You would have to adjust the position of the box or sphere used to visualize the collision area.
In the following example the avatar origin is at the bottom of his feet so the box does appear to be in the correct position. Hit the spacebar to adjust the position of the box.
Code:
import viz
viz.go()
viz.clearcolor(viz.SKYBLUE)
ground = viz.add('tut_ground.wrl')
logo = viz.add('logo.wrl',pos=[-1,1.5,6],euler=[0,0,30])
logo.collideBox()
logo.add('box.wrl',alpha=.2,scale=logo.getBoundingBox(viz.REL_LOCAL).size)
avatar = viz.add('vcc_male.cfg',pos=[1,0,6],euler=[180,0,0])
avatar.collideBox()
box = avatar.add('box.wrl',alpha=.2,scale=avatar.getBoundingBox(viz.REL_LOCAL).size)
vizact.onkeydown(' ',box.setPosition,[0,.9,0])
import vizcam
viz.cam.setHandler(vizcam.PivotNavigate(center=[0,1,6],distance=5))
The collideMesh area is the same as the object itself so it is not necessary to build anything to visualize it.