Uncertainty principle is about shape, not measurement. Unless you mean "how you define the measured aspect of the system" .
Position and momentum are Fourier transforms of each other, and therefore their waveforms cannot both be confined to small range at the same time.
It is loosely analogous to the concept that a rectangle of area 1 can never have both side lengths less than 1.
The observer effect is well-understood, there's nothing magical about it. For example, a multimeter acts a (very small) resistor, affecting the circuit it's measuring. Sometimes the error you introduce can be ignored, but get unmanageable when you try to measure really small systems, where things are more easily disturbed.
The uncertainty principle, on the other hand, is something else. There's a video that I think explains the uncertainty principle well, with a practical example (it's explaining why there's a limit to how many pulses you can send down a fiber optic and still be able to measure on the other side): http://youtu.be/0OOmSyaoAt0?t=7m28s