[ Abstract for this presentation ] [ Proceedings Contents ] [ Schedule ] [ Abstracts ]

Conclusions about the efficiencies and effectiveness of educational research methodologies in investigating how students learn

Gail Chittleborough
Science and Mathematics Education Centre
Curtin University of Technology
WAIER logo
Examining how students learn is not only looking at what they know but how they come to know what they know. In this way this research focused on the process of learning as well as the content of learning. This paper examines the educational research methodologies that were used in investigating both of these aspects of how students learn chemistry. Multiple research methodologies were adopted including surveys, cross-sectional studies and naturalistic inquiries. A variety of both qualitative and quantitative data sources were used. The choice of data sources, the samples and the difficulties and limitations of the research methodology is discussed. The validity and reliability of the interpretive analysis is demonstrated with examples from the research into how students learn chemistry. While the research was focussed on particular research questions, the choice of methodology has some commonality with all research situations. The research process itself is a dynamic and developing process and the parallel between the process and content of learning is made to the process and content of research.


The research process of identifying a research problem, establishing specific research questions which lead to the choice of the most appropriate research methodologies and in turn the selection of the most efficient and effective data sources is a standard and well-worn path. Despite this it is not always easy to identify the most efficient and effective data sources. This paper aims to critically evaluate the choice of data sources used in my research. Hindsight provides some insight into this process that may be of use to future researchers.

Factors affecting the choice of data sources include the research process, design, situation and subject. I am focusing on data sources because they are the food or fuel of the study. Researchers spend a lot of time preparing data sources, collecting data, analyzing data and drawing conclusions form the data. Consequently the data sources are crucial to the research. Shortsightedly, research can be seen as just the collection and analysing of data sources. This paper aims to demonstrate the important role of data sources, important even crucial, but not independent of the rest of the research process.

My study investigated how students learn chemistry with a particular emphasis on the role of chemical representations and chemical models. Examining how students learn is not only looking at what they know but how they came to know what they know. Because of this, the research focused on the process of learning as well as the content of learning.

The process of research

The process of research is described in detail in educational research textbooks. Students as novice researchers are inexperienced in the research process, commonly learning how to do research at the same time as conducting the research. Learning through experience is essential to fully understand the research process. Each research situation is unique and consequently has a unique research solution. There is however some aspects that is common to many research situations and has been proven to be significant in my research choices.

My intended methodology of a naturalistic inquiry collecting both qualitative and quantitative data sources was mostly achieved. Aware of the need for triangulation, corroboration and validation, my research design included a variety of data sources. By looking at the detail of the data and focussing on it's analysis it was possible that I ran the risk of loosing sight of the research methodologies. Distinguishing the research methodologies from the data sources is important because the methodologies provide the framework for the research are needed to ground the data sources.

Research situation

The research culminated in four separate studies, each having a variety of data sources, samples and situations. The styles of research were largely determined by the opportunities that were available. The four-year duration of my research resulted because of my desire to make the most of the opportunity to have access to students and to evaluate the implementation of the project.

My research questions changed, as I responded to a new opportunity that is not a problem as long as the new questions are consistent with the research problem, methodology and research subject and that the data is needed to address the research question. Research should be fluid and dynamic, but also mindful of the planning that has preceded it. Taking advantage of opportunities provides more and more data, collected easily but the processing and analysing takes time. Table 1 is presented to show the variety of data sources.

Table 1: List of data sources and the corresponding identification numbers

StudyData SourceAnonymous/
The Number of
the Data Source
1Molecular Chemical Representations (MCR)V2
1My Views of Models and Modelling in Science (VOMMS)V3
1Group work dialogueA4
1Participant researcher's observations

2Scientific Models (SM)V1
2Students' Understanding of Models in Science (SUMS)V2
2My Views of Models and Modelling in Science (VOMMS)V3

3Students' Understanding of Models in Science (SUMS)V1
3My Views of Models and Modelling in Science (VOMMS)V2
3Molecular Chemical Representations (MCR)V3
3What do you thi nk this atom looks like?V4
3Student Unit Experience (SUE)A5
3Worksheets 1-4V6
3Participant researcher's observations
31st InterviewV8
32nd InterviewV9

4Students' Understanding of Models in Science (SUMS)V1
4My Views of Models and Modelling in Science (VOMMS)V2
4Students' Evaluation of Educational Quality (SEEQ)A3
4Online surveyA4
41st Interview V5
42nd InterviewV6
4My observationsNA*7
* NA Not applicable

The comparisons among the four studies became complex. While each of the four studies was independent, there was the added opportunity for comparison. The strength of the multiple methodologies, multiple data sources and multiple samples was countered by needless complexity and lack of fair comparison. In hindsight the research became too large and almost unmanageable.

