As my contribution to this discussion I am sending an young adult
normal database of our own. Again, it is small, smaller, than
Richard's in fact (5 adults, 25 trials). I will include the GCD file
and an Excel file (containing the COV info + the GCD data) in a
separate message. As you might expect the results are very similar
to
Richard's as the same marker protocol is utilised. The most obvious
difference is that his hip flex/extension is offset from ours
into
extension by approx 10deg. This difference is made up of approx 5deg
from flatter pelvic tilt (maybe due to marker placement, but
maybe
due to age as a database of 50-65 yo we have has the same pelvic
tilt as Richard's group) and 5deg from a more extended knee (we place
the knee marker on the anterior aspect of the lateral epicondyle,
which although leading to more flexion is a more consistently
palpable landmark).
With regards to COV, these are also similar but the obvious exception
highlights one of the weaknesses of COV. The COV for knee rotation
is
62 for Richard and 247 for us!! This is due to the very small values
we obtained for this variable (the SD is bigger than the value), due
to our experience with the marker protocol. When making up this
database we compared the graphs for a 3 subject (15 trials) database
and a 5 subject database (25 trials) and found that the graphs
(including SDs) superimposed. Unfortunatley we do not have the COV
information to substantiate this.
Jeremy Linskell
Manager, Gait Analysis Laboratory
Co-Ordinator, Electronic Assistive Tehcnology Service
Dundee Limb Fitting Centre
Dundee, DD5 1AG, Scotland
tel +1382-730104, fax +1382-480194