A mixed methods design was seen as instrumental in providing comprehensive evidence with regard to the relationship between self-concept in academic writing and engagement in the AW class. Incorporating both qualitative and quantitative approaches was intended to provide various types of data, thus giving the research the rigor, and also quantitative breadth and qualitative depth. This combination increases the confidence in and the value of the data in addressing the research questions and research objectives, which is an important consideration in mixed methods research (Creswell & Plano Clark, 2007; Sammons, 2010).
In this study, it was also important to capture the broader contextual issues regarding Malaysian learners’ self-concept in academic writing and the subtle contextual
62 academic writing and their engagement in the Academic Writing class. Three key influences on the design of a mixed methods study noted in the literature: priority, implementation and integration as highlighted in Ivankova, Creswell and Stick (2006) will now be discussed.
Priority
According to Ivankova and colleagues (2006), “priority refers to which approach, quantitative or qualitative (or both), a researcher gives more weight or attention
throughout the data collection and analysis process in the study” (p. 9). In this study,
priority was given to the qualitative data collection and analysis because the purpose was to not only identify, but also to understand the relationship between self-concept in academic writing and student engagement in the Malaysian tertiary context. The first quantitative phase focused primarily on identifying the patterns and trends of self- concept and student engagement. Although this phase was crucial, it was unable to contribute greatly to the ‘why’ elements sought in the second research question. Thus, to obtain rich and in-depth description of the relationship between self-concept in academic writing and the influences for each construct necessitated that priority be given to the qualitative phase.
Implementation
“Implementation refers to whether the quantitative and qualitative data collection and analysis come in sequence, one following another, or concurrently (Green et al. 1989;
Morgan 1998; Creswell et al. 2003)” (as cited in Ivankova, et al., 2006, p. 5). In this study, there were two consecutive phases.
The mixed methods design in this study is known as sequential mixed methods design (Tashakkori & Teddlie, 2003). Within this design, one type of data informs the collection of another type of data in a subsequent stage (Mertens, 2010). The sequence of the quantitative and qualitative data collection was determined by the study purpose and research questions. In this study, the researcher first collected and analysed the quantitative data. The next phase then involved the collection and analysis of qualitative data. Figure 3.1 illustrates the sequential progression of the study.
63 ______________________________________________________________________
______________________________________________________________________
Figure 3.1. Designing and conducting mixed methods research. Adapted from Designing and Conducting Mixed Methods Research (p. 69), by J. Creswell, & V. Plano Clark (Eds.), 2010. London: Sage. Copyright 2010 by Sage Publications.
The first phase quantitative data aimed to provide statistically significant findings regarding self-concept and engagement in academic writing. In particular, the correlational analysis between self-concept in academic writing and engagement also helped in determining the level of the relationship between the two constructs. The qualitative data was intended to further explain the nature of this relationship and the antecedents which had a bearing on the direction and magnitude of the relationship. In other words, the quantitative data and statistical results were expected to provide a general understanding of the issue at hand and the qualitative data and its analysis would then explain the meaning of the relationship for individual students.
In the second qualitative phase, student case studies (Stake, 2000) were utilised. Students who represented a range of perspectives from Phase One were selected for Phase Two case studies. Nunan (1992) maintained that case studies are advantageous as
a great deal can be learnt “in general from a detailed study of one particular student” (p.
89). Case studies have also been argued to provide rich insights through depth of analysis (Gerring, 2007). In this study, the detailed data in the case studies were expected to reveal salient influences on the nature of the relationships between self- concept in academic writing and engagement, and insights into the outcomes of such influences. In particular, each case study might reveal how the interaction between a
student’s self-concept and engagement are experienced and enacted at different levels (personal and social) within the AW class context. For this reason, a case study
approach in Phase Two was chosen to “enhance our capacities for social explanation
and generalisation” within the mixed-methods study (Mason, 2006, p. 11). In this way,
accounts of the learners’ self-concepts in academic writing could be meaningfully connected to their actions and engagement in the classroom.
