OK since my last thread utterly failed I'll just remake it. Basically, first thing, is conditional means on the course? And secondly, can someone explain how it works? I don't completely get it... What I came up with is:
=\frac{E(X \cap (X>k))}{Pr(X>k)})
=\frac{E(X>k))}{Pr(X>k)})
And then E(X>k) is solved by integrating k to infinity (or whatever the upper bound of f( x) is) for
?
Lets say you had the randomly distributed variable X which follows a PDF function
 = e^{-x})
for

which has the property
 = 1)
.
If you know that

, then first of all, find the area under the curve of
)
from

to

.
It turns out that

, so now define a new random variable Y which follows a PDF function with equation:
 = e^{-x} \cdot e^{2} = e^{2 - x})
for

. We multiply by the reciprocal of the above area to ensure that the area under the NEW curve is equal to 1. We can see that
 = 3)
.
So in this case, we have
 = 3)
.
In general, this process can be summarised as:

It seems latex is down at the moment, hopefully that will be fixed soon.