Hello,
I have a MedGraphics VO2 analysis system (BreezEx) with which I have been
having problems with the FIO2. The system calibrates fine to reference and
calibration gas concentrations; and, once calibration is completed, the
system reads room air just fine. However, once I begin collecting data, the
FIO2 takes a considerable dive: presently down to 20.70% rather than the
20.93% that is room air. This is having a significant effect on the
validity of my VO2 data.

My question is has anyone experienced similar problems with this system? I
have tried everything I can think of to figure out why FIO2 is taking a
dive and no luck (nor with the company). They analyzed my O2 module and
insist it is performing properly. The O2 cell has been recently replaced,
sample lines, desiccant etc..nothing has corrected the drop in FIO2.
Strangely, the FICO2 does recover without problem following an expiration.
So, the problem seems to be only on the O2 side.
Any ideas? Thanks,

----------------------------------------------------------
William A. (Turi) Braun, Ph.D.
Assistant Professor
Department of Kinesiology and Health Promotion
California State Polytechnic University
Pomona, CA  91768
(909) 869-2771
(909) 869-4797 Fax
wabraun@...

STATE UNIVERSITY OF NEW YORK, COLLEGE AT BROCKPORT

DEPARTMENT OF PHYSICAL EDUCATION AND SPORT

ANNOUNCEMENT OF VACANCY

EXERCISE PHYSIOLOGY INSTRUCTOR


This one-year, non tenure-track position (see note below) involves
teaching and service related to exercise physiology.  Primary duties
will include 1) teaching undergraduate introductory exercise physiology
courses, which include weekly lab meetings; 2) possibly teaching
undergraduate advanced exercise physiology course, which also includes
labs; 3) advising students in the exercise physiology concentration; and
4) providing professional support to the exercise physiology program.

The required qualifications for this non tenure-track position include
1) earned master's degree in Exercise Physiology or related field; 2)
the ability to teach a culturally diverse student body.   The preferred
qualifications include:  1) Ph.D. degree in exercise physiology; and 2)
previous college teaching experience.

The rank shall be at qualified academic rank and a competitive salary
will be commensurate with experience.

Application deadline:  To assure full consideration, applications must
be received by May 23, 2000.  A letter of application, vita, graduate
transcripts and names, addresses and telephone numbers of three
references should be sent to:


Heidi Byrne, Ph.D.
Dept. Physical Education and Sport
350 New Campus Drive
Brockport, NY 14420
(716) 395-2601
hbyrne@...


Update to job listing.  Job listing was posted last week.  Position is a
one-year temporary replacement.  However, in one year we will do a
formal search for a full-time, tenure track position.  Therefore, there
is a possibility (but of course, no guarantee!) that the person who
filled the one year temp job would be hired for the tenure track
position, IF that person had done an excellent job AND met the
qualifications for the tenure track search.

We do not normally send virus alerts to the Sportscience list, but the
current situation is extraordinary.  I just received what I think has to be
a mutant of the I LOVE YOU worm.  The message was simply:

>Dear All
>
>I apologise if you were infected by the love letter virus through my Email
>directory. Some information attached.
>
>Keith

and the attached file was called VirusAlert.doc.

I immediately trashed the file without opening it.  I strongly advise
everyone to do exactly the same with any attachment that you have not
received from a trusted source.

If you receive a message with such an attachment and other people are on
the recipient list, please email them a warning not to open the
attachment.   Mark it READ THIS FIRST Re: [whatever the subject line was]
and make the message priority urgent.

Will

Faculty Position at Castleton State College (by way of Young-Hoo Kwon)

CASTLETON STATE COLLEGE
Physical Education Faculty Vacancy
Search Extended

  The Physical Education Department at Castleton State College is seeking
applications for a tenure-track position in Physical Education beginning
August 2000.  Responsibilities would include teaching courses with a special
emphasis on Exercise Physiology, Senior Research Project, Cardiac
Rehabilitation, and possibly other courses in the physical education
department.  Additional responsibilities include advising students,
professional scholarship, and supervising field experiences, student
teachers, or internships.  Knowledge of computer skills specific to exercise
analysis is desired.  QUALIFICATIONS: Applicants must have a doctorate in
physical education or related field (ABD considered-reasonable future
completion of doctorate expected).  The ability to work collaboratively with
colleagues and strong teaching, laboratory and interpersonal skills are
essential.   Castleton has placed a special value on the teaching roles of
its faculty, and candidates for this position will be evaluated principally
on the basis of their potential to be outstanding teachers and colleagues.
Rank and salary are dependent on qualifications and experience.  Review of
applications will begin on June 1, 2000 and continue until the position is
filled.  Send résumé, three letters of recommendations and a statement of
interest to:  Dr. Joseph T. Mark, Academic Dean, Castleton State College,
Castleton, VT  05735.


EQUAL OPPORTUNITY EMPLOYER

Dr. Bob Grace
Http://www.castleton.edu
Note: New Email robert.grace@...
Office: 802-468-1363
FAX: 802-468-2189

This is in response to Mel Siff's posting on whiplash injury:

<< April 24, 2000

  Re: Effect of eliminating compensation for pain and suffering on the outcome
of insurance claims for whiplash injury, Cassidy, et al. (NEJM Vol 342, No 16)

  Dear Colleagues,

  One of the most misleading pieces of medical literature that I have had the
misfortune to come across has been published recently in the New England
Journal of Medicine. The study is called the "Effect of eliminating
compensation for pain and suffering on the outcome of insurance claims for
whiplash injury." In it, the authors describe the results in Saskatchewan of
switching from a tort-based system of compensation in which an injured
claimant could sue for injuries caused by the negligence of another, to
No-fault insurance in which the claimant cannot recover any damages for pain
and suffering. The study followed time to claim closure for individuals under
the pre-1995 No-fault system (the Tort system) and did the same for
individuals who were injured under the No-fault system. The authors reported
that time to claim closure, and thus recovery, was reduced by 54% under
no-fault, and attributed the shortened time of recovery to the elimination of
compensation for pain and suf!
  fering.

  One of the most serious problems with this study is that the authors did not
discuss all of the changes that attended the implementation of No-fault, such
as mandated rehabilitation and work hardening regimes at six weeks
post-crash, and psychological evaluation at insurer-run centers after 12
weeks. The Accident Insurance Act gave the right to the insurer to close the
patient’s claim or withhold benefits if they did not "follow or participate
in a rehabilitation program made available by the insurer" (Section 185 [g]).
For this reason alone, the authors are incorrect to attribute the duration of
recovery to the cessation of compensation for pain and suffering.