Follow up with Quantitative data collection and analysis Qualitative data collection and analysis Interpretation
64
Integration
Integration of data occurred in the final stages of interpretation and explanation of results. At this point, findings from both the quantitative and qualitative stage were integrated to ascertain whether commonalities or differences exist to further enhance the understanding of the topic. This integration into a coherent discussion in the final stages of the study was intended to “provide stronger evidence for a conclusion through convergence ... of findings” (Johnson & Onwuegbuzie, 2004, p. 21). Further discussion of the data analysis procedures used in each of the quantitative and qualitative phases of the study is provided in Chapter Four.
Figure 3.2 provides a visual model of the sequential investigative procedures for this study. This shows the two phases with the respective procedures of data collection and data analysis. This visual model also elucidates the nature of data produced for each phase. These matters will be further discussed in Chapter Four which follows.
65
Phase Procedure Product
Cross-sectional questionnaire Numeric data Descriptive analysis Grouped frequencies Factor analysis Correlational analysis SPSS 16 software Descriptive statistics Factor loadings Correlational coefficient Purposeful selection of participants with a range of self-concepts
Development of interview questions
Cases
Interview protocol
Individual in-depth semi- structured interviews conducted
Text data (interview transcripts, and researcher’s notes) Image data (interviewees’ illustrations) Coding and thematic
analysis
Within-case and across-case theme development
Cross-thematic analysis NVivo 8 software
Visual model of multiple case analysis
Codes and themes Similar and different themes categories
Interpretation and explanation of the
quantitative and qualitative results
Findings Discussion Implications
Figure 3.2. Visual Model for the Sequential Mixed Methods Procedures. Adapted from
“Using Mixed-Methods Sequential Explanatory Design: From Theory to practice” by N. V. Ivankova, et al., 2006, Field Methods, 18, p. 16. Copyright 2006 by Sage
Publications. Phase One Quantitative Data Collection Quantitative Data Analysis Connecting Qualitative and Quantitative Phase Two QUALITATIVE Data Collection QUALITATIVE Data Analysis Integration of the Quantitative and Qualitative Results
66
3.4
CHAPTER SUMMARY
In this chapter, the researcher has provided a rationale for and discussed the use of mixed methods as the methodology of the study. The choice of this methodological approach was justified on the basis of the purpose of the research, and the notion that the strengths of both quantitative and qualitative approaches help to answer the research questions. The selection of a mixed methods design was a reflection of the pragmatic philosophical stance taken in the study design. The mixed methods design offers a way of advancing knowledge regarding the relationship between Malaysian learners’ self- concept in academic writing and their engagement in the AW class. In doing so, it was ascertained that neither quantitative nor qualitative research alone would suffice to capture the trends and details of the relationship between Malaysian learners’ self- concept in academic writing and their engagement in the AW class. The mixed methods design allowed for multiple methods to be used and was chosen due to the exploratory and descriptive nature of the study. The practical implementation of this approach is discussed in the next chapter.
67
CHAPTER FOUR
METHODS
4.1
INTRODUCTION
This chapter focuses on practical issues regarding methods and procedures in light of the research aims and the research questions the study seeks to address. The research process is described, including the design, and procedure for data collection and data analysis undertaken in the study. Finally, there is an examination of the ethical principles applied in implementing the sequential mixed methods design. Following the introduction, Section 4.2 discusses research procedures in Phase One of the study. Section 4.3 discusses research procedures in Phase Two. In both phases, information on rationales for the chosen methods, instrument design, pilot study procedure, data collection and data analysis procedure is provided. Finally, in Section 4.4 there is a discussion of the ethical considerations in relation to each phase of the study. The schedule of data collection and analysis for Phase One and Two is shown in Figure 4.1.
Figure 4.1. Schedule of data collection and analysis for Phase One and Two
1/02/09 23/03/09 12/05/09 1/07/09 20/08/09 9/10/09 28/11/09 Questionnaire Construction Questionnaire Pilot Questionnaire Questionnaire Analysis Interview Construction Interview Pilot Interview Interview analysis PHA SE ONE PHASE TWO
68 Phase One began with the development of questionnaire items in February 2009. A pilot was carried out on 23rd March 2009 and the questionnaires were finally distributed from the 2nd to 10th April 2009. The analysis of the questionnaires took approximately two months and this overlapped with the development of interview items for Phase Two. Similarly, a pilot was conducted for Phase Two in mid-July 2009 and the interviews were conducted from 3rd to 13th August 2009. The interview analysis for Phase Two took approximately four months.