Chapter 3: Research Methodology. "Action Research in Mathematics: Providing Metacognitive Support (As a Heutagogical Technique) to Grade 3 Students"
Author: Zoriana Myburgh
3. RESEARCH METHODOLOGY
This study aims to explore the perceptions of students on the heutagogical framework while teaching maths at an elementary school in Cambodia. The study focuses on some metacognitive elements that can assist students to become more self-determined.
Central Question: How can students’ needs be met using the heutagogical framework while teaching maths in an elementary school in Cambodia?
Subquestion 1: What tools and strategies should teachers use to implement the heutagogical framework?
Subquestion 2: How can metacognition as a heutagogical technique be used to improve students’ self-determination?
This chapter aims to explain the research design and methods that were selected as appropriate to research the issue related to the research questions.
3.1 Research design
A lack of qualitative action research in existing literature highlighted the need for this investigation into self-determination through metacognition at a young age. Compared to traditional experimental studies, action research is an amalgam of theory and practice (research and action) that “focuses on specific situations and localized solutions” (Stringer, 2007, p. 1). It is a compass leading teacher in the right direction and helping them see changes in their practice (Simon and Wilder, 2018). Action research is “grounded in a qualitative research paradigm whose purpose is to gain greater clarity and understanding of a question” (Stringer, 2007, p. 19).
Action research was promoted in the mid-1940s with the purpose to solve some practical problems in everyday life. The goal was to distinguish the problem, try to change the situation, and check the results (Coolican, 2014). Since the main purpose of action research is to improve the practice of education by studying issues or problems (Creswell, 2012), it was decided to choose this research design to align with the purpose of the study. Self-determination is like “mini action research”. It also has a cyclical model: setting goals, attempting to achieve them, self-assessing and making adjustments (Zimmerman, 2002). Metacognition itself has a cyclical model too: planning, thinking about it and making changes if needed, reflecting, and making new plans based on the results (Costa and Kallick, 2008). Thus, it was the other reason for choosing action research as a research design for this investigation.
The main attribute of almost all action research models is the cycles, specifically that each cycle is based on the conclusions of the previous cycle (Edwards and Willis, 2014) which “guides teacher preparation and instruction” (Stringer, Christensen and Baldwin, 2010, p. 1). Sometimes they are named spirals or helices (Punch and Oancea, 2014). Edwards and Willis’ (2014) model starts with reflection: Reflect – Plan – Act – Observe, while Stringer’s (2010) model commences with the observation cycle: Look – Think – Act. Even though it seems straightforward, it is not as simple as it looks. There are no arrows between them so it is not a linear process and the researchers can go back to any cycle when changes or adjustments are needed. The “Act” stage also includes reflection and evaluation. Reflection is the key characteristic of each cycle, and the results indicate whether changes should happen or an additional cycle or “mini-experiment in practice” is needed (Wagaba,
Treagust and Chandrasegaran, 2016, p. 5378). Since its purpose is not just understanding the problem but finding the solution, it is a suitable form of research for this investigation. Moreover, Nasrollahi (2015, p. 18667) noticed that Stringer’s model does not only involve teachers but also students “as action researchers collaborating in the action research process”. It is a standard model that has been similar to hundreds of other models created in the last eighty years.
The present action research will include these research tools: interviews with students and observations using the researcher journal. The research questions that require qualitative data to investigate students’ views justified the qualitative approach of this study. Interviews (qualitative methods) show the complexity of the data provided by participants (Creswell, 2012). At the same time, some quantitative methods, such as questionnaires, can be “inappropriate because of the child’s age” (Morison et al., 2000, p. 113). Qualitative methods and students’ involvement can provide more light on heutagogy in elementary schools and explore any misconceptions.
3.2 Setting and participants
Metacognition is one of the sixteen Habits of Mind (Costa and Kallick, 2008) adopted by the participating school in Cambodia. Thus, it provided an ideal setting to apply action research on metacognitive support in mathematics. Secondly, being a subject coordinator allowed the researcher to take action and adapt the math curriculum using metacognition as a heutagogical tool. Stringer (2007) underlines that conducting action research helps in curriculum construction and evaluation. Finally, as a participant-researcher, the researcher had a chance to work closely with the participants and gather data during math classes.
Previous research in self-determination in elementary school is mostly from the teachers’ (Stein, 2018) or school administrators’ points of view (Akyildiz, 2019) which indicates the need for the present study. Earlier, children were treated as “dependent” on others to guide them on what to do (Elden, 2013, p. 78). Later, Arnold and Triki (2017) argued that children might be participants in experimental research, however, the chance exists that students would only say what they think researchers want to hear instead of their honest statements. It is understandable that writing a high-quality study with children may be a noteworthy challenge (Ponizovsky-Bergelson et al., 2019), yet open-ended questions and encouragement can elicit some valuable data. Hoover (2018) recommends using short semi-structured individual interviews while working with young children. Semi-structured interviews “have in-built flexibility to adapt to particular respondents” (Punch and Oancea, 2014, p. 184) which is also preferable for interviews with children. Punch also suggests having natural settings and making sure that the language is age appropriate. All these recommendations were taken into account during this study. Interview questions can be found in Appendix III. Each interview with each student was not longer than 15 minutes, with three interviews during the experiment (the beginning, middle and end of the unit). Interviews were conducted via Google Meet, the platform used at school during online learning.
