Open or Closed - That is the Question

Open or Closed – That is the QuestionDiane HarrisUniversity of ManchesterPaper presented at the British Educational Research Association Annual Conference, University of Warwick, 6-9 September 2006AbstractThis paper explores teachers‟ use of ques tions with primary school children engaged in science activities. The research reports the profile of questioning strategies from a survey of 51 teachers in 102 lessons in Reception, Years 2 and 4: more …open‟ questions are used with the older groups of children than the reception class children and in science activities rather than literacy; additionally slightly more person-centred questions are asked by teachers in science activities than in literacy.Received wisdom in primary …constructivist‟ science teaching is that the teacher‟s questioning role in science classes, at least in the elicitation phase, is to discover what the children already know relevant to the science to be learnt and hence, teachers should ask open and person-centred questions (e.g. SPACE Project, 1987 to 1990). In contrast, some educational psychologists, probably on the basis of experience, believe that four- to seven-year-old children are confused by open questions and prefer to offer the children choices. Consequently, some psychologists have argued that children‟s cognitive potential can be seriously under-estimated by open questioning. Indeed, Donaldson (1978) argued that children as young as three-and-a-half understand the concept of choice and have no problem working with it. She argues that, by not giving children the clue that a problem involves certain choices, we are making things unnecessarily difficult for them. Inappropriate use of language and task rules, therefore, can create a barrier to finding out what childr en …really‟ know.Thus in this research (part of a larger PhD research project funded by the ESRC) we ask: to what extent do teachers of children (at different ages from 3 to 11 years) use different questioning strategies, including …open/closed‟ and …p erson/subject-centred‟ questions in science discussions, and with what effect on the quality of children‟s engagement?The survey involved twenty North West primary schools with observations being carried out in three year groups (Reception, Year 2 and Year 4) to highlight any differences in the questioning strategies used by the teachers with the different age groups. These observations were paired i.e. the same teacher was observed teaching both science and literacy so that comparisons could be made of any differences in teaching styles between the elicitation phase of the science activity and the shared text element of literacy.These observations generated over four thousand teacher questions which were categorised according to the type of question - …open/closed‟ and …person/subject-centred‟. The children‟s responses were also categorised revealing the varied impacts of different questioning strategies on classroom discourse and the children‟s engagement with the activity. In addition, the impact of …wait-time‟ was considered when the children were unable to answer.The survey reveals that teachers‟ strategy-use conformed to the expected pattern: more …open‟ questions are used with the two older groups of children than the reception class children and in science activities rather than literacy. These results also suggest that, throughout this entire age range, slightly more person-centred questions are asked by teachers in science activities rather than in literacy.Although these data sustain the importance of teacher questioning in any learning situation, it needs to be the …right‟ question i.e. one that the child can comprehend. These findings suggest that the notion of …openness‟ in questioning needs to be theorised as …distributed‟, i.e. tha t it is a function of the entire context of discourse and not only of the structure of the …question‟ as presented by the teacher. It is this appreciation of context which the younger children appear to find extremely difficult and so, in many cases, open questions seem unable to scaffold the children‟s thinking which results in the children not making connections with their own and their peers‟ ideas. It appears, therefore, that context is decisive in determining whether open and closed questions are scientifically productive.IntroductionThis survey reports the findings of an empirical survey carried out in twenty North West Primary schools where I worked with fifty-one teachers and carried out 102 observations –thirty-four in each of the three year groups (Reception, Year 2 and Year 4). These observations were paired i.e. the same teacher was observed teaching one science lesson and one non-science lesson (literacy) so that I could compare any differences in teaching styles between the elicitation phase of the science activity and the shared text element of literacy. These observations generated 4891 questions which were then categorised according to the type of question - …open/closed‟ and …person/subject-centred‟. The children‟s responses were a lso categorised revealing the varied impacts of different questioning strategies on classroom discourse and the children‟s engagement with the activity.In addition, I carried out more detailed case studies and I have selected some short extracts from these to enrich and expound the survey data.Open v. Closed Questions; Person-Centred v. Subject-Centred QuestionsI teach trainee primary school teachers to teach science and, as