4 Research paradigm and methods
4.4 Analysis methods
In this study the analyses done in phases 1 and 3 follow the general principles of content analysis. Analysing the historical data in the first phase is done as a content analysis of texts of historical documents. As particular aspects of analysing historical data the views of two historically orientated methods are applied. The first method – the critical-historical method – highlights the history of the science dimension of the work emphasizing criticality and aspects that respect contemporary science, whereas the second method – the cognitive-historical method – points out the cognitive processes of past science and learning as connecting factors (see sections, 4.4.2 and 4.4.3).
The analysis method used in the third phase of the study is also content analysis, which uses the videos of the teaching experiment as data.The purpose of content analysis is to go into the level of language of physics and deal with pupils’ external representations of DC-circuit phenomena. (See section 4.4.1)
70 Figure 14 Analysis methods in this study.
4.4.1 Content analysis for historical and empirical data
Content analysis is widely used analysis method in qualitative research. The method is suitable for analysing written, heard or seen contents; it is typically used in qualitative research. Content analysis generally includes the following phases: 1) limitation of the data, 2) transcription of the data, 3) analysing the data by way of classifying, and 4) writing a summary (Tuomi and Sarajärvi 2002, 13, 93-95). The limitation of the data can be done for example by choosing certain parts of the data to be analysed. One example of limitation is to choose a certain amount of critical episodes or critical moments of the whole data and analyse only these. According to Nurkka and Viiri (2005, 694-695; see also Viennot et al. 2004) the critical episodes can be chosen on the basis of previous research of typical difficulties on the subject matter, and by picking the most important episodes of learning. In this study the critical episodes are chosen to describe the best learning situations. The choice was made to best answer the research questions. The chosen episodes and grounds used will be examined in section 7.1. There are also a number of different ways of transcribing data. In this study the videotaped data was transcribed with the Transana program. With the program the video and transcribed text can be synchronised to help analysis.
In content analysis, analysing phase means finding out different classes of information, and arranging data to different themes or types (Patton 2002, 463; Tuomi and Sarajärvi 2002, 95). Briefly, the purpose of content analysis is to arrange the data into a compact and clear form without losing the information included (Patton 2002, 453; Tuomi and Sarajärvi 2002, 110). In phenomenographic research, which is one class of qualitative research and which does the same kind of data arranging, the data is classified by means of categories of description (Marton 1994, 4424). Pupils’ different conceptions are analysed and similarities are collected into similar categories. The data is examined many
Research method:
- Qualitative research o Design research o Case study
Research paradigm:
Pragmatism
Analysis methods:
- Content analysis
- Critical-historical method - Cognitive-historical
method Gathering data:
- Written documents - Small-group teaching - Small-group interviewing
71
times until different categories have stood out from the rest. Also Chi (1997) has proposed a method for coding and analysing verbal data. However, the main idea of Chi’s is similar to others, the purpose of the analysis is to limit the data, segment it, code the data and look for similarities or patterns. In this study the littered data was analysed with the Weft QDA program, which is software suitable for analysing qualitative data. The program was used to help organise data into classes and to interconnect earlier classes.
In doing content analysis the researcher has to choose a logic of reasoning, this can be either inductive or deductive. In this study the logic used is inductive in the content analysis of historical data, and intermediate in doing the content analysis of empirical data.
In the intermediate form of logic both inductivity and deductivity are applied, this is called the abductive approach. The rationale behind the abductive approach is to apply both the views of data grounded (inductive) and theory-driven (deductive) approaches. The abductive approach uses the earlier knowledge, but instead of purely testing the theory, the aim is to search for new thoughts. In the abductive approach the theory affects the classification of the data, because the phenomenon is already known on the grounds of theory. In this way, the theory can help the analysis. (Tuomi and Sarajärvi 2002, 95-99, 116)
The abductive approach in content analysis can be used for design research, the structure of the research used in this work. As discussed in sections 1.4 and 1.5, the structure of research in this study follows a three-phase cycle (see Figure 1 in section 1.4) and the sequential phases affect the following phases. The new phases are also based on the earlier phases. Thus, the historical models of DC-circuit phenomena found as results of content analysis in phase 1, naturally affect the teaching experiment in phase 3, and this theoretical knowledge is used as a background for content analysis.