  Another serious problem is the use of time to claim closure as a proxy for
recovery. The term "recovery" implies a restoration of normal function, as
well as cessation of pain. The problem is that the authors of the
Saskatchewan study did not report on the level of pain or function of the
people in their study at the time of claim closure. There is no indication in
the study that people did any better, or for that matter did not do worse
under No-fault insurance. The fact that claims were closed sooner may have
been nothing more than an artifact of changing to a system in which injured
people cannot negotiate or litigate an injury settlement, and thus there are
less administrative reasons to keep the claim open.

  Additionally, a system that can mandate treatment at centers with
practitioners who are employed by the insurer (so-called Tertiary Centers) is
also a system in which claim closure could reasonably be expected to be
expedited. It is also important to recognize the fact that under the Tort
system, the insurer benefited financially by delaying claim closure when a
settlement was to be paid.

  While the authors of this study purported to study the effect of removing
the financial motive for continued pain complaints on the duration of
recovery from whiplash injuries, all they really did was demonstrate that, in
a totalitarian system of health care in which the insurer is given complete
control over the lives of its premium payers, it is easier to get a claim
closed more quickly. This doesn’t come as a great surprise to many.

  Michael D Freeman, PhD DC MPH
  Forensic Trauma Epidemiologist
  Department of Public Health and Preventive Medicine
  Oregon Health Sciences University School of Medicine
  2480 Liberty Street NE Suite 180
  Salem, Oregon 97303
  drmfreeman@...
  503.763.3528


  Ben Weitz, D.C., C.C.S.P., C.S.C.S.
  LHBWeitz@...
  http://www.drweitz.com >>


Ben Weitz, D.C., C.C.S.P., C.S.C.S.
LHBWeitz@...
http://www.drweitz.com

There are only two truly infinite things, the universe and stupidity.
And I am unsure about the universe.
Albert Einstein

In reply to the upper and lower ab question, I agree with you, Kenneth,
that it should not be possible to contract the upper or lower abs
seperately.  When most people talk about upper and lower abs they are
referring to the rectus abdominis (RA).  There is one muscle that lies
anterior to the RA, the pyramidalis, but its only function is to contract
the linea alba.  The muscle fibers of the RA run the entire length of the
muscle from the lower ribs to the pubis (Gray's Anatomy and Ng et al.,
1998).
	 A couple of possible reasons exist as to why people believe the lower
abs can be trained independently of the upper abs.  The first may be due
to the "burn" individuals may feel when performing certain exercises,
most of which are of the leg lift varieties.  The RA must contract
isometrically to stabilize the pubis so that the hip flexors can lift the
legs.  This isometric contraction may cause an ischemic response causing
lactic acid to build up.
	 The second reason may be due to misunderstanding the data.  If
electrodes are placed on two different areas of the same muscle,
different readings will result (De Luca, 1997) due to different distances
from the motor point.  This does not mean that one part of the muscle
fiber is working harder than another part, because muscle fibers fire on
the all or none principle.  Either the muscle fiber contracts completely
or not at all.  Also, some studies look at the upper and lower RA accross
different exercises.  A problem with this approach is that more difficult
exercises recruit more muscle fibers, therefore, giving a higher mv
reading.  For example (using your bicep analogy), curling 50 pounds
recruits more muscle fibers than curling 5 pounds.  If readings were
taken on the lower part of the bicep for the two exercises, one could
intepret that the 50 lb. curl works the lower part of the bicep more than
the 5 lb. curl.  This is not true because the 50 lb. curl works the
ENTIRE muscle to a greater degree, not just the lower half.
	 I conducted my thesis on the upper and lower portions of the RA.  I used
the crunch, V-sit-up, and posterior pelvic tilt.  I took the arithmetic
difference (delta) between the upper and lower RA for one exercise and
compared that to the delta for another exercise.  I found no difference
between the two sections.  Piering et al. (1993) in the Journal of
Athletic Training and De Sousa & Furlani (1974) also found no difference
between the sections of the RA.  Guimaraes et al. (1991) did find
differences.
	 I hope this has helped with your question.



Darren Edgington
Assistant Strength and Conditioning Coach, MTSU
dedge007@...
phed007d@...
________________________________________________________________
YOU'RE PAYING TOO MUCH FOR THE INTERNET!
Juno now offers FREE Internet Access!
Try it today - there's no risk!  For your FREE software, visit:
http://dl.www.juno.com/get/tagj.

On 5/6/00, Darren Edgington< dedge007@...> wrote:

<< A couple of possible reasons exist as to why people believe the lower abs
can be trained independently of the upper abs. .....This isometric
contraction may cause an ischemic response causing lactic acid to build up.>>

***Agreed that ischaemia can explain the burn, but the burn does not
necessarily depend on lactate or proton accumulation.

I fully concur with the thesis that the upper and lower abdominals cannot be
recruited in total isolation of one another.  If the trunk is actively
flexed, then the entire abdominal musculature becomes involved.  Following
the principle of "reductio ad absurdum", we would be forced to conclude that
if one part of the abdominal musculature were to be entirely quiescent during
an extended period of active trunk flexion, then this would produce
instability somewhere in the trunk and increase the potential for injury.
The uninjured body does not behave like that.

However, there are other issues to consider if we extrapolate this deduction
to mean that the different parts of the abdominal musculature cannot carry
out functional activity to different degrees and in different patterns.  In
this letter, I wish to highlight such issues.

<<The second reason may be due to misunderstanding the data.  If electrodes
are placed on two different areas of the same muscle, different readings will
result (De Luca, 1997) due to different distances from the motor point.  This
does not mean that one part of the muscle fiber is working harder than
another part, because muscle fibers fire on the all or none principle.>>

****Also agreed, since geometric differences between electrode placement and
conductance or attenuation differences of the electrical signal through the
various soft tissues can affect the summated potential measured at any site
or between various sites.

<<Also, some studies look at the upper and lower RA accross different
exercises.  A problem with this approach is that more difficult exercises
recruit more muscle fibers, therefore, giving a higher mv reading.  For
example (using your bicep analogy), curling 50 pounds recruits more muscle
fibers than curling 5 pounds.>>

I conducted my thesis on the upper and lower portions of the RA.  I used the
crunch, V-sit-up, and posterior pelvic tilt.  I took the arithmetic
difference (delta) between the upper and lower RA for one exercise and
compared that to the delta for another exercise.  I found no difference
between the two sections.  Piering et al. (1993) in the Journal of Athletic
Training and De Sousa & Furlani (1974) also found no difference between the
sections of the RA.  Guimaraes et al. (1991) did find differences.  >>

***One problem with the interpretation of EMGs is that greater potentials
measured over any region do not necessarily imply that the muscle locally or
over its entire volume is producing 'functional' motor output.  Some of the
activity may be spurious (due to sub-optimal motor skill) or stabilising in
nature and not contributing solely to the dynamic action being studied.  For
example, it has been found that skilled performers recruit less muscle tissue
than less skilled individuals for executing the same activity.

