A raster image inside a vector-based program has a squishy-squashy size. The file will have the same 5300x3500px, but you can shrink it or enlarge it. It will retain the same 5300x3500 pixels regardless of this.
The point that can determine the initial size is the declared resolution. If your image has a declared resolution of 300 PPI it will fit that 5300 px in 17.66 inches. If it has a declared resolution of let's say 100 it will fit those exact same 5300px in 53 inches, making you think the image is bigger.
The internal pixel size will not change.
The main unit in a vector-based program is physical units, inches, cm or mm.
You need to differentiate some other processes.
This is a process to actually change the pixel dimension of an image.
Let's say you are using an image just for a thumbnail. You do not need it big, a small one, not in apparent size, but actually with fewer pixels, is needed. Then the process you need is resampling it.
This is just a process to view an image. It will not change the pixels of an image, it will not change also the size compared to your document. It is just a way to view it bigger or smaller on your screen.
So the image is fine regardless of whether or not I scale it up or down?
Using tools like the "Free transformation tool" will only change the size in relation to your document. The pixel count of the image will stay the same.