Is English-medium instruction effective in improving Chinese undergraduate students' English competence?Abstract: This study investigates whether English-medium instruction (EMI) has an impact on Chinese undergraduates' English proficiency and affect in English learning and use. A cross-section of 136 sophomores and juniors were drawn from an English-medium and a parallel Chinese-medium program. Data included: (a) participants' scores on two national standardized English proficiency tests; (b) their English-related affect as measured by three scales adapted from Gardner's (2004) Attitude/Motivation Test Battery; (c) their perceptions of EMI in Chinese tertiary education elicited with a survey developed by the Chinese Ministry of Education (2006), and (d) interviews with 10 focal students from the Englishand Chinese-medium programs. Results showed no statistically significant effect of medium of instruction on English proficiency or affect in English learning and use. However, extent of satisfaction with EMI, perceived necessity for EMI, and perceived increases in study burden had statistically significant effects on the outcome measures. Additionally, prior English proficiency was the strongest predictor of subsequent English proficiency and English-related affect. These findings raise concerns about the quality of the focal English-medium program and point to students' perceptions of EMI and prior English proficiency as crucial influences on further language learning and use.Keywords: affect in English learning and use, Chinese higher education, English-medium instruction, English proficiency, medium of instruction policy Brought to you by | University of Brighton Authenticated | 194.81.203.94 Download Date | 7/11/14 12:37 PM 1 All quotations from Chinese-language sources and the participants in this study are translated into English by the authors. Brought to you by | University of Brighton Authenticated | 194.81.203.94 Download Date | 7/11/14 12:37 PM 2The MOE allows students in some provinces/autonomous regions/municipalities to take the matriculation English tests developed and administered by educational authorities in these places. Participants in this study came from 29 provinces/autonomous regions/municipalities. Consequently, many participants did not take the NMET, and only 33 had NMET scores. If the NMET and regionally developed matriculation English tests were assumed to be comparable, 97 participants' scores were available for comparison. With this dataset, there was still no significant difference between the EM students (n = 45, M = 127.47, SD = 9.67) and their CM counterparts (n = 52, M = 126.06, SD = 8.56), t (95) = 0.76, p > .05. Brought to you by | University of Brighton Authenticated | 194.81.203.94 Download Date | 7/11/14 12:37 PM