Fatigue can also confound the results.  In this case, a highly trained,
strong subject in executing a given abdominal exercise uses a far smaller
proportion of his maximal trunk flexion strength than a person who is less
strong, so that fatigue even in a single repetition plays a smaller role in
the production of EMG or trunk torque.

Thus, in using the EMG to analyse human movement, it can be very helpful to
concurrently measure joint torque to examine the nature of the relationship
between motor action and EMG from different locations over the abdomen during
exercises that take the trunk over its full range of movement (ROM).  In this
respect, crunches, V-sit-ups, and posterior pelvic tilts do not involve the
entire ROM from prestretch (i.e. about 10-15 degrees of spinal extension to
full contraction with the lumbar spine maximally flexed.

Lest these comments be regarded as somewhat academic, it is interesting to
take the analysis out of the laboratory into the folowing few practical and,
sometimes most unlikely, situations.

Case 1: Anyone who has watched high level bodybuilding contests will have
noticed that some contestants are able to tense, twist, contort and rotate
different parts of the abdomen.

Case 2:  Several years ago, one of the TV learning or discovery channels
featured a belly dancer in a club who performed an amazing feat of abdominal
muscle control.  She lay supine on a table with a four coins on either side
of her lower abdomen and she contracted her abs sequentially until she had
flipped the coins over until they reached the upper abdomen.  She then
repeated the sequence from top to bottom and then left to right, right to
left and then in diagonal sequences.  As a grand finale she called for a
dollar note to be placed upon the centre of her abdomen and she proceeded to
fold the note in two movements to a quarter of its size.

This act was witnessed by millions of TV viewers around the world, so it is
not simply anecdotal - if someone out there has kept a video copy or knows
where to obtain a video copy of this exceptional act, I would be most
grateful to know.   It demonstrated to an even greater degree than the
bodybuilders' posing routines that it is possible to functionally utilise
different parts of the abdominal musculature.

Would anyone care to explain how this act is possible if there is no change
in EMG and much of the current theory seems to suggest that localised muscle
control like that is mythical?

This is certainly not to suggest that the average person executing all of
those gross (as opposed to fine motor control) abdominal exercises is capable
of voluntarily activating different parts of the abdominal musculature for
aesthetic 'body sculpting'.  Rather, this example has been raised to
stimulate us to look deeper at the use of EMG alone to analyse muscle
activity insofar as it relates to functional motor activity.    Regarding
this, has anyone used EMG or myotensiometric biofeedback to examine to what
extent one can voluntarily alter activation of different parts of rectus
abdominis?

Dr Mel C Siff
Denver, USA
mcsiff@...

This article examined differences between upper and lower abdominal
activation in skilled and less skilled  subjects.  Interestingly, one
conclusion was that, among incorrect performers of abdominal  exercises,
either exercise indistinctly activates the muscle portions - and this is
precisely what finding was reported by Darren L Edgington <dedge007@...>
in his letter to this group.

In other words, though various abdominal exercises may appear to be very
similar as performed by different individuals, the pattern of muscle and
recruitment and kinesiology may differ significantly.  One again, we note the
fundamental role played by the superordinate cortical and subcortical neural
patterns in determining motor output even in apparently rather gross
activities.

-------------------------------------------------

Sarti M et al   Muscle activity in upper and lower rectus abdominis during
abdominal exercises.  Arch Phys Med Rehabil 1996 Dec; 77(12):1293-7

  OBJECTIVE: To compare the intensity of the upper versus lower rectus
abdominis (RA) muscle activity provoked by each of two different abdominal
exercises and to contrast the intensity of contraction elicited by two
different abdominal exercises on each RA muscle portion. PARTICIPANTS:
Subjects who had practiced  endurance or strength training activities (1.5
hours 3 days a week for 3 years) and those who had not  accomplished that
criterion comprised a high and a low physical activity group, respectively.
Each of  these two groups was divided by the ability to perform the exercises
into two subgroups: correct and  incorrect performers (cp, ic).   MAIN
MEASURE: Average surface iEMG was compared  between upper and lower RA and on
each muscle portion performing curl-up (CU) and posterior pelvic tilt (PT)
exercises. The coefficient of variation, a two-way analysis of variance, and
the t test were calculated.

RESULTS: The upper RA showed significantly greater activity during
performance of  CU exercise by the cp subgroups of both high and low activity
groups. Only the cp subgroup of the high activity group showed that PT was
significantly  more strenuous than CU exercise on lower RA.

CONCLUSIONS: Among correct performers, CU produces greater activity on upper
RA. For persons who have a high level of activity, PT is more strenuous than
CU on lower RA. Among incorrect performers, either exercise indistinctly
activates the muscle portions.

--------------------------------------------------------------

Dr Mel C Siff
Denver, USA
mcsiff@...

On 5/7/00, Barrie Knox-Davies< barriekd@...> wrote:

<<I suppose I should have contacted Will Hopkins directly on this one, but if
there is anyone out there who can recommend a site or a way of gleening
examples of statistical analysis scenarios for teaching Anova's, chi-square,
Kruskal-Wallis, Mann-Whitney U, and other statistical procedures, please give
me a reply.   I teach at an honours (4th year level) to students in Human
Movement Science, that is what it is called here in South Africa, and I am
desperate for examples of these procedures that I can give them to work-out.>>

Greetings, Barrie - here is a posting that I wrote a long while ago and which
I found in the archives. Maybe this collection of stats resources will be of
some assistance:

---------------------------------------------

STATISTICS RESOURCES

Date: Thu 16 Sep 1999
Sender:  Dr Mel C Siff<mcsiff@...>

Students who do not study statistics formally in their degrees invariably
discover that a good working knowledge of stats is essential to their
research and progress. Here is some web information to assist any students
(and staff) who may need online resources on statistics:

The following websites provide numerous valuable resources on statistics:

   http://statistics.com/
   http://www.stat.ucla.edu/
   http://www.stat.ufl.edu/vlib/statistics.html
   http://sportsci.org/resource/stats/index.html

Here are even some Statistics textbooks on the web:

   http://www.psychstat.smsu.edu/sbk00.htm
   http://www.stat.ucla.edu/textbook/
   http://davidmlane.com/hyperstat/index.html

There are numerous useful texts on the subjects which I am sure others can
suggest, but here is a simple book and CD (for MAC and PC) that augments what
you will find in standard texts:

An Electronic Companion to Statistics by George Cobb Cogito Press
ISBN 1-888902-42-6

Dr Mel C Siff
Denver, USA
mcsiff@...

