Apologies if the title is confusing, allow me to clarify.
Say you have a normal distribution with µ = 50 and σ = 25. A random variable drawn from this distribution would of course have a mean of 50.
But what if this random variable is arbitrarily truncated to certain bounds? For example, if a random value drawn from this distribution were less than zero, it would instead become zero, and likewise, if a random value drawn from this distribution were greater than 75, it would instead become 75. In this scenario, the probability of obtaining a value of exactly zero is
$$\int_{-\infty}^0\frac1{25\sqrt{2\pi}}e^{-\frac12(\frac{x-50}{25})^2}\approx0.0227501$$
and the probability of obtaining a value of exactly 75 is
$$\int_{75}^\infty\frac1{25\sqrt{2\pi}}e^{-\frac12(\frac{x-50}{25})^2}\approx0.158655$$
Is it possible to determine the mean value of a random variable drawn from a distribution when it is truncated in such a way?
Note that I do not mean a truncated normal distribution, in which there is a 0% chance of selecting a value outside of the truncation range, and the probability of all other values is scaled uniformly to maintain a total integral of 1.