the loading effect
The loading effect is the unavoidable truth that the moment you connect something to a circuit, whether a meter to measure it or a next stage to use its signal, you change the very thing you were after. Picture dipping a measuring cup into a shallow stream: the act of scooping lowers the water level you were trying to read. Connecting a circuit draws some current, and drawing current from a source with real resistance pulls its voltage down.
Here is a concrete sting. A 10 kohm over 10 kohm divider on 9 V should give 4.5 V at the tap with nothing attached. But connect a load of 10 kohm to that tap and you have just put a third 10 kohm in the picture: the lower leg becomes 10 kohm parallel 10 kohm, which is 5 kohm, so the output collapses to 9 times 5/(10 + 5) = 3 V instead of 4.5 V. A cheap voltmeter with only 10 kohm of input resistance loads a high-resistance node the same way, reporting a voltage well below the true one.
Loading is everywhere and it is why high input resistance, buffers, and voltage followers exist. The cure is to make whatever you connect draw negligible current compared with the circuit it taps: use a meter with megohms of input resistance, or insert a buffer that presents a huge input resistance and a tiny output resistance between a delicate node and its load. Forgetting loading is one of the most common reasons a measurement or a design quietly reads wrong.
Probing a 1 Mohm node with a meter of 1 Mohm input resistance halves the reading to 50 percent of the true voltage. The same node read with a 10 Mohm meter barely moves, off by under 10 percent.
Connecting anything draws current and droops the node; the cure is high input resistance.
A measurement is never free: the meter or next stage always loads the circuit. The fix is high input resistance, or a buffer between the delicate node and its load.