Hi Darren/list

Although, I believe you have made some valid points, complacency and a
general
lack of empirical evidence has led exercise scientists to contend that the
abdominals are a boring one dimensional, one exercise, does all muscle
group.
I've been both guinea pig and scientist, and have been training myself and
others
for a combined 16 years.  I have learned two things:

1.    The abs are not created equally

2.    Acceptance based on lack of empiricy is ridiculous (supporting Mel's
position).

I'm not writing to debate your points.  RATHER, reexamine your Grays
anatomy, concentrating on the nerve innervations of the lower and upper
abdominal region.  What you will see in a nutshell is that the upper are
innervated by Thoracic vert. (i.e. 11 & 12)and lower through lumber vert.
1,2,3.  In my mind, at least hypothetically, different areas of innervation,
would potentially lead to different neuromuscular adaptations. I could
provide the exact nerve bundles I'm speaking of, but self-discovery is much
more gratifying.

It's seems fairly likely that the basic crunch is incapable of completely
innnervating the lower (i.e. lumbar/sacral) region.  I'm not implying that
all versions of the sit-up are similarly ineffective .  In fact, if my
abdominals could speak every Tues and Thurs, they would probably beg to
differ.

Let's start thinking optimal activation, not lack of activation.  The core
is a linked system. We need not look father than the psoas's role in hip
flexion to understand that.  The body is a complex assortment of flowing
activations and resultant movements.  I can't see why the abs wouldn't work
in a similar manner, with subtle shifting, rather than gross motor changes
in activation of linked muscles.

Chad Benson
University of Victoria

Please forward this email to your coaches.

If you are interested in Lactate measurement for training, FACT Canada
now distributes the Lactate Pro LT-1710 portable lactate analyzer.
Clinical accuracy, the size of a credit card at an amazing low price
($339US for a starter set).

This meter is used by the US and Canadian National Cross Country ski
teams, Rowing Canada, Biathlon Canada, the Wall Aquatic Center at
Northern Arizona university and by World Class Marathon runner Lidia
Simon, to name a few.

Check our website for more info: http://www.fact-canada.com

Herb Chlebek

David Driscoll, an Australian who persuaded me to run seminars in Oz while I
am talking at the "Maths and Computers in Sport" conference in Sydney next
month asked me to post this announcement here for any Aussies who may be on
this listserv:

Subject: Dr Mel Siff Down Under

On June 9-11 Dr Mel Siff will be delivering a series of strength and
personal training lectures in Sydney, Australia. The lectures will be along
the lines of the popular “Supertraining Camps’ that he runs in the USA. They
will also include unique material that he was invited to write by the
International Olympic Council  for their latest Sports Medicine volume to
coincide with the 2000 Olympics.

Details are still being finalised, but interested Aussies (or visitors who
will be in Sydney on these dates) should contact David Driscoll at:
driscoll_david@...   for further information. Places are limited so
don’t delay.

-----------------------

Dear All,

Just a bit to add to Andrew Doyle's article in the most recent
SportScience on how to get V(dot) in MS word. I have found a true type
font called 'WWW Font' and a capital 'O' gives you a perfect V(dot)
which obviously is easily changed to any size. The rest of the font is
not really usefull for anything else, but who cares. I have no idea
where I found this font but if you have trouble locating it I will send
you a copy.

Richard
----------------------
Dr R.C.Richard Davison
Senior Lecturer
Cycling Research Group
Department of Sport and Exercise Science
Canterbury Christ Church University College
North Holmes Road
Canterbury
Kent, CT1 1QU
Tel: 01227 782680
Fax: 01227 470442
Mobile: 07967 657235

----- Original Message -----
From: <Mcsiff@...>
Subject: : CLOSED KINETIC CHAIN?


>>>>> On 5/5/00, John Casler<BIOFORCE@...> wrote:
>
> << The reference is to the "standing barbell curl".  I know there is
debate
> as to exactly what a CKC Exercise actually is but I think the standing
> barbell curl fits the definition. >>

Mel Siff wrote:
> ***Let's go back to the source of all ideas about CKC and OKC.  The person
> who coined the terms Open and Closed Kinetic Chains was Dr Steindler and
he
> stated that a closed kinetic chain exercise is one in which the end of the
> limb is fixed or encounters large resistance.

> Finally, we have to ask if categorisation of exercises into closed and
open
> classes really is of much value?  Should this classification simply be
> abandoned or should we retain its use, but demand that a very precise
> description be stated on the conditions acting in any given
situation?>>>>>

*****Hi Mel,

With interest, I have watched occasional debate over the subtleties of
"closed and open kinetic chain" exercise.  I have always operated under the
idea that the body and its structure is like a "chain" of joints.  If one
end of the chain is "anchored", (lets say the feet) then the "longest" chain
we can have is the to the hands.

A "kinetic" chain would imply movement.  So I always took it to mean
applying force "through the body" (chain of joints) from the "anchored" (or
fixed) section, to the "terminal" or farthermost segment where outside force
is applied or encountered to "move" something.

I also have thought, that a single joint exercise does not create a "chain"
(accept to be acknowledged as a single link) and subsequently should not be
termed as such.

It would seem simple to accept that any multiple joint (chain forming)
exercise where the "fixed" portion is important (as a stabilizing platform)
to the application of force to the "terminating" portion would be a "closed
chain".  This would preclude exercises like the "chin" since the "force"
being actually transmitted from the "fixed (hands) "through the chain" to
the termination (feet in this case) segment is not applied to anything.
Thus the chain is "open".

My "closed chain" perception "would" include examples like the Bench Press
since the body would be fixed (to the bench) and the force from this
platform would be transmitted through more than one major joint system to
the "distal" segment.  It would just be a "shorter chain" than say the
"standing press" where the fixed portion is the feet, creating a longer
chain.

Another example of "shortening or lengthening the chain" would be a standing
vs a seated barbell press.  Sitting makes the chain shorter by taking all
the joints out from the hip down. As support for this concept; Who would
deny that an "amputee" is not performing a "closed chain" press if doing so
was from the hip.

An "Open kinetic chain" would be where we supply force on either or both
ends.  One end may be fixed (maybe termed "fixed base open chain") or both
ends ("total open chain") maybe un-fixed as in the butterfly stroke in
swimming.  So generally they would be like chins, or even walking.  Free
hand exercises get a little sticky but doable..  For example we might
consider a "push up" a closed kinetic chain" while a "dip" is "open".

Single joint, isolation exercises (like the knee extension) would not
qualify as "chains" per se, and simply be termed "single joint" or if we
must carry the "chain" designation "single link" exercises. They could in
fact be considered, "open" single link and "closed" single link, depending
on the disposition of the fixture and forces at the opposite ends of the
joint involved.