Research design

By looking at the detail of the data and focussing on its analysis it was possible that I ran the risk of loosing sight of the research methodologies. I had used a variety of methodologies: case studies, longitudinal, cross-sectional study, action-research. The research had become complex. Different methodologies go in different directions and can potentially be conflicting. Distinguishing the research methodologies from the data sources is important because the methodologies provide the framework for the research and are needed to ground the data sources.

Research methods that allowed the value of the research to be recognised were more fulfilling to me as a researcher. For example action research implementing changes to learning styles, and getting feedback from participants or stakeholders' were more fulfilling than the more descriptive studies that were interesting but had no immediate impact.

The research subject

The complexity of "learning"

The research problem concerned students learning of chemistry. Learning is described in terms of the level of understanding that is achieved. In attempts to describe the varying depth of the learning process, the following terminology is commonly encountered: shallow learning (Atherton, 2003), quick learning (Schommer, 1990), rote learning (Battino, 1992), algorithmic style learning (Mason, Shell, & Crawley, 1997), instrumental understanding (Skemp, 1976), passive learning (Yager, 1991) and a surface approach to study (Ramsden, 1992). These approaches are characterised by a lack of conceptual understanding or cognitive effort. In contrast to this are meaningful learning (Ausubel, 1968), intentional learning (Bereiter & Scardamalia, 1989), relational understanding (Skemp, 1976), active learning (Duit & Treagust, 1998) and a deep approach to study (Ramsden, 1992). The latter approaches are characterised by student-focused learning with greater conceptual understanding resulting in higher order reasoning and thinking skills (Stephens, McRobbie, & Lucas, 1999), and deep processing of information and cognitive strategies of high elaboration (Hennessey, 2003). Hennessey describes the term high elaboration to mean "deep processing of information, elaborate cognitive strategies of connecting and comparing existing conceptions with new information, and significant metacognitive reflection about what they were thinking and why" (p. 118).

The term "understanding" is associated and implied in the term learning. Wandersee and Griffard (2002) define understanding "as a dynamic epistemological status conferred on individuals by a consensually recognised referent group within a community of scholars, based upon criteria of intersubjectivity, parsimony, coherence and conceptual transparency" (p. 29).

As a subject of research, learning appears abstract and illusive however, it is apparent that theoretical frameworks from the literature provide structure and descriptors to shed light on the subject. In my study I had 3 theoretical frameworks - models, 3 levels of chemical representations and learning, connecting the learning tools, the way of thinking about chemicals - the mental models - and the learning. The theoretical frameworks provided the common foundation for the content in all four studies. Even though each study was different the commonality of the theoretical frameworks unified them.

Research questions

Considering these complex definitions and descriptions of the learning process, the task of performing research and being able to identify these attributes, establishing how students learn, and their level of understanding of a particular concept is indeed a challenging task. To make this task more manageable the research questions were written in very specific terms concerning students' understanding of particular chemical concepts. From beginning with one research problem, three research objectives were identified, and then each objective generated four specific research questions, resulting in twelve research questions. Particular data sources were chosen to specifically respond to each of the research questions. The determination of the research questions, the choice of research methodology and the corresponding data sources are all-significant in the research process. The links between the research objectives, research questions and data sources can be presented in a tabulated formats see Tables 1 and 2 to show that each objective of the research used quantitative and qualitative data sources from multiple studies. To connect all four studies, all objectives, research questions, data sources had to be tightly linked together. This is demonstrated schematically in Tables 1 and 2 showing the links between the research objectives, data sources, and studies.

