<![CDATA[The EFL learners in a multilingual society face learning problems both on skill-based and content-based instructions. The problems are caused by linguistic and non-linguistic factors. In accordance with this, better classroom techniques in skill-based and content-based instructions should be developed in such a way that the EFL instructions run well. This paper, which is derived from the result of a research conducted in 2015, particularly discusses an appropriate model of classroom technique which is better used in a content-based instruction for EFL learners at university level in a multilingual society. The analysis and pedagogical discussion are aimed at formulating and finding a better classroom technique in a content-based instruction, namely Research on English Language Teaching (RELT) subject. The data and relevant information presented and analyzed are those of a quasi-experimental research in the design of non-equivalent control group conducted at the English Department of FKIP Universitas Bung Hatta, Padang. The population of this research was students who took RELT subject in 2015/2016 academic year. The sample of this study was 44 students; there were 23 students in class A (experimental group) and 21 students in class C (control group). The treatment given to the experimental group was student-initiated question as the classroom technique, while lecturer-provided question was the classroom technique (the treatment) used in the control group. The data were statistically analyzed by using t-test; the result of data analysis tells that the classroom technique used in the experimental class, the student-initiated question, was significantly more effective than teacher-provided question. It may be argued that student-initiated question is a better classroom technique used in a content-based instruction, RELT subject, for EFL learners in multilingual societies as it ran well at the English Department of FKIP Universitas Bung Hatta, Padang]]>
