I have decided that I wish to include statistics in my own research, despite having little previous experience of using statistics. I intend to carry out a survey on innovative pedagogical methods in school science and wish to observe a sample population of pupils being taught by innovative methods (e.g. with a focus on enquiry-based learning) vs. traditional methods and carry out face-to-face interviews of these pupils, where I can ask open-ended questions. I also wish to include pupil opinion surveys, e.g. a Likert-type questionnaire based on pupils’ expressed opinions. I realise that this will have limitations, e.g. the expressed opinions may not accurately represent what pupils believe; the response rate may or may not be good; pupils may not take the research seriously. In the face-to-face interviews, pupils may not be telling what they really believe, as they may feel anonymity would not be maintained (Robson, 2011). I am yet to decide on a suitable sample size and on what questions I should include, and this will probably take me a long time until I have a clearer understanding of the ideal questionnaire. Considering my background in science (and, though I am shy of admitting this — my part-positivist stance), I do not yet know whether I will have some difficulty in accepting the ‘acceptability’ of a questionnaire, which can include responses that are not true reflections of pupils’ real opinions/attitudes.
I now have a very basic idea of how to select the appropriate statistical test, depending on whether there is a difference or relationship between variables, types of data, whether or not the samples are related etc. I have just begun to understand how a specialist software, such as the SPSS may be used to analyse the complex data collected. I need to spend a good deal of time in learning in depth how SPSS works, in order to be able to effectively employ it in data analysis. However, as Robson (2011) states, when the two different paradigms of quantitative and qualitative research are combined, there can be a greater variety of potential benefits. This suggests that multiple strategy research is possible and that qualitative and quantitative research paradigms need not be incompatible. In the case of my research, quantitative data analysis on its own may not be able to lead to an in-depth explanation of the results. I would therefore wish to employ both quantitative and qualitative research methods, the former to acquire the needed data, on the basis of which I shall try to provide an interpretation based on the latter method, e.g. the implications of using innovative teaching-learning methods in different social, cultural and economic contexts. This should not only help me in my efforts to answer a range of research questions to a greater depth and present a more complete view of my research, but will also enable me to develop a wider range of skills and competencies that could be of great use in any future research that I carry out .
- Robson, C. (2011). Real World Research. 3rd ed., Chichester: John Wiley & Sons Ltd., pp. 162-168.
I have been a teacher/researcher in chemistry, with hardly any experience of doing statistical analysis. Quantitative research is therefore a new field to me, just as qualitative research also is. During my PGCE course, I did carry out a very basic survey of high school pupils’ attitudes towards the motivational aspect of science practical work, using a questionnaire that I had myself made up. I involved two Year 10 groups in my research – a top set and a mixed ability set, each consisting of 30 pupils. I prepared a questionnaire which I handed out to every pupil in the two groups. This included questions on pupils’ opinions on the use of practical work in science lessons, availability of resources, the use of the science laboratory and out-of-school settings, the extent to which pupils were able to record, analyse and interpret data etc. The pupils’ responses were obtained in the form of both ordinal and scale data. I then calculated the percentages of pupils giving answers within specific data ranges and interpreted my results in terms of the relative numbers of students favouring or challenging the use of science practicals and its various aspects. Of course, I did all this as a complete novice, as I had no prior exposure to quantitative surveys and had very little idea about how to make up a questionnaire. I also did not have access to software packages, such as the SPSS. Now, with a slightly better understanding of quantitative research in education, I feel that this can be a useful tool in carrying out surveys in my field, particularly when it is backed up with a qualitative analysis of the results. However, I am yet to build up an idea of a number of important aspects of quantitative research, e.g. what should be an ideal sample size? How exactly should I link statistical data to a qualitative interpretation? How do I deal with a situation where I do not get the expected number of responses? At the moment I have a basic understanding of how SPSS works, how to choose a statistical test based upon whether there is a difference or a relationship between data and types of data, how to interpret a distribution curve, as far as T-tests and Chi Squared tests are concerned. But I am yet to grasp the significance of other statistical tests. I know I shall need to devote much time to reading about this kind of research, in order to gain a deeper understanding.
My interest lies in the field of science education research and I shall look at a journal article published in 2007 in the International Journal of Science Education by Rob Toplis, of the School of Sport and Education, Brunel University. The paper is about evaluating school science practical work and focuses on how 14-16 year old students approach anomalous results. Earlier research had shown that the majority of pupils did not know what to do with anomalous results in science investigations. Toplis aims at providing a deeper insight into pupils’ understanding of what anomalous results are and what needs to be done about them. I think it has been written for science education researchers, school science teachers as well as for those involved in curriculum development.
