The answer to this question comes from the very definition of measurement. In plain words, by definition, when you measure the state of a system you interact with that system and therefore perturb its state. This is known and is very well documented.
Take the double-slit experiment for example:
Your setup is an electron beam that you fire towards a wall with two slits. In front of the wall there is a screen where the impact of every electron is registered. If you do this experiment without determining (measuring) wich slit the electrons go trough, you will get an interference pattern like this one:
This pattern can be explained if the electrons behaved as wave so that you get maximums and minimums depending on wether you get constructive or destructive interference by the superposition of the wave coming trough slit 1 and the one coming trough slit 2.
If you, in the other hand, measure which slit each electron went trough, you will end up with this:
From this experiment we conclude that the mere act of "looking" (measuring) the electrons immensely affects their distribution on the screen. Clearly, their state is modified when one watches them. This is the very quantum mechanical principle that states that measurements interfere with the states of microscopic objects.
It's not possible to determine the state of something without measuring it, but it's not possible to measure the state of something without interacting with it. When we are not measuring the state of a system, you can only say that it is in a superposition of all its possible configurations/states, and it is only when you measure it that your interaction makes the system collapse to one particular configuration/state.