According to McCracken (1988), the core of the long interview is the four-stage
method of inquiry which includes the following stages:
Review of analytic categories (Literature Review from Chapter Two)
Review of cultural categories ( Personal Views and Experience and Pilot Studies from Chapter Three and Chapter Four)
Discovery of cultural categories (Data Collection from Chapter Five, see figure 5.1)
Discovery of analytical categories (Data Findings and Analysis from Chapter Six)
5.3.1 Stage One: Review of Analytic Categories (Chapter Two)
The first stage was named by McCracken (1988) as a review of analytic categories
which was based on a review of literature in COO studies and international higher
education. This stage has been completed in the Chapter Two. Gaps in the literature
were identified and initial knowledge was constructed within this stage. The review of
critical idea collection and to aid in the construction of an interview questionnaire for
later stages. It not only provides a framework for the research of COO studies in the
context of service evaluation, but also enables me to place my work in the context of
what has been done in the literature of international higher education allowing
comparisons to be made for the second stage.
5.3.2 Stage Two: Review of Cultural Categories (Chapter Three)
The second stage aims to use myself as an instrument of inquiry by taking advantage
of the intimate acquaintance with my own knowledge and experience (McCracken,
1988) and my close involvement in the evaluation of British higher education.
Therefore, it is concerned with the relationship between my own understanding and
my target study of interest. The review of my personal views and experience becomes
the way of breaking down the complexity of the research that informs me how to
design the interviews for the next stage.
5.3.3 Stage Three: the Discovery of Cultural Categories (Chapter Five)
The third stage can be seen as a data collection process which includes questionnaire
construction and interview procedure. Questionnaire construction in this study is
semi-structured which has three parts including biographical questions, grand tour
questions, and the main questions designed based in the previous stage (i.e. the review
of culture category). The interview procedure continued until the data was sufficient
enough to gain theoretical saturation (Glaser and Strauss 1967). Twenty-three long
interviews along with three focus group interviews were conducted. Section 5.2 of
this chapter will discuss the third stage in details.
5.3.4 Stage Four: the Discovery of Analytic Categories (Chapter Six)
Data analysis is the last stage of McCracken (1988)’s Long Interview Methodology. McCracken (1988) named this stage as Discovery of Analytic Categories. The data
collection from the long interviews and focus groups are the main sources of the information. Although ‘absolute rules’ do not exist for qualitative data analyses, there
are two general guidelines to follow. First, the analytical method must serve to
analyse the research questions generated at the very beginning of the study (Patton,
1990). Secondly, data analysis must contain explanation. According to Geertz (1973)
and Dezin (1989), such data explanation is the first task in qualitative analysis, as the
discipline and rigour of qualitative analysis depends on presenting solid explanatory
data in such a way that others reading the material can understand and draw their own
interpretations.
In accordance with these two guidelines, the presentation of data in this chapter
consists of two parts; explanation and interpretation. This is achieved by directly quoting verbatim extracts from the students’ responses in the interviews then commenting upon these remarks with reference to the extant literature.
Since the interviews were semi-constructed, a cross-case analytical procedure (Patton,
1990) is deemed as appropriate in presenting the data. This procedure requires that
data not be reported on a group-by-group or informant-by-informant basis. Instead,
the answers from different informants are grouped together in terms of common
questions or central issues. Given the nature of the data, it is neither feasible nor
appropriate to present it in a raw and unrefined manner, such as in transcript form.
Instead, it has been subjected to analysis and organisation. To make data analysis
serve the specified research purposes, student responses have been organised into
three parts: COO effects on service evaluation, intrinsic influences of COO effects,
and external influences of COO effects.
Specifically, the data would go through the refined McCracken (1988)’s five steps of analysis (Figure 5.2) advocated by Herrington and Oliver (2000). The first stage refers to the judgement and observation of utterance of the transcript. “The first stage treats each utterance in the interview transcript in its own terms ignoring its relationship to other aspects of the text. (McCracken 1988: 42)” The second stage is the stage of meta-observations. The transcripts would be evaluated by the priori categories which
were identified by the literature review in Chapter Two and my personal views and
experience in Chapter Three. Observations were developed by these priori categories
themselves. In addition, according to the evidence of the data new categories were
fully examined. In the third stage, the draft data matrices were constructed. The relationships among the category were analysed. “The focus of attention has now shifted away from the transcript and toward the observations themselves” McCracken (1988: 42). It must be noted that the data matrices obtained in the third stage were in a
fluid state and represented a first step at organising the data as I begun to develop
isolated observations and their possible connectedness to other observations. This
process was continued in the fourth stage. In addition, the fourth stage examined the
observations in the third stage and refined the data matrices. This can also be seen as a
process of data reduction and organisation. Therefore, the final data matrices were
obtained. The fifth stage reviewed former stages and concluded the themes discovered
from the data.
Figure 5.2: Data Analysis Process (Herrington and Oliver 2000: 13) Stage Four
Judgement of data and analysis, and identification of themes and their interrelationships
Stage Two
Meta-observations where implications and possibilities of the data are examined more fully
Stage One
Judgment of individual utterances with little concern for their larger significance
Stage Five
Review of the above four stages to form conclusions Stage Three