The study area is the Phoenix Metro area comprising of the city of Glendale and Phoenix city. This area has an estimation of around 1.6 million people (U.S Census Bureau).In order to carry out the research, a sample of 400 census tracts were utilized
Type of Data Collection
Methods of data collection
Various methods were applied. To collect data on both cities on water usage involved use of primary data source. It entails obtaining monthly water use from census block while data on socioeconomic variables was also from primary source. From American Communities Survey and Census, both population and socioeconomic variables data was obtained.
Methods of data presentation
Using census tract the acquired water use data for both cities was spatially aggregated, and both converted into total gallons per census tract. From American Communities Survey and Census several variables were selected and represented in a table so as to act as description of a measure
Methods of data Analysis
In order to provide a clear understanding of the relationship that exist between socioeconomic and water use. There was application of the following. A regression analysis, which entails calculating Spearman’s rank for each variable. It was aimed at showing an association that exists between population and water use. The coefficient of determination(r-squared) in an attempt to measure the goodness of fit, as well as a spatial error (SE), , were established. Each variable was tested for autocorrelation which was corrected by SE regression.
Was conducted to measure variables (explanatory and dependent) joint degree of association. The confidence interval was also performed at 5%, and variables non-significant at that level were removed. Before obtaining final equation, there was a test for multicollinearity. Involved pacing variables significant in spatial error into a simple multivariate model.
Principal Component Analysis
Aims at transforming data into a new set of linear weighted components. The data on socioeconomic variables was placed in a matrix linked by census tract identifiers. This led into transformation of the data into components. Each component was analyzed for their explanatory power based on their variable make up. The variables were then ranked according to their power to influence the component.
One of the plans is to explain water usage based on socioeconomic characteristics. Plan to show how the existence of high–income and elderly members explains most variability of water usage in Phoenix area. The group is undertaking plans to provide awareness in decision making by educating public on water use and conservation. In addition, there is a critical need to show the importance of socioeconomic variables (income, household size, and education) in decision making and a rough estimate of how much water is used in certain areas.
Multivariate and simple regression approach, which involve the collection of data on socioeconomic variables are essential. Data on social economic and population size was obtained from census survey and American communities survey (U.S Census Bureau).