Reliability and validity of three Chinese-version tasks of Mayer–Salovey–Caruso Emotional Intelligence Test.

Aim This study examined the reliability and validity for Taiwanese populations of three Chinese-version tasks of the Mayer--Salovey--Caruso Emotional Intelligence Test (MSCEIT): a face task (FT), emotional-management task (EMT) and emotional-relations task (ERT). Background For patients with schizophrenia and affective disorders, one aspect of successful outcomes of treatment and nursing care in primary healthcare is patients' ability to manage emotions (Russell & Potter 2002). Emotional management is the ability to modulate one's own and others' feelings to deal with stress and to promote self-understanding and growth (Mayer et al. 2003, Green et al. 2005). This ability can be assessed in western populations using the MSCEIT, as recommended by a National Institute of Mental Health conference breakout group, Measurement and Treatment Research to Improve Cognition in Schizophrenia (NIMHMATRICS) (Green et al. 2005). The MSCEIT is a valid and reliable self-administered tool to assess emotional intelligence that is a cognitive- and emotional-processing ability (Mayer et al. 2003). However, no scale assessing these abilities in Taiwanese populations was found in a literature review. The MSCEIT includes eight tasks in four branches: perceiving emotions, facilitating emotions, understanding emotions and managing emotions (Mayer et al. 2003). The perceiving-emotions branch includes face and picture tasks that measure ability to recognise feelings. The facilitatingemotions branch includes two tasks that measure ability to generate emotions to foster ideas. The understanding-emotions branch includes two tasks that measure ability to understand simple and complex emotions. The managingemotion branch contains EMT and ERT that measure ability to manage emotions in solving problems. This study followed the NIMH-MATRICS suggestion to assess emotional-management ability using three tasks (FT, EMT and ERT). In the FT, participants identify degree of facial expression. In the EMT, participants identify effectiveness of strategies to manage their emotions to solve problems. In the ERT, participants identify effectiveness of strategies needed to incorporate emotions into decision-making involving other people (Mayer et al. 2003). Methods A non-experimental cross-sectional survey was used with a convenience sample of three groups of adults (age ≥18 years). Group 1 (53 undergraduate and graduate students) was used for reliability testing. Group 2 (40 security guards and maintenance personnel) and group 3 (38 clinical patients with schizophrenia and affective disorders) were used for validity testing. Participants in group 2 and 3 had no significant differences in age and educational level. After the study site's institutional ethics committee approved the study, data were collected. The study purposes and procedures were explained, and subjects' confidentiality and anonymity were assured. All individuals signed written informed consent before participation. Data were collected by personal data sheet, three Chineseversion MSCEIT tasks (Mayer et al. 2003) and the Wong and Law Emotional Intelligence Scale (WLEIS) (Shi & Wang 2007). The personal data sheet recorded sex, age, educational level and marital status. The three Chinese-version MSCEIT tasks tested participants' emotional intelligence. The WLEIS tested the criterion validity of MSCEIT subscales. The WLEIS is a 16-item self-report measure with a seven-point Likerttype scale for assessing emotional intelligence based on a trait-oriented model. The 49 items of the three Chinese-version MSCEIT tasks were satisfactorily translated and back-translated from English to Mandarin Chinese in this study. The 20-item FT includes four item parcels, each with five responses on a fivepoint scale. The 20-item EMT includes five parcels with four responses each. The nine-item ERT includes three parcels with three responses each. The stability of MSCEIT subscales was established by intra-class correlation coefficients (ICCs) for group 1. Item-scale correlations were analysed by Pearson's coefficients. MSCEIT internal consistency was established by Cronbach's alpha for three groups. The criterion validity of the MSCEIT tasks was analysed by the ICC correlations between their scores and the WLEIS scores for group 3. Differences between the three groups were assessed by ANOVA ANOVA. Results The total sample was on average 29.9 years old (SD 9.54), with 13.91 years (SD 2.78) of education. The majority was women (n = 78, 60%). Subjects' mean scores by group and Cronbach's alpha for each task are reported in Table 1. For group 1, the six-week, test--retest correlation for 49 items was 0.76 (ICC), with confidence interval 0.337-0.802. High ERT scores were associated with high EMT (r = 0.53, p<0.01) and FT (r = 0.37, p < 0.05), but EMT and FT scores were not significantly associated. In validity testing, high WLEIS scores were mildly correlated with high FT (ICC = 0.28), EMT (ICC = 0.31) and ERT (ICC = 0.23) for 38 clinical patients with schizophrenia and affective disorders, but only EMT reached significance. In cross-sample validity testing, the three groups differed significantly in scores for the three tasks (Table 1). Conclusions This study provides evidence for the stability, internal consistency, criterion validity and cross-sample validity of three Chinese-version MSCEIT tasks, which are stable subscales for assessing emotional intelligence. Cronbach's alpha and item--scale correlations indicate that FT and ERT form coherent subscales. Our EMT scores had a lower internal consistency than that previously reported (Mayer et al. 2003), suggesting that EMT may not be appropriate to use separately for Taiwanese people. However, we found the three MSCEIT tasks had low correlations with the WLEIS for clinical psychiatric patients. This unexpected finding may be attributable to the two scales assessing different types of emotional intelligence (Goldin et al. 2008). The MSCEIT measures test-related emotional intelligence and the WLEIS measures self-perceptions about emotional intelligence. Finally, the cross-sample test showed that the three tasks had enough sensitivity to measure performance differences between different groups. To conclude, the construct validity of the three Chinese-version MSCEIT tasks needs to be examined by factor analysis on larger clinical samples with psychotic disorders. The tasks are reliable tools that may be suitable for measuring ability to manage emotions in Taiwanese individuals with psychosis. Relevance to clinical practice Emotional intelligence plays a meaningful role in clinical nursing practice (Akerjordet & Severinsson 2007). Managing emotions such as aggression, anger, fear, anxiety and depression is a mental health issue in primary healthcare for both health providers and psychotic patients (Russell & Potter 2002). The three Chinese-version MSCEIT tasks include 49 items that can easily be applied to assess ability to manage emotions. These tasks can help in assessing the emotional intelligence of Taiwanese individuals with mental illness. The three tasks may also be used to evaluate treatment effects or nursing care outcomes aimed at improving patients' ability to manage emotions in solving daily problems. [ABSTRACT FROM AUTHOR]