part of their course, they learn about the theories of teaching science. During the summer of 2001, I decided to update my teaching skills and gain some experience of working with the new …Curriculum Guidance for the Foundation Stage‟. I worked as a volunteer in a private day nursery with three and four-year-old children for two full days each week for four months. I initially used the strategies I had been teaching to the traineeteachers. The majority of science teaching literature, including the …Curriculum Guidance for the Foundation Stage‟ and other literature specificall y written for the Foundation Stage, advises the use of open questions for finding out children‟s ideas. Open questions should encourage a child to share his or her thoughts and ideas, whereas the use of closed questions will only generate short answers and therefore limit further discussion. Very quickly, however, I discovered that using open-ended questioning did not necessarily elicit the answers for which I was hoping. I found that very often the child would appear confused and be unable to answer the question at all. This observation appeared to be very significant, although I acknowledge that the problem may not be related to the type of question at all, so I decided to search for anyone else having similar experiences. There are various factors which may also be at work here including the vocabulary and grammar used by the teacher and understood by the child and the maturity of the child‟s language generally so I expected that I would have to investigate a variety of disciplines in order to come to a better understanding of my problem.The Science Processes and Concept Exploration (SPACE) Research of 1987 to 1992 was classroom–based research which explored the ideas that pupils in primary school (5 to 11-year-olds) already possessed in science and whether, within a normal classroom situation, the children might be encouraged to modify their ideas following relevant experiences (Osborne, et al., 1994).It was the teacher‟s role in the …elicitation phase‟ to discover what the children already knew by questioning. It was found that pupils from all three age groups do have prior ideas about most things, although their explanations are not always very scientific. The study then went on to investigate whether, within a normal primary classroom, the ch ildren could be encouraged during an …intervention phase‟ to develop their prior ideas into a more scientific understanding (Russell and Watt, 1990).The SPACE Research concluded that 5 to11-year-old children did indeed have prior ideas and that it was possible for teachers to help them develop these ideas into a more scientifically useful form – a constructivist approach to learning (Osborne, et al., 1994). The research led to the development of curriculum materials, Nuffield Primary Science (1993), which conform to the constructivist paradigm and remain very popular in schools. There are separate schemes for children in Key Stage 1 (5-7 years) and in Key Stage 2 (7-11 years).The SPACE Research (Russell and Watt, 1990) also recommended that, in order to encourage the children to share their ideas, teachers should ask open and person-centred questions. More recently, Harlen and Qualter (2004) have identified open and person-centred questions as giving access to a child‟s own feelings, views, ideas and reasoning and therefore encourage the child to investigate.The significance of constructivism and productive questioning remains pivotal to children‟s learning of science in primary schools. Trainee teachers across the Primary Phase, i.e. those who intend to teach Foundation Stage, KS1 and KS2 children, are all advised to ask their pupils open and person-centred questions in science as put forward by the SPACE Research. This strategy has been reiterated in one of the Association for Science Education‟s more recent publications, “Guidelines forScience in Early Years”, which provides examples of open and person-centred questions to promote the children‟s science skills (Newport Advisory Service, 2002). The breakthrough came when I delved into the discipline of psychology and discovered a paper by Springer and Keil (1991, p. 768)which reported that, “...relying on open-ended interviews, as has been done in Piagetian studies of causality, would have run the risk of seriously under-representing children‟s competencies.” Their preferred questioning strategy with four to seven-year-olds is to offer the children choices. Donaldson (1978) reiterates the view that children as young as three-and-a-half understand the concept of choice and have no problem working with it. She argues that, by not giving children the clue that a problem involves choices, we are making things unnecessarily difficult for them. Adult uses of language, therefore, can create a barrier to finding out what children really know. On the other hand, as the experiments of Donaldson (1978) and Tizard and Hughes (1984) all attest, …meaningful‟ context is crucial in providing children with access to tasks. Piaget‟s underestimation of children‟s cognitive ability can be partially attributed to h is reliance on verbal interview techniques. How can young children provide answers which truly represent their ideas if they do not fully understand the question?The use of language must therefore be considered as it is central to communication in the classroom – the teacher communicates information through the spoken word and the children generally demonstrate their understanding