'reading the Mind in the Eyes' in Bengali Populations in India and England: Assaying Effects of Language and Culture

Background:  Studies of social cognition and its disorders must accommodate local cultural norms if they are to be valid, but must be normed and referred to international standards of diagnosis and assessment if they are to be credible. There thus is a need for translation of standard instruments into local languages, and adaptation for local cultures. The 'Reading the Mind in the Eyes' Test (RMET) is a widely used instrument assessing social perception of complex affective and cognitive states from pictures of eyes, based on linguistic labelling of these states. Despite translations into many European and other languages worldwide, till recently no version of the RMET had been formulated for South Asia. A companion study by Dasgupta et al. has translated the adult RMET into bangla, the language of Bangladesh and the Indian state of West Bengal.

Objectives:  To assess cross-cultural validity of the bangla RMET.

Methods:  Data from the bangla RMET collected from 135 (70 female) adult native (first language) bangla speakers in Kolkata, India are contrasted to those from a matched sample of 60 native bangla speakers in Nottingham, England. Further samples of native English speakers in Kolkata and native English speakers of South Asian ethnicity in Nottingham, matched to these other samples for age, sex, education and socioeconomic status, are contrasted against these in a 2 x 2 (culture x language) analysis of variance, using both the full 36-item RMET and a version that omits the six items identified by Dasgupta et al. as not reaching criterion validity in the bangla translation.

Results:  Preliminary results indicate an effect of language but no effect of culture, and no interaction, on 36-item scores but not on the 30-item shortened version. Data analysis is ongoing.

Conclusions:  Despite its faults and confounds, the RMET has demonstrated value as a quick index of one aspect of social cognition. Disentangling effects of language and culture can be difficult, especially for a test such as the RMET that (1) relies so heavily on verbal encodings, and (2) depends on monoracially caucasian images with which most people in the world have little or no perceptual experience. Our preliminary results suggest that in the particular case of the RMET, language is a greater determinant than culture, but that comparability can be achieved by excluding a small set of linguistically problematic items. Future RMET versions should make use of ethnically diverse facial emotion models representative of worldwide populations.