So to recap,

"closed" terminated, fixed or encountering forces from both ends so that
significant forces are experienced through the chain

"open" encountering or exerting force from only one end of more than one
joint or causing action with no fixed base (as in the butterfly stroke),
with out significant external resistance or significant external force
against the end or ends

"kinetic" moving or providing resistance or movement against an external
force or the body itself

"closed kinetic chain" exercise that has a fixed (base) and a distant
(termination) body segment which is "more" than one joint in length .  The
"chain" is closed by one end being fixed and the other exerting force on an
external force or resistance.

"open kinetic chain" has  either no, or one, fixed base but the other end of
the chain has no external force applied to it or resistance to its movement.
It is "open" because it is not terminated to any external force.
There are basically two types of "open chains"
1) where one end is fixed ("fixed base open chain") and
2) where both ends are "open" ("fully open chain")

"single link" exercise is a single joint isolation exercise

"closed link" where both ends of a single joint exercise are either fixed,
acting against significant external force, or any combination.

"open link" where one end is fixed or has external load applied and the
other does not.

a "chain" refers to the number of major joints through which force is
actively applied and can be any length including at least two major links
(joints)

"active" means where either tension for stabilization or the resulting
movement occur.

"static or Isometric" chains and links can also exist but are not kinetic.

NOTE: the above is my opinion only as to what the terms "should" mean based
on a logical interpretation of their individual definitions, and may not be
the generally accepted concepts from the originators and those who currently
use the terms.

The importance of this terminology is that it allows us to examine, evaluate
and categorize an exercise in a way that permits us to understand the
participating movers, stabilizers and co-contracting partners.  It allows us
to see how "reducing the chain" while not necessarily a "bad" thing, changes
these "moving, stabilizing and
co-contracting patterns.

If function is of importance then generally "longer chain" exercises that
most similarly match the activity, may be more valuable.  Balancing the
"weak" links (if they are perceived and evaluated properly) can provide
great training.  Developing specific chains beyond others can also create
"functional imbalances" if they are not "integrated" and "translatable" into
coordinated motor activities.

(Geeeeesh, I was just going to make a couple comments and look what
happened)

Regards,

John A. Casler

BIO-FORCE, Inc.

"We strongly recommend the use of FORCE"

In reply to Barrie Knox-Davies...

A couple of months ago I taught a small third-year undergrad class
the basics of design and analysis.  The only way to go, in my
opinion, is to get the kids to each make up their own numbers for
something basic like a controlled experiment to change weight with a
drug.  Then you do t tests and the like.  You as teacher have to do
it yourself first, though.  That way you encounter all the bugs that
the kids will encounter.

We used SPSS, but only because that's the only package we could
access that runs on Macs and PCs.  It's still an unbelievably
unfriendly package, so we never did get to using ANOVA.  You can do
most things with unpaired and paired t tests of change scores, and
the analyses are more robust than the full ANOVAs.

By the way, if anyone out there thinks they have found the
friendliest stats package, please tell me.  It's got to be cheap,
too.  SAS has several potentially good products, but the cost is
prohibitive for general student use, so I haven't assessed them.  (I
use the main SAS package, but that is utterly out of the question for
students.)

To do analysis of relative frequencies (chi-squared tests), get them
to make their subjects a mix of sex and race.  Don't forget to get
them to interpret the magnitudes of the relative frequencies.  It's
not enough to say the difference is significant or non-significant.

I don't teach Kruskal-Wallis and Mann-Whitney as such.  I introduce
the concept of non-parametric analyses through the idea of rank
transformation.  All non-parametric analyses that we meet in our work
are actually the usual t tests and ANOVAs performed on the
rank-transformed dependent variable. (It's misleading to call them
non-parametric analyses, actually.)  Why rank transformation?
Because when you have a non-normally distributed dependent variable,
your left-over bits in a statistical model (the residuals) are also
not normally distributed, so the confidence limits or p values aren't
accurate, and the actual outcome statistic may also be biased.  Rank
transformation is one way to try to get around the problem.  You
usually end up with normally distributed residuals after rank
transformation, but not always.  You are supposed to check the
distribution of the residuals, but not many people do, and even I
can't be bothered sometimes, or I forget.

Make sure you include the idea of log transformation to get better
residuals. In fact, always try log transformation first: people
should use log transformation a whole lot more than they do at the
moment.  The outcome statistics are then percent changes or
differences, which anyone can understand. The outcome statistics
after rank transformation are harder to interpret.  The best outcome
statistic after rank transformation is the effect-size statistic
(change in mean divided by the between-subject standard deviation).
It's unbiased by rank transformation, which is a simply wonderful
result that not enough people know about.

All this and more at http://sportsci.org/stats, of course.  BioMedNet
reviewed stats sites lately.  Go to
http://www.biomednet.com/hmsbeagle/68/reviews/insitu

Will
--
Will G Hopkins PhD FACSM
Physiology and Physical Education,
University of Otago, Box 913, Dunedin, New Zealand
Phone +64 3 479 7330   Fax +64 3 479 7323
Work: will.hopkins@...  Home: will@...
Personal website: http://sportsci.org/team/will.html
A New View of Statistics: http://sportsci.org/stats
Sportscience: http://sportsci.org

List members,

I haven't seen any research examining selective recruitment of upper and
lower abdominal muscles, however it is theoretically possible that upper
Rectus Abdominus (RA) could be more active during upper trunk flexion, and
lower RA more active during pelvic stabilisation or kyphotic movements.  I
have described below a theoretical bases for selective activation of RA to
stimulate further thought discussion on the topic.  I would be very
interested in hearing list member's reactions, however please direct your
replies to the list.

   The theory requires substatiation of 3 main points.

1.  Fibre length change tends to occur with changes in the length of its
sarcomeres (Heslinga & Huijing, 1995; Matano, 1994; Willems & Huijing,
1994) and fibres are often of different lengths in different regions of
muscle such that they are longer in distal regions of long muscles (Scott
et al., 1991; Muhl et al., 1992). As such, sarcomeres have different
resting lengths in different regions of a muscle (Morgan, 1990; Van Eijden
& Raadsheer, 1992; Van Eijden et al., 1997; Willems & Huijing, 1994)
.  Therefore, in long muscles like rectus abdominus (RA), it is possible
that sarcomeres of proximal fibres are at different resting lengths than
distal fibres.

2.  Research by David Morgan and his colleagues has shown that during
muscle contraction, sarcomeres with near-optimum or less-than-optimum
overlap of actin and myosin are likely to shorten whereas sarcomeres that
are on their descending limb (stretched past optimum) often lengthen and
only provide passive force as they reach their physiologically maximum
length.  Therefore, if an entire muscle is activated, only motor units
containing fibres with sarcomeres near, or shorter than, optimum need to be
maximally activated.  Fibres with sarcomeres past optimum length may not
need to be well activated, even during a maximum voluntary contraction,
since these sarcomeres are likely only to lengthen and provide passive
force anyway (this would be an 'inefficient' activation strategy).

