work, borrowing money from bank, mortgage assets, stopping schooling of children,
reducing expenditures on food and other consumption, selling household assets, and
migrating to another place.
Composite Adaptation Strategies of SBR Villages
Principal component analysis (PCA) was used to develop a composite adaptation
index. The PCA transforms numerous variables into a concise and coherent set of
uncorrelated factors called principal components (Hotelling 1933). The principal
components of each variable account for much of the variance among the series of
actual variables. Each component is a linear weighted combination of the initial
variables. Each component is structured in a way that the first component signifies
the largest possible amount of variation in the actual variables. This method was first
used by Boelhouwer and Stoop (1999) for integrating various indicators into a single
index. It is noteworthy that the UNDP Human Development Index was also modified
by PCA, replacing the simple aggregation procedures (Lai 2003). It is a multivariate
statistical technique used as a data reductionist method that transforms the actual set
of variables into a lesser number of linear varieties by illuminating the underlying
factors (Kaźmierczak and Cavan 2011; Solangaarachchi et al. 2012). PCA has been
effectively used by scholars, statisticians, and planners to prepare socioeconomic
indices during the past few decades (Tata and Schultz 1988; Fotso and Kuate-defo
2005; Rygel et al. 2006; Antony and Rao 2007; Havard et al. 2008; Messer et al.
2008; Krishnan 2010).
Before applying the PCA framework, all the variables were tested and authenticated by qualitative interviews with village Panchayat (representatives elected by
the village assembly) and the Gram Sabha (general assembly of village people)
because variation was seen in the data collected from the households. PCA as a
reductionist technique permitted a strong as well as reliable set of variables for
assessing adaptation strategies in the SBR. A composite adaptation index was
established by using 45 different adaptation strategies opted by the Sundarban
community for extreme climate events (cyclone, flood, etc.), changes in economic
activities (agriculture and fish collecting), and household coping mechanisms. The
PCA was performed to resolve the number of factors to be retained for further
analysis. Correlation analysis and screen plot were performed using these 45 parameters. Finally, factor analysis identified the strategies most often adapted from the
45 factors by the sampled households. After applying the factor analysis, 14 important factors were identified. The rank of these responses was then normalized and
correlated with household socioeconomic characteristics. PCA constructed a composite adaptation strategies index for the sampled villages. Prinscore was calculated
by using following formula:
5 Evaluating Adaptation Strategies to Coastal Multihazards in Sundarban. . .
107
reducing expenditures on food and other consumption, selling household assets, and
migrating to another place.
Composite Adaptation Strategies of SBR Villages
Principal component analysis (PCA) was used to develop a composite adaptation
index. The PCA transforms numerous variables into a concise and coherent set of
uncorrelated factors called principal components (Hotelling 1933). The principal
components of each variable account for much of the variance among the series of
actual variables. Each component is a linear weighted combination of the initial
variables. Each component is structured in a way that the first component signifies
the largest possible amount of variation in the actual variables. This method was first
used by Boelhouwer and Stoop (1999) for integrating various indicators into a single
index. It is noteworthy that the UNDP Human Development Index was also modified
by PCA, replacing the simple aggregation procedures (Lai 2003). It is a multivariate
statistical technique used as a data reductionist method that transforms the actual set
of variables into a lesser number of linear varieties by illuminating the underlying
factors (Kaźmierczak and Cavan 2011; Solangaarachchi et al. 2012). PCA has been
effectively used by scholars, statisticians, and planners to prepare socioeconomic
indices during the past few decades (Tata and Schultz 1988; Fotso and Kuate-defo
2005; Rygel et al. 2006; Antony and Rao 2007; Havard et al. 2008; Messer et al.
2008; Krishnan 2010).
Before applying the PCA framework, all the variables were tested and authenticated by qualitative interviews with village Panchayat (representatives elected by
the village assembly) and the Gram Sabha (general assembly of village people)
because variation was seen in the data collected from the households. PCA as a
reductionist technique permitted a strong as well as reliable set of variables for
assessing adaptation strategies in the SBR. A composite adaptation index was
established by using 45 different adaptation strategies opted by the Sundarban
community for extreme climate events (cyclone, flood, etc.), changes in economic
activities (agriculture and fish collecting), and household coping mechanisms. The
PCA was performed to resolve the number of factors to be retained for further
analysis. Correlation analysis and screen plot were performed using these 45 parameters. Finally, factor analysis identified the strategies most often adapted from the
45 factors by the sampled households. After applying the factor analysis, 14 important factors were identified. The rank of these responses was then normalized and
correlated with household socioeconomic characteristics. PCA constructed a composite adaptation strategies index for the sampled villages. Prinscore was calculated
by using following formula:
5 Evaluating Adaptation Strategies to Coastal Multihazards in Sundarban. . .
107