As described in section 4.3.2 small group learning is understood to be an effective way of learning. In this study the empirical data gathered is mainly talk and discussions in a small group. Gestures are also used to check the meaning of talk in ambiguous situations (Mavrou et al. 2007, 167). Certainly, the data also includes some drawings and text, but the greater part is oral or visual data. The form of gathered data is chosen from the standpoint of the research questions set, the purpose of gathering empirical data is to get information about the development of small group’s external representations of DC-circuit phenomena (see section 1.3 for research question 2).
Pupil’s or small group’s external representations are special conceptions of the subject matter. On the other hand, developing external representations are understood as a proof of an active learning process, which can be studied by content analysis.
Classroom discussions have recently been researched in many studies. For instance Boulter (2000), Mortimer and Scott (2003), Viiri and Saari (2006), Mortimer and Machado (2000), Nassaji and Wells (2000), Wells (1999), Scott (1998), Chan et al.
(1997), and Chinn and Brewer (1998) have studied discussions from different standpoints.
72 4.4.2 The critical-historical method
According to the critical-historical method one aim of the history of science is to “function as an analytical instrument for the critical evaluation of methods and concepts that appear in modern science”. Thus besides the historical importance of the description of the developments of concepts, the special role of the historical course of events is also in its influence on the physics of today. (Kragh 1987, 32-33) Jammer clarifies the aim of the history of science by pointing out that the aim of it is to lead “to a profounder comprehension of the meaning of the term and to a higher level of understanding of its role and significance in physics” (Jammer 1961, VII).
The typical features of the used method are horizontality of study, anachronism vs.
diachronism and criticality of study. Horizontality (Kragh 1987, 80) means that the concept formation processes are studied through time, and the topics chosen are studied during this period. The other feature of the method is the axis of anachronism vs.
diachronism (Kragh 1987, 89-90). A purely anachronistic method would only study past science from the viewpoint of physics of today. In this study, however, past science also has a value of its own, so there is an aspect of diachronical history of science too. In this work diachronism can be seen in the valuation of qualitative sensory experiments and comparative experiments as an autonomic part of concept formation. The theoretical framework used is modified from the view of today’s physics, but the aim of it is not to impact on or distort the conceptual structure of the past. The anachronical view of this study appears in the clearest way in the connection of prequantitative experiments when properties are understood as forerunners of modern quantities.
4.4.3 The cognitive-historical method
Nersessian’s (2002b; 1995) cognitive-historical method is a tool to make use of past science’s practices and use it to understand learning processes for example in the case of conceptual change. The cognitive-historical method combines historical and contemporary scientific practices with cognitive science investigations of aspects of human cognition.
By studying past science it is possible to get information of the practices, which scientists employ in their work. This is the “historical” dimension of the method. On the other hand, the “cognitive” dimension is used as a “lens” of the examinations of the past science. The objective of this kind of research is to “identify various cognitive practices employed in scientific cognition; to develop explanatory accounts of the generativity of the practices;
and to consider, reflexively, the implications of what is learned for understanding basic cognitive processes generally”. Further, it is always possible that there might emerge some new aspects of cognition in the cognitive-historical analysis, which could even alter current cognitive science understanding.
To get information on human’s reasoning, representing and learning the cognitive-historical method makes use of all kinds of cognitive-historical records of science like notebooks and publications. According to Nersessian cognitive-historical analysis can include for
73
example dimensions of concept formation and change, and using and developing modelling tools and instruments. (Nersessian 2002b, 135-136)
Like other historical studies in general, the nature of the first phase of this study is to find out a historical course of events, in this case the development process of the historical models of DC-circuit phenomena. The interest is thus in the historical modelling processes and also in modern conceptions of the nature of DC-circuit phenomena.
The special feature of the first phase of this work, resulting from the chosen method of research, is the interpretation, not only representation, of historical events. The interpretation is done from the point of perspective of the designed framework for the modelling process. For this purpose, schematic, graphical maps were chosen. The work has been organised so, that by using these maps a summary of the historical models is given step by step. These maps are shown Figure 8 in section 3.5 where the "master map", summarizing the features taken to be essential for historical models of DC-circuit phenomena, are displayed. Each map concentrates on the phase of concept formation under discussion.
74