23 grade three students (9-10 years old) and their parents were informed about this research. 15 of them returned the signed consent forms, 8 parents did not reply to the email sent, so their children were not involved in research. In the end, 7 boys and 8 girls took part in the study.
All the students were Cambodian. Their native language is Khmer (the official language of Cambodia). The interviews were conducted in English as this is the language they use to study at this International School. This language barrier is why math in English might be particularly challenging and metacognitive strategies might be beneficial to them. All interviews were transcribed verbatim by the researcher.
Consent forms were sent on April 28, 2021, as soon as permission was granted by ERGO. Parents were called by the academic assistant from the school who informed them about an email sent. As soon as some consent forms signed by parents were returned, consent forms were sent to their childern. Observations began as soon as both parents and students signed the consent forms. The first interviews were held on the same or following days as soon as the consent forms were returned:
29/04: S2, S3, S13, S14
03/05: S17, S21, S22
04/05: S16, S23
05/05: S5, S7, S8, S12, S19
07/05: S20
The second round of interviews was held on June 17-21. The third round was recorded on June 14-15, 2021.
3.3 Data collection
The experiment lasted 9 weeks starting on April 29 until June 30, 2021. Three cycles of Stringer’s model were incorporated into three phases of instruction (Stringer, 2007, p. 9, fig. 1.1).
PHASE 1: PLANNING
1. LOOK (3 days)
Consent forms are sent to parents and students (Appendices I and II). The first semi-structured interviews with some students are recorded. Field notes are taken every day.
This first stage allows the researcher to collect data about participants’ perspectives and to define the problem.
2. THINK (1 week)
Data are organised, the answers are coded and analysed, and observations continue during this time. Metacognition Diaries (MD) and the final assignment that incorporated metacognition elements are created based on the results of the interviews and on the theory of the process of metacognition which consists of three dimensions – commitment, value, and capacity (Costa and Kallick, 2008). Even though the authors presented five dimensions in their book, and they are discussed in the previous chapter, the core of this study constituted only three dimensions. First, HoMs are included in the school curriculum since kindergarten, it is assumed that by grade 3 most of the students should know the meaning of 16 HoMs, thus, Expanding Meaning as a dimension was not covered by this research. Secondly, considering the age of students and the time constraints it would have been overwhelming for students to reflect properly on all five dimensions at the same time. Since the focus of this paper was self-directedness through metacognition, Increasing Alertness as a dimension was not included in this study.
3. ACT (1 week)
MDs are shared with the students in Google Classroom (Appendices VI, VII, VIII). The final assignment is explained to students (Appendix IV). Before being given to participants, MD and the final assessment were examined by our school curriculum coordinators who supported that both assignments can be used to monitor students’ metacognitive growth.
The Flipgrid platform is introduced to students. Students have a choice to either complete MD in Google Classroom or answer these questions by recording videos in Flipgrid (video diaries). To set the tone of reflection, at the end of each math lesson students have five to ten minutes of “silent thinking time” where they hav a chance to reflect and answer the questions in the diaries or record the videos. Observations continue during this time. Organisational skills are taught by using checklists.
This part of implementing practical solutions distinguishes action research from other types of research. Stringer (2007, p. 142) calls it “the sharp end of the stick”. This is the part where the action happens.
PHASE 2: INSTRUCTION
1. LOOK (1 week)
The second semi-structured interviews are recorded to check what modifications should be made. Observations continue during this time. MDs and Flipgrid Videos are checked by the researcher. Most of the student answers in the diaries are short and not specific. Multiple students do not stay online until the end of the class and do not complete the MDs or do not record the videos.
2. THINK (1 week)
Answers are coded and categories are added or modified. Observations continue during this time. Reasons for students leaving online classes early could be family circumstances, bad Internet connection, no interest in the topic, other distractions. Student answers might be improved by providing more scaffolding and teaching how to set goals, how to monitor learning by using checklists and rubrics, how to take notes, etc. The benefits of reflections should be discussed with the students.
3. ACT (2 weeks)
This cycle takes longer because it is necessary to spend more time on interventions. All math lesson plans are modified and include metacognitive elements (Appendix V), namely goal-setting, self-assessment, peer feedback, HoM Discussion, assessing using a rubric, exit tickets, note-taking, post-assessment. Not all lessons include all these elements at once because of time constraints. It is decided to add metacognitive elements in the beginning and middle of the lessons so that all students have a chance to reflect before they leave the class. Observations continue during this time.
PHASE 3: EVALUATION
1. LOOK (1 week)
The third semi-structured interviews are recorded on June 14-15 and observations continue during this week. All lesson plans continue to have metacognitive elements and students continue to write MDs or record video diaries on Flipgrid.
