Interpreting Heteroskedasticity from a Scatterplot Graph

I am currently doing my MBA thesis. I work at a university and I want to know what factors from the 7Ps of marketing (product, place, price, promotion, people, process and physical evidence) influence student choice of university. The dependent Y variable is new student intake for the academic year 2011/12. I am analyzing students who enrolled at the university for intake 2011/12 and students who passed the university entrance test but did not enroll for intake 2011/12.

1. I have used SPSS 20 to compute the scatterplot graph but I cannot interpret the results. First, there are two parallel perfectly negative lines about 1 and half inches apart on the SPSS scatterplot graph. I presume the two lines represent the two halves of the population: those who study at my university(1) and those who passed the entrance test but did not enroll (0). How can I interpret the scatterplot related to my research? I have never seen a double negative line in a scatterplot, let alone two totally parallel and totally negatively perfect lines!

2. I have performed a Durbin-Watson test and the value is 0.542, which indicates auto-correlation. What exactly does this mean for my particular research? How can I relate this to the 7Ps model and new student intake?

I do not have a great understanding of statistics so I would appreciate a VERY simple and easy to understand answer for my questions. Thanks.

James.