through the same medium. It had been felt by teachers that the children who used scientific vocabulary had a better understanding of science than those who used everyday language to describe their ideas. The SPACE Research found that this was not necessarily true because children could competently explain their understanding of a scientific concept even though they did not know the appropriate scientific term (Russell and Watt, 1990). Understanding is therefore far more important than the vocabulary used in the early stages of learning any new information. Often if a child is introduced to a new scientific term and does not properly understand its meaning, then the term will be used inappropriately. In science, therefore, “…the teacher should focus on the exploration of ideas through relating concepts to existing experiences using language which is familiar to the learner” (Heywood, 1997, p.48).The SurveyIt was most appropriate to investigate the problem in two stages as it seemed necessary to explore what questioning strategies are actually used with different ages of children in science and, for comparison, literacy and then to examine the effects of these different strategies on the teaching and learning of science in more detail.An empirical survey was conducted to address the first stage of the problem. Robson (1999) advises that a survey sample needs to reflect the population to be examined and therefore schools were selected from inner city, city suburbs and rural locations throughout the North West. Selecting the sample size also needed careful consideration because, due to the format of the school day in primary schools throughout England, the survey size had to be restricted because of the time available to observe science. Currently literacy and numeracy occupy every morning in school and so science is usually taught for only one afternoon each week. There was, ofcourse, more flexibility in Foundation Stage classes. According to Robson (1999), however, a sample size of one hundred observations would be appropriate for a small-scale survey and I therefore carried out 102 – thirty-four in each of three year groups (Reception, Year 2 and Year 4). These observations were paired i.e. I studied the same teacher teaching one science lesson and one literacy lesson so that I could compare any differences in teaching styles between the subjects.For the survey I was an observer and only attended to the categories included in my classroom observation instrument: teachers‟ questions - …open versus closed‟ and …person-centred versus subject-centred‟ - and children‟s answers which were categorised according to the response of the teacher i.e. …correct‟, …incorrect‟ or …no response‟. I recorded some of these lessons for validation purposes. These data gave high inter-rater reliability. The results were consistent with my own experience: it seems that open questions are preferred in the older classrooms rather than the younger and the effect was present in the literacy classes too.A limitation in the methodology for this research was the lack of corroborative evidence regarding the silences or …wait-time‟ between the end of the teacher‟s question and the beginning of the child‟s reply. There are five possible interpretations of these silences:∙the child was not paying attention∙the child was totally absorbed with the activity∙the child did not know the answer to the question or∙the child did not understand the question itself, the meaning being clouded by the question∙the child was actually thinking.The fifth point is of prime importance because children have to be given time to formulate their ideas before they can possibly provide an answer (Rowe, 1974). Rowe (1974) discovered that these …wait-time‟ periods of undisturbed silence rarely lasted more than 1.5 seconds in normal classroom discourse and the typical length was between 0.7 and 1.4 seconds. She discovered that, if teachers simply increased …wait-times‟ to three or more seconds, then there were benefits to be gained for both teacher and pupils. According to Rowe (1974), Tobin (1987) and Stahl (1994), these positive effects include:∙an incr ease in the length of the pupils‟ answers∙an increase in the appropriateness of the pupils‟ answers∙an increase in the number of volunteered appropriate answers∙ a decrease in the number of unanswered questions∙the pupils‟ marks improve.Increasing …wait-time‟ to three seconds or more should give almost every child sufficient time to complete the cognitive processes needed in each specific situation. The teacher‟s task is therefore to ensure that the pupils have sufficient information, prior to each period of silence, so that the necessary processing can be successfully completed.The Teachers’ QuestionsAll the teachers I have spoken to in the course of my research have been familiar with the idea of open and closed questions. Most have then gone on to explain the reasons why they do or do not use open questions with their particular group of children. The majority admitted that they were not familiar with the terms …person-centred‟ or …subject-centred‟ and therefore did not consider these aspects wh en questioning children. Nevertheless, it is apparent from the survey data collected to date that both person- and subject-centred questions are employed in all the classrooms I have visited.Although I collected data from complete lessons, I have only used the elicitation phase of science lessons and the shared text activity of the literacy sessions. These