2. THINK (1 week)
Final interviews are analysed. Student MDs, videos and observation notes are reviewed, final codes and categories are added, the strengths and weaknesses of the experiment are identified. All core categories, abstract concepts and specific indicators were organised in a spreadsheet.
3. ACT (3 days)
General and brief results were discussed with colleagues during the subject meeting at the end of the school year.
These continuous cycles of looking, thinking, and acting allowed the researcher to identify the necessity of metacognition as a heutagogical technique in fostering self-determination in young students.
3.4 Data analysis
Metacognition is not only about planning and knowledge activation but also the intentional monitoring of students’ cognitive processes, reflection, time management, and self-evaluation. That is why the participants frequently had chances to set goals in using HoM 5 Metacognition, self-reflect on whether they achieved those goals, monitor their progress in solving math problems, reflect on their progress using the rubric, apply different strategies in solving word problems and reflect on them. These topics were observed by the researcher and discussed during each of the three interviews. The analysis began after the data was collected from the initial interviews, starting from April 29, 2021, as soon as some parents and students returned the signed consent forms. From that point onwards, data collection (from the following interviews and field notes) and analysis occurred simultaneously. This approach has become a regular practice in qualitative research (Charmaz and Belgrave, 2015).
In order to understand the context, Stringer (2007, p. 100) suggests researchers should start with interviews and then move to other sources of information in the next cycles of action research. Observations written in the research journal were held during all three cycles of action research. The focus was on behaviours related to self-determination, particularly the metacognitive skills related to goal-setting, self-reflection, reflecting and monitoring the progress in solving math word problems, and the emotional aspect after solving problems (feeling of difficulty, feeling of confidence, etc.) . As soon as the student showed some evidence of applying metacognition, it was recorded as a memo or a sentence. Thus, the protocols were kept for each student who signed the consent form. Observations were overt since both students and parents were asked to give their consent before the beginning of the study. The primary data were derived from interviews but observations further clarified or extended understanding of the issue being investigated. Punch (2014) underlines that combining interviews and observations is a good method that can lead to high-quality data.
Students’ interview data, as well as observation notes, were coded and analysed using a grounded theory approach. Creswell (2012) believes that it helps the analysts remain close to the data during the whole process. Codes and themes were developed, and connections were identified between different themes in order to generate conclusions. Based on the systematic design of the grounded theory, there are three types of codes represented diagrammatically by Punch and Oancea (2014, p. 238, fig. 10.4):
a) substantive (or open coding) is done at the beginning of the analysis to generate more abstract concepts,
b) theoretical (or axial coding) to see how data is interconnected, and
c) core (or selective coding) concentrates on core categories around which the theory is designed.
Charmaz (2006) emphasizes that it’s not a linear process and a researcher can go back to the initial data and make new codes at any moment of action research. Moreover, the grounded theory implies that the experiment doesn’t commence with an already formulated theory but rather “allows the theory to emerge from the data” (Strauss and Corbin, 1998, p. 12). Analysts are encouraged to (a) remain open to various opportunities, (b) produce a few options, (c) investigate different opportunities before selecting one, (d) use cycles to go back to the experiment and get a new vision, (e) believe in the study, (f) circumvent shortcuts, (g) enjoy the research (Ezell, 2017, p. 72).
Information was collected, analysed, and compared until data saturation was achieved.
3.5 Ethics and risk assessment procedures
The approval to conduct this research was confirmed through the Ethics and Research Governance Online system of the University of Southampton. In order to start the study, it was obligatory to obtain confirmation. After that, the potential participants and their guardians were invited to take part in the study via email.
First, the Board of Directors of the participating school were informed about the study. Since the children are 9-10 years old, the consent forms were first signed by students’ guardians and after that, if the parents approved, they were sent to students. Both the guardians and the participants agreed that the interviews will be recorded. Confidentiality was guaranteed to all participants. The participants had the right to withdraw from the interviews at any moment before June 30. Personal information was anonymised during the transcription process and coded as Student 2, Student 3 etc. Since February 20, all schools in Cambodia moved to online learning, thus, both interviews and observations took place via Google Meet. Only the researcher had access to the audio-recorded data which were held on a University of Southampton file storage space and were destroyed after the transcriptions were complete. The transcribed interviews and the observation notes were stored on a password-protected University of Southampton file storage space too.
3.6 Validating findings
To avoid an incorrect interpretation of data, multiple methods of qualitative data collection should be used (Oliver-Hoyo and Allen, 2006). The accuracy of the findings was validated through the triangulation analysis:
- interview with students,
- observation notes,
- literature review.
Triangulation of data ensured the chosen key themes, providing an insight into self-determination from the students’ views. Punch (2014) also suggests getting feedback from responders. After the interviews were transcribed verbatim, the participants were asked to check them. Triangulation and member checking are the most common validation techniques (Creswell, 2012).
In the next chapter, the results of qualitative action research will be presented.
icons on the left to see existing comments.
0 General Document comments
0 Sentence and Paragraph comments
0 Image and Video comments
General Document Comments 0
