Problems with specification of ordered probit model

[thommy84]

Dear all,

I'm currently writing my diploma thesis about demographics of inflation perceptions and expectations.

As perceptions and expectations of inflation are not published in quantitative data in Europe, but only in qualitative data (risen a lot, risen moderately, risen slightly, stayed about the same, fallen) i figured to introduce an ordered probit model.

I want to model the influence of various socio-economic factors on this qualitative inflation perception. The available data includes percentages for the different answer categories for total perceptions, different age groups, different educational groups and so on. As I'm not used to programming in EViews i would be glad if someone could point out how to proceed from this point.
At first, is the information from the data as described sufficient for an ordered probit model?

So far i would conclude, that i'd have do define my dependent variable as p(yi =x) with i=1,2,3,4,5, but i don't really get how to put the exact equation.

Thanks in advance for any help