Our preliminary analysis showed very high inter-rater reliability for the scoring of our newly developed Japanese version of the LEAS (LEAS-J). Although the relationship with TAS-20 showed rather poor correlation, except for EOT, the results for NEO-FFI and IRI indicate that the LEAS-J has good concurrent validity. The current findings confirm that the Japanese written descriptions of each scenario are essentially equivalent to the original, which indicates that the LEAS-J would be useful for assessing the level of emotional awareness of Japanese subjects.
The LEAS-J was shown to have good internal consistency. The finding for α coefficients was very similar to the findings of a previous study by Lane, et al. , suggesting that the LEAS-J provides sufficient reliability. We strictly classified our subjects' descriptions, using the glossary of words in the LEAS scoring manual .
In the age cohort comparison of LEAS-J and LEAS scores, however, the mean scores of almost all of the LEAS-J scenes were significantly lower than those of the LEAS. This could be due to a number of factors. The original LEAS was given to students in a classroom setting, while the LEAS-J subjects were requested to complete the scenarios at home. The difference in the collection method might have had some influence on the difference. Another factor may be the Japanese style of expressing feelings, which has been reported to be much different from that of people in Europe and the U.S. . Japanese are said to generally use fewer emotion related words and, in fact, the emotion-related word count of LEAS-J was significantly correlated with the LEAS-J score for every scenario for Self, Other, and Total in the current study. Another possible explanation is that expressing feelings depends on social context: Japanese, in general, tend to attenuate and control emotional expressions more than Americans . Such a tendency modulates the social desirability in the style of expressing feelings, especially when writing; for example, a mature Japanese person should express feeling in a rather reserved way. Perhaps consciously accessible thoughts and rules about what is appropriate directly influence which emotions do and do not reach conscious awareness . It will be necessary in the future to determine if the Japanese style of expressing feelings contributed to the lower LEAS-J scores.
In addition to scoring by the glossary of the original LEAS as stated above, even after the same subjects' words and descriptions were re-scored assessing the individual level of emotional awareness as a whole at the conceptual level, the mean scores of all scenes on the LEAS-J remained significantly lower than the level of the mean LEAS scores. Nisbett & Masuda  compared the context sensitivity of Americans and Japanese, and suggested that Asians appear to attend more to the field and Westerners more to salient objects. Further study will be necessary to clarify the differences in the inventories; especially the different styles of emotional life between Americans and Japanese.
Construct and concurrent validity were obtained by examining the relationship between LEAS-J scores and TAS-20, IRI, and NEO-FFI scores and are as follows.
The LEAS-J Self score showed a significant, negative correlation only with the EOT scores of TAS-20. This finding is similar to the finding that the LEAS subscales were negatively correlated with only the EOT of patients with somatoform disorders [30, 31] and the finding that in other studies the LEAS tends to correlate most highly with the EOT factor of the TAS-20. In contrast to the DIF and DDF of TAS-20, EOT is more accurate in ratings because the items of EOT ask subjects to rate themselves on a skill or habit that they could easily be aware of . EOT is also less influenced by depression or anxiety . Furthermore, EOT has been more closely associated than the other two factors in various objective measurements regarding affect regulation, such as physiological indices like baseline heart rate , and in the Affect Consciousness Interview . Based on these observations, the more specific relation between LEAS-J and EOT should increase the validity of LEAS-J as a psychological measure that can objectively probe an individual's levels of emotional awareness.
The present study is the first report of the relationships between LEAS-J and IRI and NEO-FFI. The significant correlations of all aspects of LEAS-J with fantasy (FS), perspective taking (PT), and empathic concern (EC) in IRI suggest that the ability to take into account the feelings of oneself and others with sensitivity underpins the imaginative function that places others in the position of oneself (FS), an ability associated with understanding the inner emotional state (PT) of others and of consideration for others (EC).
The LEAS-J scales were significantly associated with the NEO-FFI scores for Extraversion (E), Openness-to-Experience (O), and Agreeableness (A). Extraversion includes sociability, liveliness, and the general experience of positive affect. Openness-to-Experience includes aesthetic sensitivity, intellectual curiosity, need for variety, non-dogmatic attitudes, high curiosity and interest in the internal and external world. These positive associations with E and O are consistent with their negative associations with TAS-20 [14, 35]. In addition, the present findings for Agreeableness (A) are supported by our previous study showing its negative associations with TAS-20 among Japanese subjects . Agreeableness includes a high degree of a sense of cooperation with others, such as trust, altruism, and sympathy. It appears that in Japan higher emotional awareness contributes to greater agreeableness, which seems consistent with the hypothesis that sensitivity to the others is highly valued in many Asian cultures. On the other hand, no significant correlations were observed between an excessive interest in the negative feelings of oneself (neuroticism) with the Self, Other, or Total LEAS-J scores. This observation confirms evidence that the LEAS is not influenced by depression or a state of anxiety in healthy subjects [13, 30, 34] and stands in contrast to findings that the TAS-20 is significantly correlated with neuroticism [14, 22, 36].
The scores of women were higher than those of men for all of the LEAS-J subscales. This is consistent with the report by Barrett, et al. , and it is evident that the gender-related difference of emotional awareness is present cross-culturally. The LEAS-J scores were well correlated with the IRI scores in the present study, showing that females tend to be more aware of their own and other's emotions than males. Actually, the emotion-related word count of females on LEAS-J was significantly higher than that of males in the current study, similar to the findings from Barrett et al. . In addition, current findings of gender differences are confirmed by IRI scores  and the EOT scores of TAS-20 . Given that this gender difference in emotional awareness has been demonstrated in children as young as 10 years old  and early adolescents 12~15 years old , these findings raise the distinct possibility that a sex difference in emotional awareness is a biological universal.
In this study, the LEAS-J scale scores were not significantly different by socioeconomic status. However, Lane, et al. [13, 34] found significant differences in LEAS scores between groups of differing socioeconomic status. Our findings indicated very similar socioeconomic status for our participants, as might be expected in a homogeneous sample of Japanese undergraduate students, and, indeed, the vast majority of the subjects belonged to the middle class based on the occupational prestige scores . Therefore, we were not able to precisely assess the influence of socioeconomic status on the level of emotional awareness because of the constricted range in our sample. Further, because all subjects were university students who had passed entrance examinations of similar difficulty, it is possible that their comparability in linguistic ability had an equalizing influence on LEAS-J scores.
Further work is necessary in the following areas: LEAS-J should be given to subjects with a wider range of ages to examine the possible influence of age as an indicator of emotional development on LEAS-J scores; to determine if LEAS-J adequately detects impairments of emotional awareness in clinical samples, such as patients with alexithymia; to develop an additional scoring system that is optimally suited to the Japanese language and culture to supplement the current glossary that is derived from LEAS protocols completed by subjects in the United States and is based on a direct translation from English; and finally, the reliability and validity of LEAS-J will need to be examined carefully by a concurrent structured interview, such as the SIBIQ .