As with most lenses the optimal results are achieved when stopped down - in the case of the 18-55mm VR, to ƒ/8 or ƒ/11. Best results for sharpness seem to be achieved with the lens set to wide angle (18mm) and ƒ/11, where sharpness is excellent across the entire frame, between 1 and 1.5 blur units. At other focal lengths and apertures, we see some evidence of corner softness, but the central region is still quite sharp.
There are some odd deviations from otherwise stable image sharpness, for example at 18mm and 35mm you get fairly consistent results, but at 24mm there are some large spikes of corner softness. This might not be an issue with a properly centered lens - or at least not really visible - but it's definitely something to look out for, and may speak to sample variation. For example, at 24mm and ƒ/4, we see upwards of 6 blur units on the right corners.
To the lens' credit, diffraction limiting sets in only at ƒ/22 or greater, meaning ƒ/16 is actually quite usable. Images sharpness is above average, even fully stopped down, at 4 blur units pretty consistently across all focal lengths.
In summary, for the money, it has more than acceptable results, especially for a plastic kit lens.
Chromatic Aberration
The lens shows good resistance to chromatic aberration, much improved over the previous version of the 18-55mm. This is noteworthy as Nikon removed the ED glass element, in favor of an improved lens coating. Chromatic aberration is present at the wide angle (18mm), consistently in the corners and overall as the aperture is stopped down, but as the focal length increases chromatic aberration all but disappears by 35mm. Nikon is definitely ahead of the game when it comes to dealing with CA.
Shading (''Vignetting'')
Where the previous version of the 18-55mm showed significant corner darkening only at 18mm (over a stop darker in the corners), the VR version shows less at 18mm, but vignetting is now seen at all focal lengths where the aperture is set to between ƒ/3.5 and ƒ/5.6. With the previous version, vignetting was a non-factor if you used a focal length other than 18mm. However, we're only talking about 2/3 of a stop in the worst case scenario (18mm and ƒ/3.5); in other focal lengths wide open we see a half-stop darker in the corners than the center. By ƒ/8, corner shading has disappeared (less than 1/4 stop).