3.  Many muscles appear to be compartmentalised, with different nerve
branches supplying different regions of a muscle (Ter haar Romeny et al.,
1982, 1984; Van Zuylen et al., 1988; Tonndorf & Hannam, 1994).  Much
research has also been performed by Gans and colleagues suggesting that
many long muscles of animals (particularly the cat sartorius) contain
compartments that are activated differently depending on the movement in
which the muscle is involved.  Tonndorf & Hannam in fact showed that often
compartments are separated by tendinous tissues with motor units rarely
crossing into neighbouring compartments.  The RA muscle in humans is
separated into anatomical compartments by the linea alba running
longitudinally and the linea transversae running transversely.  The Email
by Chad Benson suggested that different parts of RA are innervated by
different nerve branches.  Therefore, RA is a prime candidate for
functional compartmentalisation.

Collectively, these three points suggest the following:

If RA is functionally compartmentalised, the upper and lower sections (or
perhaps there are more than two functional compartments) could be activated
differentially.  It may be therefore that different nerve branches activate
RA sections differently depending on whether upper or lower parts contain
fibres with sarcomeres at optimum length for maximum force or endurance
during a specific contraction.  In regions of the muscle where sarcomeres
are on their descending limb, there would be no need to activate maximally,
even during a maximal contraction, since they would lengthen to exert
passive force anyway.  These regions of muscle would act in a tendon-like
manner transferring force from more active regions of the muscle.

*To speculate further, the length of sarcomeres in upper or lower RA might
be related to the moment arms of the insertions of these abdominal
regions.  Differences in the moment arms of different parts of RA would
change the length range of motion throughout which the fibres must shorten
during a contraction.  If the moment arm of one region deemed the active RA
region to contract over a long range of motion, sarcomeres active during
movements involving that RA region might contain longer sarcomeres with a
longer active length range of motion. This active region would
theoretically be close to the insertion point.  There would be no point,
for example, in lower RA containing optimum sarcomere lengths for upper
trunk flexion since force might be dissipated by series elastic structures
and the lengthened sarcomeres of the upper RA. It would be more efficient
for the lower RA to contain optimum sarcomere lengths for lower trunk
flexion since force would be applied directly to the bone via its tendinous
insertion, and the upper abdominal region (which might be less active and
contain more sarcomeres lengthened to provide passive force) would act like
an anchor and an extension of the upper tendinous insertion.

Such a theory is speculative, but it seems possible that differential
activation of RA sections is not only possible, but could be beneficial,
during different movements.  Often anecdotal evidence precedes scientific
proof so our inclination that abs can be targeted differently by different
exercises may indeed be correct.

I would be interested in any feedback or discussion on these suggestions.

Tony.


********************************************************************************\
*******

Anthony Blazevich
School of Exercise Science and Sport Management
Southern Cross University
PO Box 157Lismore Australia 2480

Email:   ablaze20@...
Phone: +61 2 6620 3231
Fax:     +61 2 6620 3880

********************************************************************************\
*******
*

Greetings!

I know that our esteemed colleague, Will Hopkins, makes snide (and
justified) comments on his statistical website about one-tailed hypothesis
testing, but I was looking for an example of when using a one-tailed test
would be appropriate. As I was taught, a one-tailed test is appropriate when
there is absolutely no possibility of the true result falling in the other
tail (e.g., being hit by a truck decreases your risk of injury) or when the
body of evidence is so overwhelming, there is no reason to assume that the
true result would fall in the other tail (e.g., cigarette smoking decreases
your risk of lung cancer). The example that I am seeking is when, if ever, a
one-tailed test is appropriate in sports medicine/science. I'd prefer a
realistic example if possible. Thank you in advance for your thoughts.

Regards,

Stephen Lyman, Ph.D.
Director of Epidemiology & Clinical Research
American Sports Medicine Institute
1313 13th Street South
Birmingham, Alabama 35205
Office: 205-918-2132
Fax: 205-918-0800
e-mail: StephenL@...
web: www.asmi.org

In Reply to Will Hopkins

I have recently been introduced to Statistica which also works both on
PC and Mac. I have not used it for teaching as we currently have
only one copy, but I have used it for my own research and it works very
well, particularly in repeated measures designs which we tend to do a
lot of in sport science.
Whether it will be any better for teaching students I dont know but I
suggest it is worth a look, but personally I wouldnt be without it for
my own work.

Richard

----------------------
Dr R.C.Richard Davison
Senior Lecturer
Cycling Research Group
Department of Sport and Exercise Science
Canterbury Christ Church University College
North Holmes Road
Canterbury
Kent, CT1 1QU
Tel: 01227 782680
Fax: 01227 470442
Mobile: 07967 657235

Wil:

You and I exchanged memos a while back about a SAS product called JMP.
The full package price is way too expensive for students, as you noted.
However, Thompson Learning (1-800-354-9706) distributes a student
version call JMP-IN Student Edition that sells for $67.95 USD. This
includes the software and 522 page stats text/manual.

There are differences between the full package and the student version.
The following is copied from the www.jmpdiscovery.com web site regarding
package differences:

Analysis Platforms
1. Fit Model: Standard Least Squares JMP IN does not do the exact p
value for the Durbin-Watson statistic.
2. Fit Model Screening: JMP IN does not support Pareto effect charts,
the Box-Meyer Bayes plots, or Box-Cox transformation plots.
3. Fit Model: JMP IN does not have this Log Variance analysis, which was
an experimental feature in JMP to provide for maximum likelihood
variance modeling through log-linear models.
4. Fit Model: Proportional Hazards, Parametric Survival, D-Optimal are
removed.
5. Fit Model: The Clustering platform is not included with JMP IN. In
JMP, it does both hierarchical and k-means clustering.
6.Fit Model: JMP IN supports the Kaplan Meier platform or the Cox
Proportional Hazards Model. The parametric survival model can be done
using the nonlinear platform. JMP IN does not support several other
survival models that JMP does.

  Graph Platforms
       Ternary Plots: Removed from JMP IN.
       Contour Plots: Removed from JMP IN.
       Control Charts: JMP IN does not support the following three
features: CUSUM control charts, periodograms, and variability charts.

