as apparent in poor imagers.This suggests that individual differences in image ability are a critical factor in the presence or absence of the bizarreness effect.Although there is inconsistency in the results on image formation and imaging ability, image seems to be one of the critical factors in determining the likelihood of the effect.
Another important factor in determining the bizarreness effect is emotion.Bizarre sentences arouse strange feelings, unusual emotions, and humorous feelings, which can make bizarre sentences more distinctive, and thus, more memorable.One study found that the bizarreness effect was only observed when it induced humor (Worthen & Deschamps,2008).Perhaps humor mediates the facilitation of the bizarreness effect, which suggests that that emotion created by bizarre sentences is critical in facilitating recall.
Additionally, the relationship between individual differences in the ability to process emotion and performance of memory has been examined.Individual differences in processing emotional attributes of episodes elicited by words has been studied in relation to emotional intelligence (EI) (Toyota,2011).Salovey & Mayer (1990) defined EI as 'the subset of social intelligence that involves the ability to monitor one's own and others' feelings and emotions, to discriminate among them, and to use this information to guide one's thinking and actions ' (p.189).This definition has been followed by a number of similar definitions with minor distinctions (Davies, Stankov and Roberts,1998;Mayer, Caruso and Salovey,1999;Toyota, 2008;Toyota, Morita & Takšić, 2007).Research indicates a difference in recall performance between participants with a high EI and low EI (Toyota, 2011).When words were associated with episodes expected to occur in the future, participants with High EI recalled more of the words associated with both pleasant and unpleasant episodes.Participants with low EI recalled with a similar level of words across the three target types.These findings suggest that EI is associated with the processing of emotional attributes associated with future episodes.
The self-choice effect in intentional memory also differs among participants with high and low EI (Toyota,2013).Participants were given three types of pair words: pleasant-pleasant, pleasant-unpleasant and unpleasant-unpleasant pairs.They were asked to choose one word that reminded them of pleasant episodes and to remember it for each pair in the self-choice condition, or to remember the underlined word while recalling pleasant episodes in the forced-choice condition.Results indicated differences in the size of self-choice effect (the superiority of the self-choice to the forced-choice condition) between participants with high and low EI.Namely for the pleasant-pleasant pair, a larger self-choice effect was observed in participants with high EI compared to low EI, because participants with a high EI could process the emotional attributes of a pleasant emotion associated with the pleasant word precisely, and choose a word with a more pleasant emotion.For the pleasant-unpleasant pair, the self-choice effect was observed for both types of EI participants.For the unpleasant-unpleasant pairs, it was difficult for participants with high EI to process the emotional attributes of a pleasant episode associated with an unpleasant word.As a result, a small self-choice effect was observed for participants with high EI.The reverse effect (the superiority of the forced choice to the self-choice condition) was observed for participants with low EI, which infers they have difficulty processing emotional attributes associated with pleasant episodes elicited by an unpleasant word.EI plays a major role in regulating emotion (Salovey & Mayer, 1989;Toyota et al. , 2007).If EI has an effect on emotion processing elicited by a target word, the magnitude of the bizarreness effect would be influenced by the level of EI.Participants with high EI can suppress the arousal of emotion elicited by bizarre sentences, whereas for those with low EI, they cannot suppress the same emotional arousal.The more emotional arousal elicited by bizarre sentences, the more distinctiveness could be increased.If true, EI would be negatively correlated with the recall performance of bizarre sentences.
The purpose of the present study was to investigate the relation between EI and recall of bizarre sentences.. Two types of conditions were provided: an emotion condition and an image condition.The emotion condition involved rating the degree of unusual feelings of each target and its sentence frame.In this task, participants have to be sensitive to unusual feelings or strange emotions elicited by bizarre sentences, and oriented to process the emotion from the sentences.Higher EI may function to suppress emotions elicited by bizarre sentences, which would indicate a negative correlation between EI and recall of bizarre sentences.During the image condition, participants were asked to rate the vividness of images and their sentence frames.In this task, participants were less sensitive to emotion, because they focused on the images created by the sentences, which would indicate no correlation between EI and the size of bizarreness effect.It was predicted that EI would be negatively correlated with recall of bizarre sentences in the emotion condition, but no correlation between EI and recall would be observed in the image condition.

Design and participants
A 2 x 2 design was used where the first (between-participants) factor was the type of condition (emotion vs. image) and the second (within-participants) factor was the type of sentence (bizarre vs. common).Participants were 35 students in nursing school with a mean age of 22.56 yrs.(SD=5.03yr.).Seventeen students (5 males and 12 females) were assigned to the emotion condition and 18 students (4 males and 14 females) were assigned to the image condition.