activities had several similarities which made the comparison the most suitable:∙They took approximately the same amount of time(dependent on the age of the children)∙Both activities called for children‟s ideas and thereforethere was a large amount of teacher questioning∙They were carried out as whole class or whole groupactivities by the teacher which meant that data collectionwas more reliable.I had hoped to be able to compare the plenary sessions in both science and literacy but, for a variety of reasons, teachers very often omitted the plenary in one of the subjects which meant that I was unable to make a direct comparison.Table 1: Types of Questions asked by the Teachers (Harris, 2005)Table 1 records the type of questions asked by the teachers of the selected year groups. These data can be shown graphically for the different types of question as follows:Figure 1: Graph of Percentages of Open Questions used with each Year GroupAs can be deduced from Table 1 and Figure 1, for the Reception Classes, 24% of the questions asked by the teacher in science were open questions whereas only 17.7% of the questions asked in literacy for the same age group were open questions. For Year 2 the percentages of open questions asked by the teacher rose to 50.4% in science and 47.4% in literacy. There is only a small increase in Year 4 in the number of open questions asked by the teacher: 56.1% in science and 53.9% in literacy.Figure 2: Graph of Percentages of Person-Centred Questions used with each Year GroupRegarding person-centred questions, these can be deduced from Table 1 and Figure 2. For the Reception Classes, 42.7% of the questions asked by the teacher in science were person-centred with only 40.1% asked in literacy for the same age group. For Year 2 the percentages of person-centred questions asked by the teacher rose to 51.9% in science with one per cent less (50.9%) in literacy. Once again there is only a small increase in Year 4 in the number of person-centred questions asked by the teacher –52.9% in science and 52.4% in literacy – a difference of only 0.5%.The survey therefore reveals that there is a marked inc rease in the number of …open‟ questions used with the older children in Years 2 and 4 as compared with the children in the reception classes in both science and literacy with more open questions being asked in science. Also, there is a gradual increase in the number of person-centred questions being asked by the teachers as the age of the child increases.The Pupils’ AnswersIn the course of the survey I also kept a record of the type of response the pupils gave to each question asked by the teacher:∙acknowledged by the teacher to be correct∙acknowledged by the teacher to be incorrect∙no response given by the pupil.For the purposes of this paper, the …success rate‟ of each question type was determined by combining the scores for science and literacy and then taking a percentage of each total.Figure 3: Graph of the Percentages of each Category of Question asked in the Reception ClassesFrom the graph, it can be seen that the reception class children have more difficulty in answering open questions than closed questions because their success rate is 59% for both person- and subject-centred open questions as compared with 85% and 86% respectively for person- and subject-centred closed questions. The next most obvious feature of the bar graphs is the large percentage of unanswered open questions – 24% for person-centred questions and 32% for subject-centred questions. This feature will be discussed further in this paper.Figure 4: Graph of the Percentages of each Category of Question asked in Year 2From the graph, it can be seen that Year 2 still have more difficulty in answering open questions than closed questions. Their success rate with open questions is 65% for person-centred and 62% for subject-centred as compared with 88% and 90% respectively for person- and subject-centred closed questions. There remains a large percentage of unanswered open questions –19% for person-centred and 25% for subject-centred.Figure 5: Graph of the Percentages of each Category of Question asked in Year 4By Year 4, the pupils are able to answer open person-centred questions almost as successfully as both types of closed questions because their success rate is 84%, 88% and 93% respectively. These eight- and nine-year-olds, however, still seem to have problems with open subject-centred questions. They continue to be able to answer only 62% of these questions successfully although the number of incorrect answers has fallen from 13% to 10% since Year 2. For my sample pupils, the percentage of unanswered questions has actually increased during this same period, from 25% to 29%.Explaining the Unanswered QuestionsI next decided to turn my attention to the questions that the pupils were unable to answer. A total of 4891 questions were asked by teachers in the course of my survey of which 10.69% were unanswered. I considered that this was important because, even if the pupils answered teachers‟ questions incorrectly, at least they could comprehend the meaning of the question sufficiently to try to answer the question. As previously discussed, according to Rowe (1974), the …wait-time‟ periods of undisturbed silence should last a minimum of three seconds but that in normal classroom discourse the typical length was considerably less. Analysis of my own data revealed that only 9% of the unanswered questions had …wait-times‟ of less than three seconds. I can only