Ridge has the comm, so I won't add much in terms of the effect of parallax on POI vs. 'perceived POA' except for a couple of pictures I put together a few years ago to illustrate the effect of parallax setting error.
First off, the parallax setting on your scope is a control that focuses the crosshair image in the same plane as the target image as light passes through the scope. If the parallax is set properly, no matter how the shooter's eye is positioned, the images of the crosshair and the target will stay locked together, as depicted that the image inverts perfectly in the plane of the reticle (top case). If the parallax is incorrect, but the shooter has PERFECT eye alignment, it will still make the hit (center case). If the shooter has a parallax error and 'less than perfect' eye position, the crosshairs will appear to drift proportionately around the target. The bottom case shows how if the eye isn't aligned, even if the scope IS on target, the crosshairs won't appear on target, so the shooter would move the crosshairs to center on target, but they'd be inadvertently moving the rifle OFF of target.
(This isn't exactly correct, but for all intents and purposes, the theory is there).
To illustrate this, I took these photos below. This scope is set up, intentionally, with parallax errors and imperfect "eye" position, at 10yrds. The red dot at the bottom at the bottom of the target is a laser boresighter, intended to show that the rifle was not moving in the rest, even though the crosshairs clearly move around the target:
Without that parallax error, the crosshairs would not move around the target like they do in this series. I've DRASTICALLY exaggerated the extent of the eye position misalignment to better show the drift, as no shooter would ever fire with the 'moon shadows' in their scope like these examples, BUT, smaller misalignment errors aren't so obvious, and still have dramatic effects on the relationship between POI, POA, and PERCEIVED POA.
You can see at +/- 1/2" drift in these photos, only at 10yrds. At 100yds, that's an extra 10" of error in your groups, at 200yrds, that's an extra 20" or more.
Of course, that's EXTREME misalignment of the eye position, but if you shoot with even 10-25% of that degree of error in your eye position, that could yield 2-4" of additional error in a 200yrd group, even though the rifle isn't moving and the shooter's fundamentals aren't changing.
As has been said, when you change the magnification, ALL scopes have a drift in parallax, some more than others. All it takes is a quick second of shifting your eye side to side, top to bottom, and verifying that the crosshairs don't drift on target, and a second to ensure that the crosshairs AND target show up perfectly focused together.