  Design of Experiments
  JMP IN does not support:
       Cotter Designs.
       Mixture designs.
       mixed level designs,
       3-level fractional factorials.
       D-Optimal design search

So, for an introductory course, it has all the needed procedures. There
are some nuances of the software that take some getting used to (what
package doesn't). E.g. it doesn't import speadsheets very easily and
post hoc tests on 2+ factor anovas is not real straightforward to the
uninitiated. Computing new variables is a breeze and all outputs are
presented graphically first and the user requests the numeric output.
And it is lightning fast. I have 4-5 years experience with it and find
it to meet all my needs.

Don Kirkendall
Dept of Orthopaedics
University of North Carolina

In agreement with Will,

My approach has been to create experimental data from simple imagined
experiments like
"steroids and weight gain" or "teaching methods and test performance"
paying some attention to the data distribution and such to ensure it was
realistic. I also have pulled down some data files from the internet,
created student data disks and gone form there.

I also collected data on my first year students who I teach strength
training and performance assessment to, and then use it on my second year
spports students during their statistics course.  Jump testing, agility
tests, medicine ball throw etc.  We can
then do gender comparisons, group comparisons, correlation analysis,
reliability tests etc.  Along the way we look at learning effects issues
and other study design aspects.

The approach I am debating for next Fall is to build the entire course
(only 15 hours classroom time) around a large data collection effort where
the whole class would go out into the community and measure physical
activity participation at a variety of sports centers, jogging trails etc.
as well as collect some basic interview info
from a sample of men and women.  Then we would have a whole spectrum of
nominal, ordinal and ratio level data to use as a basis for a survey of
basic inferential methods.

regards,

Stephen Seiler
Stephen Seiler PhD
Assistant professor
Institute for Sport
Agder College
Serviceboks 422
4604 Kristiansand S, Norway

email: Stephen.Seiler@...
phone: (47) 381 41 347
fax    (47) 381 41 301

Endurance performance physiology website:
http://www.krs.hia.no/~stephens/index.html
Sportscience News and journal website:
http://www.sportsci.org

I'm interested to know whether anyone has recently assembled a
list of sport and exercise science journals and their impact factors?
This information does not appear to be readily available but
certainly is helpful in deciding where to submit our work for
publication.

Thanks for any information the list can provide.

Gordon Sleivert
*******************************************
Gordon Sleivert, PhD
Senior Lecturer, Exercise Physiology
Director, Human Performance Centre
School of Physical Education
University of Otago, PO Box 56
Dunedin, New Zealand
Phone: (64)3 479 9109
Fax: (64) 3 479 8309
E-mail: gsleivert@...
******************************************
"Work is for people who don't know how to fish!"


******************************************

Dear list,
wanted to let you know that there are over 30 biomechanics related job
postings at

http://www.biosolutions.net

as always, you are invited to post more and/or use the discussion forums.

Best regards,
Robert Heppe, M.A.
Biomechanical Solutions

Hi..
	 Can anybody supply me with the details of an upcoming
conference in Sydney titled "Maths and Computers in Sport"???

Much appreciated...

Stuart Morgan.

* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
Stuart Morgan
School of Biophysical Sciences & Electrical Engineering
Swinburne University of Technology
Hawthorn, Victoria, 3122
Australia
Ph: 03 9214-8502
Email: swm@...
Web: www.swin.edu.au
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *

On 5/14/00, Stuart Morgan<swm@...> wrote:

<<  Can anybody supply me with the details of an upcoming conference in
Sydney titled "Maths and Computers in Sport"??? >>

***Here is the web page for this Conference:

http://www.maths.uts.edu.au/5MandCS/

Mel Siff

Dr Mel C Siff
Denver, USA
mcsiff@...

Stephen Lyman wanted an example of when to use a one-tailed test in
sport science or medicine.  He offered the following amusing
non-sport examples: being hit by a truck decreases your risk of
injury, and cigarette smoking decreases your risk of lung cancer.
But I don't think even these examples qualify for one-tailed testing.
I'll illustrate with the second example.

You do a prospective study for however many years with however many
subjects.  You find 8% of smokers in your sample get lung cancer,
whereas 2% of non-smokers get it.  Therefore smokers are 4.0 times as
likely to get lung cancer as non-smokers.  But that's never going to
be the true (population) value exactly.  Your stats program will give
you the likely range within which the true value is likely to be.
Let's say it's 0.9 times to 18 times as likely. (This isn't a
particularly powerful study, but that's usually an issue with
one-tailed testing.)  The 0.9 is less than 1, so it really is
possible that smoking is good for your lungs, on the basis of these
data anyway.  That's the way to analyze the data using confidence
limits.

Now let's analyze the data using a two-tailed test.  The p value
corresponding to the above data might be 0.06.  Which means you can't
quite exclude the possibility that smoking could REDUCE lung cancer.
With the new interpretation of the p value, the possibility that
smoking does indeed reduce lung cancer is p/2, or 0.03, or 3%, or
odds of 1 to 32 (3 to 97).

But, you say to yourself, there's no way in the world that smoking is
going to reduce lung cancer.  Apparently that means you can do a
one-tailed test, so you do it and get a p value of 0.03. (The p value
for a one-tailed test is half that of a two-tailed test.)  It's now
statistically significant, but how do you interpret that?  I'm not
absolutely sure, but I think you still have to conclude that there's
a 3% chance that smoking reduces lung cancer.  But, but... you can't
conclude that, because you just assumed that there was no way that
smoking could reduce lung cancer.  So it doesn't make sense.

I feel a bit more comfortable with the explanation of the confidence
interval that corresponds to a one-tailed test.  You see, in
principle you don't have to put a confidence interval symmetrically
around the observed value.  (In the present example, with relative
risks, the confidence interval is skewed anyway, but stats programs
normally place it in such a way that there is equal probability for
the true value being above and below the observed value.)  So you can
slide the confidence interval all round the place: it's still a 95%
confidence interval.  A one-tailed test corresponds to just one
extreme of sliding the interval, when you slide the upper confidence
limit all the way out to plus infinity.  In the present example the
lower limit may then be something like 1.5.  In other words, there's
a 95% chance that the true value of the risk of getting lung cancer
from smoking is somewhere between 1.5 times through infinity times
the risk of getting it from not smoking.  That's the interpretation
in principle, but I don't buy it.  An upper limit of infinity is
unrealistic.  My conclusion is that a one-tailed test is not tenable.

Notice that you can have a one-and-a-half-tailed test if you like, or
anything between two tails and one tail, by sliding the upper limit
only part of the way out towards plus infinity.  If that seems
absurd, isn't a pure one-tailed test just as absurd?  On the other
hand, I seem to recall an argument for one-tailed testing based
on--and I can't remember exactly--something to do with how much
importance you place on smoking being harmful as opposed to
beneficial.  But I think the way to address that issue is to use 99%
or whatever limits, or to use the p value to express the outcome as a
probability that smoking is harmful (or beneficial).

