[Colleen]

I will try to describe this as best I can.

This user works with fairly long documents (50-75 pp) which have tracked
changes and automatic paragraph numbering tied into styles. When he wants
to edit on a page far down in the doc, he typically holds down his Page Dn
key until he gets to a page half to ¾ way down, and usually he has overshot
his mark and needs to press the up arrow key to get up to the text he needs.
A few seconds into his up arrowing, the screen will briefly freeze, then his
cursor has suddenly transported to page 3 or 4. It happens consistently when
he is up arrowing, and it never places the cursor at the very top of the doc,
or on later pages, (page 20, e.g.) just page 3, 4 or 5. He prefers the
keyboard, so I dont know if it would happen if he tried scrolling with the
mouse. I did suggest he use Edit Find, to get to a spot that far down in a
long document, but he prefers the Page Dn, up arrow method.

He says it happens in many different documents €“ including documents created
from other offices. I tried it out on my PC and didnt have any problems
doing what he did. He is the type of user who doesnt want anyone touching
his computer, so I wasnt actually able to test it out on his PC.

I did find an article in the KnowledgeBase about slow scrolling in Word docs
with 100 or more consecutive automatic numbered paragraphs. His could easily
have that many numbered paragraphs (I havent counted), but I think the
article described the behavior when using the mouse. His is always using the
arrow keys.

He rarely uses other programs, so I dont know if this happens in anything
other than Word. I did have him pull up a long web page in Internet
Explorer, but we didnt see any problems.

One other thing hes noticed, which may be related, is that sometimes when
hes moving up, a line of text will suddenly be repeated many times on the
screen. He was not able to demonstrate this. To me, it sounds like a screen
refresh or low memory problem, but he didnt describe in very much detail.
It doesnt seem to be as consistent as the first problem I described.

Thanks for any suggestions,

Colleen