Correlation Matrix revisited


Return to Multi-way Page 2

This time, as we did previously for a larger matrix, we will use the routine in SPSS to run Pearson's correlation for water, sanitation, education and the outcome variable waz score. Try this exercise:

1.  Open keast4j.sav

2.  Click on Statistics, Correlate, Bivariate.

3.  Enter the list of variables (waz, dlowedn, notoilet, dpiped, dwell) one-by-one into the variables box using the arrow key.

4.  Click on the box to highlight Pearson's Correlation Coefficient and the box for Two-tailed.

5.  Click on OK.

 

wpeB.jpg (34416 bytes)

INTERPRETATION:

Among the independent variables in the model, it is evident that there is some collinearity.  For example, the Low Education variable has a significant correlation coefficient with toilet access (r = 0.169, p= .000) and piped water (r = -0.152, p= 0.000). Toilet access has a significant correlation with both piped water (r = -0.153, p=0.000) and well water (r = -0.158, p= 0.000). The collinearity which has been detected here will need to be addressed when a model is developed to show the association between water and sanitation with nutrition status.  Return to the main text for some ideas on how to confront this issue.

Return to Top of Page