You are correct that in quantum mechanics, the act of observation or measurement typically involves interaction with the system being observed, which can lead to a change in the system's state. This phenomenon is known as the "measurement problem" or the "observer effect."
In the double-slit experiment, when a measurement is made to determine which slit the particle passes through, the act of measurement disturbs the system and causes the wave function to collapse into a specific state. This collapse eliminates the interference pattern and leads to the particle being observed as if it had behaved as a classical particle, showing a particle-like behavior with a definite path through one of the slits.
However, it is important to note that when physicists talk about observing the double-slit experiment, they typically refer to detecting the particle at the screen to observe the resulting pattern. In this case, the observation does not involve determining which slit the particle went through. The interference pattern can still be observed without disturbing the system significantly.
The reason physicists don't claim that the act of observation changes the results of the interference pattern is that the disturbance caused by the detection at the screen does not disrupt the interference. The measurement of the particle's position at the screen does not provide information about which slit the particle passed through. The interference pattern arises from the superposition of the particle's wave-like behavior passing through both slits, and this behavior is preserved even when the particle is detected at the screen.
So, while the act of observation or measurement does generally have an impact on quantum systems, in the case of the double-slit experiment, observing the interference pattern on the screen does not significantly alter the overall behavior of the system. It is the specific measurement of which slit the particle went through that disturbs the system and leads to different results.