Because the interpretive analysis is made in light of previous research, it is advantageous to write the research question in line with the literature e.g. the research question, "What are students' perceptions of the role and purpose of generic models and scientific models?" is related to the theoretical framework of four types of models - teaching, scientific, mental and expressed models - showing their relationship to learning.

Table 2: The studies investigating the research questions

Objective 1Study
1.1What are students' perceptions of the role and purpose of generic models and scientific models?1, 2, 3, 4
1.2What are the criteria which students identify as being significant when classifying scientific models?1, 2, 3, 4
1.3 How does students' modelling ability affect their use of models and their ability to understand chemical concepts?1, 3, 4
1.4How and why do models help students learn?1, 2, 3, 4
Objective 2
2.1What are students' perce ptions of the role and purpose of chemical representations, including chemical models, teaching models, chemical equations, diagrams, and pictures in learning chemistry?1, 3, (2 & 4 indirectly)*
2.2What are students' understandings of each level of chemical representation in relation to the chemical phenomena they experience?3, (1 & 4 indirectly)**
2.3How does this understanding enable students to effectively transfer from one representational level to another?1, 3, 4
2.4How do the variety of representational forms, which students encounter in chemistry, impact on the epistemology, ontology and social factors that have been shown to contribute to conceptual change?1, 3, 4
Objective 3
3.1What are the factors that influence how and why students learn chemistry?3, 4
3.2What learning strategies do students use in learning chemistry?1, 3, 4
3.3How do learning strategies contribute to the development of students' personal mental models of chemical phenomena?1, 3
3.4How does students' metacognitive awareness influence their learning of chemistry?4

Table 3: Outline of the four studies

Research Questions
for Objective:
Type of StudyData Sources*
and Sample Sizes
1High schoolYear 113-4 weeks1.1
Interpretive research -
descriptive and explanatory
case studies
Questionnaires - Models n=36
MR n=36
VOMMS n=36
Class group activity - audiotapes and videotapes, observations
2High schoolYears 8, 9, 101 week1.1
Quantitative analysisQuestionnaires - SM n=228
SUMS n=228
VOMMS n=228
3UniversityFirst year2
Interpretive research -
descriptive and explanatory
case studies
Initial Questionnaire (containing SUMS, VOMMS and MR) n=18
SUE n=61
Worksheet 1 n=9
Worksheet 2 n=8
Worksheet 3 n=9
Worksheet 4 n=10
1st interviews n=7
2nd interview n=12
4UniversityFirst year2
Interpretive research -
descriptive and explanatory
case studies
Introductory Questionnaire (containing SUMS and VOMMS) n=49, SEEQ n=98
Online Survey n=115
1st interview n=17
2nd interview n=5
* abbreviations include: MR - Molecular Representations; SM - Scientific Models; SUMS - Students' Understanding of Models in Science; VOMMS - My Views of Models and Modelling in Science; SUE - Student Unit Experience; SEEQ - Students' Evaluation of Educational Quality.

Data sources

Direct and indirect data sources

In order to respond to the each research question, various data sources were designed, selected and administered. Direct and indirect data sources were collected throughout the research. The direct sources asked students for their experiences, opinions and their understanding of how their personal learning was proceeding using both qualitative and quantitative methods. The indirect sources looked for examples of how the students were learning in activities, dialogue, tests, worksheets and questions, making inferences and implications from their responses. Both sources of data are valuable and both are needed for validation. As the researcher, I was alert to examples of students saying one thing and actually doing another.

Investigating the learning and understanding of chemistry used direct sources to gain data about students' understanding of their own learning e.g. What strategies do you use in learning chemistry? And also indirect sources e.g worksheet responses to chemistry questions, to investigate how the student is learning.

Quantitative and qualitative data sources

Multiple sources of both quantitative and qualitative data were collected from a variety of the student population. The quantitative data primarily consisting of surveys and questionnaire provided broad general indicators and were quick and convenient to collect even from large sample sizes. The qualitative data consisted of the participant researcher's observations; students' written comments, dialogue between students and interview data. This data took more time and effort to collect but provided greater insight into the students' understanding through the expressions of their opinions, experiences and expressed models of chemical phenomena. The classification of the data sources is shown in Table 4.

