Descriptive frequencies and bivariate analyses were analysed to test for statistical significance by time period of VPD policy implementation (prepolicy vs postpolicy) and reported using ORs, 95% CIs and p values.
Longitudinal data were analysed by multivariate analyses of variance (MANOVA) from four measurement occasions at 2-year inter-intervals.
Because the organic signatures from glutaraldehyde compromise the Raman analyses naturally only the untreated samples were analysed by Raman.
To test that hypothesis, the spatial and temporal characteristics of the coupled groundwater-stream water system before and after the restoration were analysed by principal component analyses of time series of groundwater heads and stream water levels.
Data were analysed by means of repeated measures analyses of variance and multiple regression analyses.
Sterol contents of algae were expressed on a per carbon basis and differences between algae were analysed by Analyses of Covariance (ANCOVA) with light intensity as continuous variable and medium phosphorus concentration as factor.
Some individual consultation communications were analysed by χ analyses.
Nine gill samples from each condition and sampling point, including the five replicates analysed by microarray, were analysed by qPCR.
Differences in gene expression analysed by qPCR were analysed by t-test and a significant threshold of P < 0.05.
Distribution of the dependent and indendent variables were analysed by frequency analyses.
