Improving teacher questioning

Abstract

This study investigated how non-native speaker teachers made questions from academic text. It looked for improvement of both the form and content of teacher questions through practice and minimal feedback, the strategies teachers used, and the effect that text type had on teacher questions. Before improvement in the form of teacher questions could be measured, a series of stages outlining the learning process of question-making in L2 learners was designed. Errors in teacher questions were then held up to these stages thus showing improvement to be not just a smaller number of errors over time but also an increase in the complexity of errors. To focus the teachers on the content of their questions, they were told that their questions were for students who would later have to summarise the text. Because research on summarization showed that an effective measure of a summary is the percentage of main ideas covered, a similar measure was used to judge improvement in content of teacher questions. The data for this study comprised approximately 1700 questions that were gathered over ten one hour sessions from a class of 17 teacher trainees taking part in a summer language course at the English Language Institute. During each session, all the subjects were given a text which they were asked to read, and from which they were asked to find the main ideas and to make 10 questions that would focus their students on those main ideas. It was predicted that with practice on task, teachers would slowly move through the designated stages in errors of form just as the L1 learner does between 2-4 years old. Unfortunately results only suggested improvement. This was due to both uncontrolled factors like text difficulty and background knowledge of the teachers, the problems involved in working with errors and the stages of measurement themselves. One definite area of improvement was in teacher attitude to the task. At the end of the study, subjects completed a questionnaire and overall most subjects felt more confident and better at question-making and found the study to be worthwhile. Investigation of the strategies teachers used to make questions came down to three basic question types: General questions - where the teacher made a general question that could be applied to many different texts eg. "What is the main idea of this text?"; Text-linked questions - where the teacher made a question directly linked word by word to a proposition in the text; Thoughtful questions - where the teacher really read the text well, understood it well and thought carefully about what the writer was trying to say before making questions. Advice to teacher trainers from this part of the study is that "general questions" are fast and easy to write, provide little chance of error on the part of the teacher and make the students think but can only be used in small numbers. 'Text-linked" questions immediately orient the students to where the answer is in the text but are prone to errors due to text-interference. "Thoughtful questions" are more likely to be good questions in terms of content and are far less likely to contain errors due to text interference. This backs up the idea that a stage in the teaching or learning process is only as good as the one before it. If the teacher masters the text at the reading and comprehension level, good teacher questions in both form and content are much more likely.