114
3 eThekwini’s Green and Ecological Infrastructure Policy Landscape
change adaptation, an ecological infrastructure places the concept of ‘water security’ at centre stage. The uMngeni and Palmiet Rivers are, in this sense, important
ecological infrastructure components that link them, ontologically, to water security.
3.4.2.1 Matrices
To discuss the ‘appearance’ of the theories in the interviews, I needed a practical and
communicable framework that would accomplish the ‘…scientific, in the positivist’s
sense of the word, and aim towards an interpretive understanding, in the best sense
of that term’ (Miles and Huberman 1984: 21). With regard to communication, I
realised during my analysis that the data contained in the interviews were comprised
of the assumptions of more than one theory, and that they were, in some instances,
simultaneous. This indicated that the data cannot be neatly compartmentalised into
each of the identified theories (Table 2.1). The data were more fluid during the analysis
than was the case in the graphic presentation of the identified theories (Fig. 3.16).
This dynamism called for me to pioneer a way of displaying the data from which I
would be able to draw conclusions.
Since the propositions of the perspectives appeared more than once during the
interviews, I had a practical problem, namely, how I should display the data without
repeating lengthy quotes throughout the text? What is more, I had already written
down elaborate quotes during the paradigm analysis above.
To avoid repeating quotes, I decided to utilise Microsoft Word’s cross-reference
function to refer to either the entire quote, or parts thereof. I will use this practical
method in the rest of the analysis.
With regard to the analysis of PULSE
3 ’s repertoire of theories (p. 61), as well
as the interpretation of the causal mechanisms (p. 58) and the problem-solving and
critical theories (p. 63), I will rely on tables or matrices to present the data. Miles
and Huberman (1984: 26) describe these tables as a ‘descriptive matrix’, which they
found ‘uncommonly fruitful’. My matrices will be in the form of a conceptually
clustered-matrix that brings together the variables that are connected by the theoretical assumptions (Miles and Huberman 1984). Before I proceed with my theoretical
conceptually clustered matrices, I will share a few thoughts on the development and
use of matrices.
According to Miles and Huberman (1984: 26), ‘Matrix formulation is, in our experience, a simple, enjoyable and creative process, as anyone who has ever constructed
a dummy table knows. It is also decisive: matrix formats set boundaries on the type
of conclusions that can be drawn’. Since I am dealing with several theories, theory
classes and causal mechanisms, these boundaries will come in handy for organising
the densely packed data contained in the numerous interview quotes.
Furthermore, Miles and Huberman (1984: 26) note that: ‘There is a catch, of
course [in using matrices], because one is limited to the data in the display. Thus,
a great deal depends on the core [data] that have been selected from the field notes
[interviews and photographs], and how far they are aggregated or abstracted’. It is for
this reason that I will rely on Microsoft Word’s cross-reference function to confirm
3 eThekwini’s Green and Ecological Infrastructure Policy Landscape
change adaptation, an ecological infrastructure places the concept of ‘water security’ at centre stage. The uMngeni and Palmiet Rivers are, in this sense, important
ecological infrastructure components that link them, ontologically, to water security.
3.4.2.1 Matrices
To discuss the ‘appearance’ of the theories in the interviews, I needed a practical and
communicable framework that would accomplish the ‘…scientific, in the positivist’s
sense of the word, and aim towards an interpretive understanding, in the best sense
of that term’ (Miles and Huberman 1984: 21). With regard to communication, I
realised during my analysis that the data contained in the interviews were comprised
of the assumptions of more than one theory, and that they were, in some instances,
simultaneous. This indicated that the data cannot be neatly compartmentalised into
each of the identified theories (Table 2.1). The data were more fluid during the analysis
than was the case in the graphic presentation of the identified theories (Fig. 3.16).
This dynamism called for me to pioneer a way of displaying the data from which I
would be able to draw conclusions.
Since the propositions of the perspectives appeared more than once during the
interviews, I had a practical problem, namely, how I should display the data without
repeating lengthy quotes throughout the text? What is more, I had already written
down elaborate quotes during the paradigm analysis above.
To avoid repeating quotes, I decided to utilise Microsoft Word’s cross-reference
function to refer to either the entire quote, or parts thereof. I will use this practical
method in the rest of the analysis.
With regard to the analysis of PULSE
3 ’s repertoire of theories (p. 61), as well
as the interpretation of the causal mechanisms (p. 58) and the problem-solving and
critical theories (p. 63), I will rely on tables or matrices to present the data. Miles
and Huberman (1984: 26) describe these tables as a ‘descriptive matrix’, which they
found ‘uncommonly fruitful’. My matrices will be in the form of a conceptually
clustered-matrix that brings together the variables that are connected by the theoretical assumptions (Miles and Huberman 1984). Before I proceed with my theoretical
conceptually clustered matrices, I will share a few thoughts on the development and
use of matrices.
According to Miles and Huberman (1984: 26), ‘Matrix formulation is, in our experience, a simple, enjoyable and creative process, as anyone who has ever constructed
a dummy table knows. It is also decisive: matrix formats set boundaries on the type
of conclusions that can be drawn’. Since I am dealing with several theories, theory
classes and causal mechanisms, these boundaries will come in handy for organising
the densely packed data contained in the numerous interview quotes.
Furthermore, Miles and Huberman (1984: 26) note that: ‘There is a catch, of
course [in using matrices], because one is limited to the data in the display. Thus,
a great deal depends on the core [data] that have been selected from the field notes
[interviews and photographs], and how far they are aggregated or abstracted’. It is for
this reason that I will rely on Microsoft Word’s cross-reference function to confirm