The effect of non-prepared ecological data on the decision of multivariate analysis: A case study of water quality data of the Shatt Al-Arab River, Iraq

Many ecological studies deal with a multivariate approach in the statistical analysis of the data. Aquatic ecologists advise applying the multivariate techniques to interpret environmental data information through various techniques such as biplots plotting in PCA, RDA, and CCA to achieve this goal. This work examines the importance of the application of the preparation of raw data before performing the statistical techniques in terms of scaling and transforming data and its important impact on the results of the PCA analysis of the water quality. The results showed the effect of data processing on the outputs of the analysis. However, raw data must be presented to evaluate appropriate methods of data processing before applying analysis techniques. The results of the study showed that some data did not show a clear change when converting their raw data, while other variables had a significant and clear effect on improving the normal distribution of values after the conversion of raw data, and this was evident through the Shapiro–Wilk test which was conducted for the variables where the values increased significantly.