THINKING PROCESS DURING ENGLISH IMPROMPTU SPEAKING INDONESIAN STUDENTS OF DIFFERENT LEVELS OF FLUENCY

This research is qualitative descriptive of Multicase design. The descriptive research studies are designed to obtain information concerning the current status of phenomena which directed toward determining the nature of a situation as it exists at the time of a study (Ary, Donald, 1979). The objective of the research is to find out what happen when thinking processes takes place prior to speaking impromptu and also to find out what kinds of phenomena occur during the English impromptu speaking of students of different levels of fluency. The subjects of this research are six Indonesian students with different levels of fluency in English. They are classified as elementary and advanced English level. The data obtained by way of field note taking, tape recording and unobtrusive observation. The data analysis procedure of this research is referring to Miles and Huberman models which stages are data reduction, data display, then followed by conclusion drawing/verification. In this research, the writer classified the data by the subjects age, level of English fluency, time of studying English and type of obstruction occurred. The speaking disruptive phenomena found from the data reduction, is classified into kinds of pauses to be listed based on the frequency of occurrences. This processes draw the researcher into the findings which show that when thinking processes takes place, speakers tend to find ways to draw the thought and stall the time by producing sounds or silence which known as pauses. Since the research is focusing on the subject of non native English, other phenomena following the pauses which occur are the occurrence of repetition, self correction and grammatical error. The conclusion to be drawn from this research is spontaneous or impromptu, given no time or several time to prepare, pauses, error and other disfluencies in speaking still occur, despite the English competence of the speakers.