Table 4: Classification of data sources

Data sourcesQuantitativeQualitative
Molecular Chemical Representations (MCR)
My Views of Models and Modelling in Science (VOMMS) Scientific Models (SM)
Students' Understanding of Models in Science (SUMS)
Student Unit Experience (SUE)
Students' Evaluation of Educational Quality (SEEQ)
Online survey
What do you think this atom looks like?
1st and 2nd Interview (3)
1st and 2nd Interview (4)
Online survey
IndirectWorksheets 1-4Group work dialogue Researcher's obs
Worksheets 1-4
1st and 2nd Interview (3)
1st and 2nd Interview (4)

It is worthwhile classifying the data sources to gain an appreciation of the range and type of data sources, providing further validation, and identifying gaps in the research design. Being aware of the number of interpretations within each data source is worthwhile. The qualitative data is more generalised and deals with the sample as a group whereas the qualitative data is more individualistic- and may or may not be representative of the whole group

Direct and quantitative data sources

Direct and quantitative data sources include questionnaires and instruments that are suited to a statistical analysis. In my study there were yes/no, a) or b), and Likert-type responses. In these direct sources it the participant who was expressing an opinion, making a judgment, or expressing an understanding and the researcher is reporting what the participant has said or what the group results have indicated. Therefore the data source usually had one interpretation.

Examples 1 and 2

These data sources are well suited to a statistical analysis: percentages, frequencies, standard deviation, Cronbach alpha reliability, factor analysis, scales, ANOVAs, and correlations. This research highlighted the difficulty in formulating an instrument so that it is unambiguous and statistically valid.

Direct and qualitative data sources

Direct and qualitative data sources included interviews written responses observations that require the participants to volunteer their understanding, interpretation, opinions and judgments. It was relatively easy to formulate these straight-forward questions and to understand the results. For example the reason the students gave to support their Yes/No response to Example 1 (above), and written responses to Online Survey questions e.g.
Example 3: Written answers

  1. What aspects of the pre-laboratory exercises are helpful to your learning of chemistry?
  2. What aspects of the pre-laboratory exercises are NOT helpful to your learning of chemistry?
Similarly an interview question asks students to reflect on their learning. Gabby's response to this interview question illustrates two interpretations- firstly Gabby's interpretation and response and then the researcher's interpretation of her response including the words, intonation and intent.

Example 4: Interview

Models are frequently used in chemistry teaching and learning, can you recall any chemical model that you have learnt in your chemistry module?

Gabby: I think it is important for someone to understand the basics. If there were no pictures I would not be able understand anything but now I have pictures in my head. I think the pictures that you get are very important because that is what you remember.

Indirect and quantitative data sources

Indirect data sources such as students' responses to content specific worksheets requiring a quantitative type response provided some insight into their understanding of the content. The students were not graded on the worksheets. The student's response is one interpretation and the interviewers interpretation of the response in terms of the research question is a second interpretation.
Example 5
Example 5
Example 6

Not at all confident
Very confident
Don't know

1.1.35 grams of Na2CO3 contains (0.05x .25x 6.022 x1023) ie 7.53 x 1021 molecules12345X
2.A standard solution has a known number of particles that is indicated by the molarity of the solution.12345X

Indirect and qualitative data sources

Indirect and qualitative data sources included worksheets and interviews. In responding to a worksheet on equilibrium, one student was able to draw a picture of an equilibrium situation at the molecular level and from this the researcher assessed her level of understanding. Her response showed how she was thinking about and understanding the concept.
Example 7

Example 7

Making interpretations from students' responses required the researcher to assess what and how the student was learning. Wading through the chemistry and identifying the way the student is tackling the learning helps to gain an understanding of how the student is thinking. This is a complex process.

Example 8

Interviewer But when you were doing the lab, you probably didn't have a good understanding of what was happening.
MatNo. I didn't know what was happening at the molecular level. But when I sat down after looking at the procedure that we had carried out in the lab, and looking at the questions, and then what we were asked to find out, I could link the experiment to the question. Try to develop some point of view.
InterviewerAll right. So when you did that, how did you go about making that link. Did you actually draw pictures or did you just think about 'well I've got this organic layer and I've got this aqueous layer - Did you do this sort of thing what I've done here?
MatI did have an idea of, I had an idea of the setups, and what was added to each setup. So, in some way, I didn't have the diagram on paper, but I had an idea in my mind that this was what was happening.
InterviewerAnd you definitely understood that the equilibrium constant you measured was for which equilibrium system?
MatThe equilibrium that I measured?
MatWell it was for the iodine moving from the aqueous, the iodine in the aqueous layer reacting with the iodide that was added in the aqueous layer.
InterviewerSo it was for the equilibrium just within the aqueous layer.
MatJust within the aqueous layer. And I think there was some point where we had to subtract the organic layer.