conclude that the reason for this is that Rowe‟s research has been disseminated through to the classroom by teacher trainers over the intervening thirty years. Also, the teachers in my survey knew that I was interested in questioningstrategies and this could well have influenced their teaching.Figure 6: Graph of the ‘Wait-Times’ for the Unanswered Questions in the Reception ClassesIn the reception classes, the unanswered open person-centred questions and the closed subject-centred questions had almost the same average length of …wait-time‟ at 4.59 and 4.60 seconds respectively. The teachers allowed the longest average …wait-time‟ for the open subject-centred questions at 4.98 seconds and the shortest average …wait-time‟ for the closed subject-centred questions at 4.31 seconds.Figure 7: Graph of the ‘Wait-Times’ for the Unanswered Questions in Year 2By Year 2 the longest …wait-time s‟ were allowed for open person-centred questions at an average of 4.87 seconds. The shortest average …wait-times‟ were now given for the closed questions: 3.33 seconds for person-centred questions and 3.1 seconds for subject-centred questions. Open subject-centred questions were given an average …wait-time‟ of 4.10 seconds.Figure 8: Graph of the ‘Wait-Times’ for the Unanswered Questions in Year 4In Year 4 the average …wait-times‟ for the four types of question had become less diverse with less than a second between them. The open questions were given the longest average …wait-times‟ at 4.12 seconds for the subject-centred questions and 3.74 seconds for the person-centred questions. The …wait-times‟ for the closed questions were slightly shorter at 3.27 seconds for the person-centred questions and 3.52 seconds for the subject-centred questions.Throughout this survey the open subject-centred questions have been the most problematic for all three age groups including the eight- and nine-year-olds in Year 4. This is partly attributable to …wait-times‟ of less than three seconds but this only constitutes 13% of the open person-centred questions in the reception class and 7% in both Years 2 and 4.The Case StudiesCase studies were also undertaken to examine the effects of different questioning strategies on the teaching and learning of science in more detail. The teachers who took part in the case studies were chosen on the basis of the survey as likely to be interesting or anomalous. For the purposes of this paper I have included two short Year 4 extracts which highlight the context-related problems which exist even with eight- and nine-year-olds. For both the survey and the extracts from the case studies,I classified the questions initially on strictly traditional grammatical grounds. In practice, however, I was aware that an open question may well be interpreted as closed in certain contexts.The first class observation-extract follows a practical introduction to the lesson where the Year 4 pupils are given the opportunity to explore the properties of four materials – a solid, a liquid, a gas and a solid that behaves like a liquid. The materials are all contained in transparent jars and so the pupils‟ investigations are limited to visual inspection and rolling and shaking the jars. The teacher holds up the jar of water and encourages her pupils to explain why they believe that the water is a liquid using a closed subject-centred question. Her pupils concentrate their attention on their own ja rs of water and Jordan comes up with “you can‟t hold it because it‟s not a solid” and he then goes on to explain that it “can‟t be a gas”. Subsequently the teacher queries the fact that a liquid cannot be held as his pupils are holding their jars. She uses an open person-centred question and Jordan swiftly corrects her, “No, Miss. It‟s held in the jar”. Jordan then considers for two seconds before explaining that, “it‟s not a solid because…if it‟s a solid, you can‟t put your hand through it‟ and then re words his comment so that …liquid‟ is the subject of his statement, “you can put your hand through a liquid”. A second closed subject-centred question generates further reasons for believing that the contents of the jar are liquid from other members of the class. These include “it doesn‟t stand still so it can‟t be a solid”, “water isn‟t hard so it‟s got to be a liquid”, “a liquid can…flow into things” and “ if you put your hands in water and then a breeze moves it, they‟ll dry”. All these statements are more than the one-word responses one is led to believe will result from closed questions.Figure 9: Describing the Properties of a Liquid in Year 4One of the reasons I believe that this teacher made such good progress with her pupils was that there was immediate practical scientific context to be observed, examined and explored.She deliberately initiated the discussion with the liquid which was an unusual starting point and the pupils seemed to struggle with this. Many of their initial comments referred to solids and they needed to rethink their explanations. Nevertheless there was no shortage of ideas and both open and closed questions generated high-quality responses.In contrast, the following class observation-extract follows an exchange where the teacher tried to encourage her pupils to come up with a title for the afternoon‟s investigation of why pulse rate changes when running. The teacher then uses a closed question to direct her Year 4 pupils, “think about doing PE, what do we talk about?” which gets a response “your heart goes faster”. Subsequently, the teacher asks an。