Materials
The target words to be remembered and their sentence frames were selected from a previous publication (Toyota, 1987(Toyota, , 2002)).The targets and their sentence frames were written in Japanese Hiragana characters which were familiar to the participants.Two types of image sentence frames were provided: bizarre image and common image sentences.Examples of the two types of sentence frames for a target (e.g., Baby ) are as follows:" drinks beer"as a bizarre image sentence (bizarre sentence), and " drinks milk"for a common image sentence (common sentence).Four types of lists were included in the incidental study phase.The construction of the list was the same as in previous studies (Toyota, 1987(Toyota, , 2002)).Each list had 15 of each of the two sentence types (bizarre and common) and four buffers.Two of the buffers were at the beginning of the list and two were at the end of the list.These buffers were intended to minimize the primacy and recency effects in the free recall test.The buffers were excluded from data analysis.Four different orders were constructed for each list and they were counterbalanced.For both the emotion and image conditions, the sentences were presented in a booklet with the following information on separate pages: a target, its sentence frame, and a 6-point rating scale.Two types of rating scales were provided: a scale for the degree of strange feelings elicited by the sentence and a scale for the vividness of the image for each sentence.

Procedure
The experiment was conducted with an incidental memory procedure in a small group session.The procedure was the same as a previous study (Toyota, 2002) except for the task used in the incidental study phase.
(a) Incidental study phase.For the emotion condition, each participant received a booklet with the following instructions."A word and a sentence are shown on each page.Your task is to rate the degree of strange feelings or unusual emotions using a 6-point scale where 1 indicates usual or ordinary and 6 indicates unusual or strange.Fill in the blank in each sentence frame with your rating."For the image condition, each participant received a booklet with the following instructions."A word and a sentence are shown on each page.Your task is to rate the vividness of each image using a 6-point scale where 1 indicates dull and 6 indicates vivid.Fill in the blank in each sentence frame with your rating."A period of 10 seconds was allowed for each page.
(b) Interpolated task.A sheet of paper with Japanese words printed in Hirakana was given to each participant.Participants were told to quickly circle as many nouns as possible with more than three letters.This task was used in the previous study as the incidental memory procedure (Toyota, 1987).Participants were given 3 minutes to perform the interpolated task.
(c) Free recall test phase.Unexpected immediate free recall was conducted next.Participants were asked to recall and write down as many as possible of the words (targets) presented.Three minutes were allowed for free recall and writing.
(d) Questionnaire on emotional intelligence phase.The Japanese version of the Emotional Skills and Competence Questionnaire (J-ESCQ) (Toyota et al. , 2007) was administered to assess the level of EI in each participant.The J-ESCQ is comprised of 24 items divided into three subscales: Perceiving and Understanding Emotions, e.g., "I notice when somebody feels down"; Expressing and Labeling Emotions, e.g., "I am able to express my emotions well"; and Managing and Regulating Emotions, e.g.,"I try to stay in a good mood."Previous studies (Toyota, 2011;2013) used the J-ESCQ and found it a valid measure of EI.All participants were provided the J-ESCQ and asked to rate on a 5-point scale (1: Never, 2: Seldom, 3: Occasionally, 4: Usually, 5: Always) how often they felt or thought each of the statements.The total score of the J-ESCQ was used to assess the level of EI in each participant.

RESULTS
The mean proportion of target words recalled in the free recall test is shown in Table 1 as a function of the type of condition (emotion and image) and the type of sentence (bizarre and common).An analysis of variance found a main effect of the type of sentence (F 1,33 =10.00, p<.01.η 2 =.13), where recall performance of bizarre sentences was higher than common sentences.The main effect of type of condition (F=.23) and the interaction between the type of sentence and type of condition (F=.72) were not significant.To examine the relationship between EI and recall performance, Pearson's correlation coefficients (r) were calculated for performance in both groups (See Table 1).A negative significant correlation was found between EI and recall performance of bizarre sentences in the emotion group.

DISCUSSION
The purpose of this study was to examine the relationship between individual differences in EI and the bizarreness effect.We predicted that the correlation between EI and recall performance of bizarre sentences would be significant in the emotion condition.Results indicated that the recall performance of bizarre sentences was negatively correlated with EI when participants were oriented to process the emotion.These findings indicate that EI can suppress emotions aroused by bizarre sentences.It has been proposed that humor mediates the facilitation of the bizarreness effect in recall (Worthen & Deschamps, 2008).Humor is a critical factor because it plays a role in controlling and regulating emotions caused by bizarre sentences.
Conversely, EI has the potential to suppress emotions related to humorous, strange or curious statements, because the degree of distinctiveness is lessened (Schmidt, 2006).The degree of distinctiveness in bizarre sentences can impact the ability to recall them.Consequently, participants with high EI appear to have a tendency to suppress the emotion that arises from processing a bizarre sentence.This tendency could lead to a lower recall performance of bizarre sentences.The results of the present study corroborate findings from Toyota (2011) and Toyota (2013), indicating a relationship between EI and free recall performance with an intentional memory procedure.In future research on recall and the bizarreness effect, the implications of EI should be further examined.EI plays a significant role in interpretation and memory of experiences and additional research should continue to examine this connection.

Table 1
Mean proportion of recall as a function of types of condition and sentence, and correlation coefficient (r) between recall and EI.