Mat is a very able student and worked at understanding the changes that were occurring at the sub-microscopic level in the experiment. He demonstrated transferring from the macroscopic level to the sub-microscopic level. This analysis requires two interpretations - the student and the researcher.

Sample and sample sizes

Efforts were made to ensure that the volunteer samples were representative of the wider population for that field however there are constraints such as the availability of subjects, time limitations and accessibility to students so their validity in being a true representation of the whole population cannot be guaranteed.

Strength of this research is in the variety of data that the four studies provide. The samples of students are diverse, with different motivations, ages and background knowledge.

The evolution of the data collection

Over the period of the research, there has been a development and evolution of the data sources. The research literature was used as a basis to develop the quantitative instruments used in the first study. These instruments were modified and used in subsequent studies. The trialling and evaluation of the inst ruments has been a dynamic part of the research process involving responding to the results, acting on feedback, eliminating unnecessary or repetitive items. In this iterative process, the number of instruments has reduced and the instruments have become more valid in achieving the desired objective.

Over the period of the research, the emphasis on the quantitative data declined and the emphasis on qualitative data increased. Studies 1 and 2 have substantial quantitative data sources, while in Studies 3 and 4 the quantitative data sources are used to support and complement the qualitative data sources. Similarly, the emphasis on the direct data sources declined and the emphasis on indirect data sources increased. This evolution is typical and expected because the qualitative data provided a good overview of the research topic highlighting some misconceptions or weaknesses and then the quantitative data sources provided opportunities to examine these areas in more detail. These aspects add to the rigour and robustness of the research.

The validity and reliability of data sources

According to Anderson (1997) validity refers to the extent to which what we measure reflects what we expected to measure" (p. 13) and "reliability refers to the consistency in measurements" (p.12). The validity of the analysis has been justified through the validity of each data source, the use of triangulation by using multiple data sources and the crosschecking of any analysis with associates. The validity of particular quantitative instruments was secure in their administration in a serious and significant manner encouraging respondents to provide honest and accurate responses. It is often difficult to ascertain the validity and reliability of qualitative data such as interview transcripts and their analysis. The transcripts are the words but the delivery and intonation, the way it is said can differ and the interpretation can differ. I conducted all the interviews and was present at all classes involved in this research thereby providing consistency. I can only rely on my own interpretation and understanding of the participants' responses. The interviews and analysis required vigilant attention to my personal biases, for example guarding against pre-conceived ideas or leading the participant, in order to minimise the amount of personal bias (Cohen & Manion, 1994).

Research results can be inconclusive or dubious whether the research is rigorous or not, consequently, research methods must be rigorous, appropriate, reliable and valid so that all results are dependable. Without this certainty, the analysis and conclusions are worthless. I focused on the qualitative and quantitative data sources because of the need for triangulation, corroboration and validation without stepping back to make sure that they were achieving this objective. Infiltrates all levels of research methodology: appropriate methods to answer the research question, the analysis of the data; a mix of quantitative /qualitative and direct/indirect data sources.

Interpretive analysis

The interpretive analysis was closely associated with the multiple theoretical frameworks. Multiple levels of analysis were used in the interpretive analysis including the three levels of chemical representations; meaningful learning - considering the aspects contributing to conceptual change or relational learning; the students' mental models; internal factors and external factors contributing to learning; and the perspective of the learning situation. The benefits of closely aligning the research question with the theoretical framework become apparent in the analysis - making the interpretation obvious. For example: in answering the research question "What learning strategies do students use in learning chemistry?" various data sources were selected to provide a fair representation of student behaviour. The frameworks of internal and external factors contributing to learning; meaningful vs. rote learning and the perspective of the learning situation provided structure for the analysis.


