Applying a Multi-pronged Approach to Assessing Adaptive Capacity - Climate Change and Water Governance: Adaptive Capacity in Chile and Switzerland - page 89

and act as an important reliability check (Miles and Huberman 1994 ) when there is
just one researcher working on a case study (as in this topic), as opposed to multiple
coders working in a research team.
However, there are studies that have shown that while multiple coders can add
reliability; this may not be true with 'data that require coders to evaluate the impor-
tance of the information in the context of a larger response' (Crittenden and Hill
1971, in Engle 2010 ), such as complex processes across multiple scales (e.g. gover-
nance mechanisms for adaptive capacity). In such cases, multiple rounds of coding
may provide better results for improving reliability (Engle 2010 ; Perreault and
Leigh 1989 ). Therefore, an internal coding check was performed through the mul-
tiple rounds of coding (coding and recoding the same texts), but an inter-coder
check with a researcher familiar with the research topic was also performed.
The inter-coder reliability check took place with another researcher, also studying
water resources management. Due to the number of codes and the depth of codes, it
was considered inappropriate to conduct a traditional form of inter-coder reliability
test, as it would have taken too long for the researcher to become familiar with the
codes and their detailed descriptions in the memos. Therefore, the researcher read
through a section of coded passages and assessed the relevance and appropriateness
of the codes assigned to those sections, and asked for clarifications where needed.
The result was positive with a fairly high agreement and also provided useful
insights into where some of the codes may have been duplicated or needed to be
re-organised within the coding hierarchy. The final issue in coding methods is to
know when enough is enough. For the purposes of this project, the advice of Miles
and Huberman ( 1994 , p 62) was followed, who suggest that this stage of analysis is
complete when the 'analysis appears to have run its course, incidents readily
classified and categories are saturated, and regularities emerge' .
Having refined the core codes and variables that arose out of the interview pro-
cess, the next process was to complete the investigation of the linkages between
governance context across the different cases and scales (independent variables)
and adaptive capacity (dependent variables). The analytical process relied mostly
on qualitative data gathered through the interview process, but also drew from archi-
val data to construct the case and attempt to answer the key questions at the start of
this chapter. In order to construct the case, there were four main steps:
1. Characterising adaptive responses and outcomes according to different forms of
adaptation (Transformation, Persistent Adaptation, Passive)
2. Correlating categorised adaptation actions (Transformation, Persistent Adaptation,
Passive) with different governance mechanisms and contexts
3. Identifying bridges and barriers to mobilising adaptive capacity in each case area
and with regards to governance mechanisms across different scales of governance
4. Deeper characterisation of governance indicators according to analytical steps
a b o v e
The different steps comprised the multi-pronged approach to understanding and
assessing adaptive capacity as detailed in Fig. 5.3 . The starting set of governance
determinants of adaptive capacity were used to frame the exploration of adaptive