The paper reports a case study of six groups of Key Stage 4 pupils, forming part of a wider case study on pupils’ evaluations in science investigation, using 10 small groups of KS4 pupils with 2-4 pupils in each group. This was an in-depth qualitative study based on observation of the groups when they carried out science practicals, subsequent interviews and analysis of every pupil’s written work. The six groups involved in this study consisted of two from a selective girls’ school, two from a selective boys’ school and two from a non-selective, co-educational school. The group selection was done by teachers, who chose pupils on the basis of their readiness for interviews. Toplis says that there was a compromise between including a range of attainment levels and a mix of genders in the sample groups with the need for an in-depth qualitative study of pupils’ understanding and actions. The study shows that pupils could identify anomalies in data tables or from lines of best fit and tries to explain the lack of action after this identification by pointing out the constraints placed on pupils. It relates these to the National Curriculum policy and model used. On the basis of the evidence provided, the author’s claims appear to be justified.
Investigative science practicals are an important part of enquiry-based science learning, which is one of the areas I wish to work in. I feel that this study addresses a very important aspect of pupils’ understanding. I find it interesting that Toplis mentions that school science investigations involve a non-scientific epistemology as opposed to authentic science research. The author himself identifies the limitations of his research. He admits that this study, based on working with small groups cannot provide generalisations, but can contribute to developing a theoretical understanding of pupils’ response and reasoning in practical science investigations. Also, it would be difficult to record pupils’ verbal and social interactions in a busy laboratory environment. However, the study had the following strengths: it could confirm the individuals and variables and also provide information on results and procedures used, which could be used for subsequent interviews. If I were to carry out a similar research, I would perhaps consider allowing more time to pupils for the actual investigations followed by analysis of their results. I would also allow conversation and collaboration among the groups, e.g. group discussion on what action to take about anomalous results. It may be worth looking at the effect this might have on pupils’ understanding and reasoning.
- Toplis, R. (2007) ‘Evaluating science investigations at ages 14-16: dealing with anomalous results.’ International Journal of Science Education, 29 (2), pp. 127-150.
I shall try to evaluate the design of a piece of research carried out by Abrahams and Millar in 2008, which I first came across a couple of years ago. This paper discusses the effectiveness of practical work as a teaching and learning strategy. During my teaching years, I had often been uncertain about how much practical work to include in lessons and have found this paper to be very relevant.
The authors have used a case-study approach. According to them, the large scale quantitative studies of school science practical work in the UK carried out previously studied the rhetoric of practical work rather than the reality, as they provided insights into the views of students and teachers, but did not compare those views with observations of actual practice. Again, questionnaire-based surveys and interview studies are more likely to reproduce existent rhetoric, rather than provide accurate insights into the reality of teaching within its natural setting. The authors therefore wanted to critically analyse the reality of school science practical work and they needed a strategy that would bring the researcher into close contact with teachers and students, collecting data in the teaching laboratory, observing actual practices and carrying out interviews on the basis of these observations.
The advantages and disadvantages of their choice of research design are as follows:
- A case-study approach would help in critically exploring the relationship between reality and rhetoric within an educational context.
- The authors used a multi-site approach involving a series of 25 case studies in different settings, rather than a single in-depth case study. According to Schofield (1993), the possibility of studying numerous heterogeneous sites helps in increasing the generalizability of qualitative work.
- The responses to interview questions would be more connected to reality when an interviewee knows that the interviewer has observed the practice under discussion.
- The authors had limited control of the content or subject matter of the lessons actually observed.
- The study distinguishes two levels of effectiveness of a practical task – whether students did the things the task the designer intended (Level 1) and whether they learned the things they were intended to learn (Level 2). However, the design of the study allowed much less to be said about the effectiveness of practical tasks at Level 2 than at Level 1.The authors had permission to observe single lessons. Observing subsequent lessons would not have been possible in many cases because of the perceived disruption to routines. Other actions, such as follow-up visits to assess students’ understanding either shortly after the lesson observed or later, were also impossible, because different diagnostic instruments would have to be devised for each observed lesson. This would introduce many new variables and make general conclusions almost impossible to draw. Therefore data collection was limited to a single visit for each practical task.
If I were to do the research, I would focus on a more in-depth understanding of the effect of practical work on pupils’ learning by carrying out my study in one or two schools. Though this might not aid me in generalising my findings very effectively, I would be able to plan a series of subsequent observations over time and build up a deeper insight by following the pupils’ gradual progress in learning.
- Abrahams, I. and Millar, R. (2008) ‘Does practical work really work? A study of the effectiveness of practical work as a teaching and learning method in school science.’ International Journal of Science Education, 30 (14) pp. 1945-1969.
2. Schofield, J.W. (1993) ‘Increasing the generalizability of qualitative research.’ In M. Hammersley (ed.), Educational research: Current issues. London: Paul Chapman Publishing, pp. 91-113. (Cited in 1)