In teaching there is content that has to be covered, and in research there are findings that need to be made. But just as significant is the process: in teaching the process of learning is crucial and in the case of research, the methodology of research is most important. If the processes in each case are not set in place properly then the value of the content that is learnt - in the case of teaching - or the findings that are made - in the case of research - are jeopardised.

Focussing on the structure and process of educational research can improve the efficiency and effectiveness of the educational research. More specifically, narrowing and restricting the research by writing very specific research questions can keep the research manageable and achievable. Identifying the relevant theoretical frameworks and writing the research question consistent with these makes the analysis and interpretation easier. Exploring and selecting the most appropriate research methodology will provide direction and a foundation for the research. Having a range of substantiating data sources and being confident in the validity and reliability of those data sources provides credibility to the research. Linking each data source with particular research question should make the research more efficient and effective.


Anderson, G. (1997). Fundamentals of educational research. Hampshire, UK: Falmer Press.

Atherton, J. S. (2003). Two dimensions of practice. [viewed 6 May 2003, verified 9 Aug 2005] http://www.doceo.co.uk/tools/willwit_1.htm

Ausubel, D. P. (1968). Educational psychology: A cognitive view. New York: Holt, Rinehart and Winston, Inc.

Battino, R. (1992). On the importance of rote learning. Journal of Chemical Education, 69(2), 135-137.

Bereiter, C. & Scardamalia, M. (1989). Intentional learning as a goal of instruction. In L. B. Resnick (Ed.), Knowing, learning and instruction: Essays in honor of Robert Glaser (pp. 361-392). Hillsdale, NJ: Lawrence Erlbaum Association.

Cohen, L. & Manion, L. (1994). Research methods in education. New York: Routledge.

Duit, R. & Treagust, D. F. (1998). Learning in science - from behaviourism towards social constructivism and beyond. In B. J. Fraser & K. G. Tobin (Eds.), International handbook of science education (Vol. 1, pp. 3-25). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Hennessey, M. G. (2003). Metacognitive aspects of students' reflective discourse: Implications for intentional conceptual change teaching and learning. In G. M. Sinatra & P. R. Pintrich (Eds.), Intentional conceptual change. Mahwah, NJ: Lawrence Erlbaum Associates.

Mason, D. S., Shell, D. F. & Crawley, F. E. (1997). Differences in problem-solving by non-science majors in introductory chemistry on paired algorithmic-conceptual problems. Journal of Research in Science Teaching, 34(9), 905-923.

Ramsden, P. (1992). Learning to teach in higher education. London: Routledge.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82(3), 498-504.

Skemp, R. R. (1976). Relational understanding and instrumental understanding. Mathematics Teaching, 77, 20-26.

Stephens, S., McRobbie, C. J. & Lucas, K. B. (1999). Model-based reasoning in a year 10 classroom. Research in Science Education, 29, 189-208.

Wandersee, J. H. & Griffard, P. B. (2002). The history of chemistry: Potential and actual contributions to chemical education. In J. K. Gilbert, O. De Jong, R. Justi, D. F. Treagust & J. H. Van Driel (Eds.), Chemical education: Towards research-based practice (Vol. 17, pp. 29-46). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Yager, R. E. (1991). The constructivist learning model: Towards real reform in science education. The Science Teacher, 58(6), 52-57.

Author: Gail Chittleborough, Science and Mathematics Education Centre, Curtin University of Technology, GPO Box U1987, Perth WA 6845. Email: G.Chittleborough@curtin.edu.au

Please cite as: Chittleborough, G. (2005). Conclusions about the efficiencies and effectiveness of educational research methodologies in investigating how students learn. Proceedings Western Australian Institute for Educational Research Forum 2005. http://www.waier.org.au/forums/2005/chittleborough.html

[ Abstract for this presentation ] [ Proceedings Contents ] [ Schedule ] [ Abstracts ]
Created 9 Aug 2005. Last revised 18 May 2006. URL: http://www.waier.org.au/forums/2005/chittleborough.html
The Forum Proceedings are © Western Australian Institute for Educational Research. However
the copyright for each individual article remains with the authors of the article.
HTML: Roger Atkinson and Clare McBeath