I may sound confident in these assertions, but I could be wrong.
Please let me or the list know if you think there are other ways to
make sense of one-tailed tests.

Will

Dear Group,

The following are questions from my friend Jeff who is an amateur cyclist and
journalist:

Jeff wrote:

"I've got a question for you about climbing muscles, for a story I'm writing.
I'm trying to go beyond the simplistic Bicycling Magazine style of "sing to
yourself as you climb" advice. Accepting that the best way to train is to "go
ride hills," I'm trying to use scientific language, in part, to explain
climbing. I'd like your input about training climbing muscles. Does a cyclist
use ANY different muscles when riding up? Use them differently - in laymen's
terms, a different pull? Any other factors you can talk about for
understanding climbing?"

To this, I responded:

As you know, there are two primary ways to climb, seated and standing. The
differences physiologically and biomechanically,as I know them, are as
follows:

       Seated - enables you to use more muscles in the leg group including
quadriceps for knee extension (i.e.for the top dead center (TDC) to bottom
dead center (BDC)), hamstrings for knee flexion and hip flexors (when you
pull up from BDC to TDC), glutes (involved in hip extension when the quads
are knee extending), calf muscles and anterior tibialis (as you "ankle"
through the pedal stroke). Usually when one is seated they are pedalling at a
steady tempo (whether they're mashing a monster gear at low cadences or
spinning at higher cadences). Most of this you already know. As far as the
heart and lungs go this slow steady rise in oxygen uptake and the comensurate
heart rate is easier to adapt to and since we are using more muscles it seems
"easier" and yet the heart rate may reach pretty high levels (Zone 4 of 5).
All that considered one may be able to maintain this tempo, with a lot of
grimacng,  for a considerble length of time (fitness levels will determine).

       Standing - As you know this all changes as soon as someone feels
frisky and jumps up the road. You now have three basic options. Go with it or
keep tempo and hope they blow up or hope that your tempo will enable you to
catch them. The third is a modification of the other two where as you
accelerate for shorter periods and sit/stand to keep from blowing up and
slowly (hopefully) reel the attacker in. This can only be maintained for a
little while before you need to rest/recover and get back into a rhythm
before you resume your pursuit. Biomechanically standing is much less
efficient. It also cannot be done for long stretches of time at a high
intensity (there are a few exceptions to the rule). Standing is meant for
high intensity, short duration attacks and jumps or accelerations. Think back
to the Miguel Indurain days when he could keep up with Pantani and Chiapucci
and yet they were the "Mtn. goats." As far as pure climbing goes they were
lighter and "flew" up the hills, but his ability to maintain a strong power
to weight ratio kept him within reach.

Your strength/power to weight ratio has a lot to do with climbing (a high
pain tolerance and a lot of mental fortitude helps too). The vertical aspect
of climbing also adds an obstacle. Remember there are three main disciplines
for cycling: sprinting, climbing, and time trialing. A good time trialist
should also be a pretty good climber but the opposite isn't always true
because if the person is lighter they will prbably have problems in windy
conditions. Power/weight is basically how much can someone overcome and how
fast can they do it? Really big guys (i.e Gre Nichols) can TT very well and
can climb with a group, particularly if they're allowed to set the pace, but
as soon as a quick acceleration happens they just can't respond as quickly.
Medium sized guys (i.e. you and me) can TT and climb (when we're fit) with
the best of them except when the mity mites dance up the hills. Climbing is a
skill just like anything else (skill - consistent , competent behavior).

P.S. As far as the power tap system goes, it's a great device for measuring
and providing feedback. It's also less expensive vs. SRM. You'll hear about
more people using the system. I personally don't have experience, but would
love to have one. It's a great way to determine improvement and measure it
regularly vs. testing in an awkward environment.

P.P.s - I like to teach people to "Waltz" up the hills to a three count (i.e.
1 2 3) when they're seated. What this means is that you apply more force on
the downstroke of the 1 and then follow with a little less on the 2 and 3.
The next time the 1 comes around your using the other leg. Eventually you
teach yourself to have a more powerful stroke and you get faster and more
efficient. If you've
got a good "tune" in your head it helps.

After this response Jeff wrote back:

"One question came to mind - could a rider, by climbing in a standing
position for the majority of the time spent on hills, become more
biomechanically effecient? Do lightweight climbers do that naturally over
time by standing so much when they climb? Also, do you have more info on
strength/power to weight ratio? I remember reading in Velonews a while back
exactly how much weight Indurain lost to reach the same level of p/w to match
Pantani, and how much 1lb of weight loss equaled in time on a climb. Is there
a Web site you know of that has that kind of detailed info?"


Here is where I defer to other more knowledgable coaches, etc.

Thanks,

Michael Porter, MS
michejp@...

An invitation to the First Latin American Congress of Science in Football
follows. Please do NOT reply to sender, but to the addresses provided below.

Luis Aragón.
----------------------------------------------------------------------------
-----------------------------------------------

UNIVERSIDAD DE COSTA RICA
Escuela de Educación Física y Deportes
First Latin-American Congress of Applied Sciences and Football
July 11th to July 13th, 2000

San José, COSTA RICA


The University of Costa Rica’s School of Physical Education and Sports gladly
invites to the “First Latin-American Congress of Applied Sciences to Football
(Soccer)”. This scientific event will be held in San José, Costa Rica, from
July 11th to July 13th , 2000.

During the Congress, national and international investigators will present,
discuss, and analyze, the most recent football-related research. Trainers,
students, physical education teachers, professors, and all audiences are
invited to participate in this important event.

Three kinds of events will take place during the Congress:

a)      Oral presentations (lectures): a researcher will orally present a
research/case study. The duration of a regular presentation is limited to a
maximum of 15 minutes. However, for selected magistral conferences, the
duration will be of 30 minutes . Presenters are encouraged to use high-quality
visual aids (e.g., PowerPoint). All this presentations will be held at the
“Scientific City Auditorium”.
b)      Workshops: this is an event that will take place in the football field,
located at the School of Physical Education facilities.
c)      Roundtables: a panel of experts will discuss selected topics related to
the game.

Main Topic Covered in the Congress

The main topics covered in the congress will be:

1.      Nutrition in the football player
2.      Sports psychology
3.      Injuries rehabilitation
4.      Sports medicine
5.      Exercise physiology
6.      The biomechanical basis of football
7.      Strength, power, and speed training in football
8.      Sports Chiropractice
9.      Hydration at different environmental conditions
10.     Journalism
11.     Contracts in football (Sports Law)
12.     Football management
13.     Women’s football
14.     The football referee

However, other areas may be covered.

Information for authors

Authors are encouraged to send their works (presentations, workshops) as soon
as possible (the deadline is June 12th), either via e-mail (MS-Word formatted,
IBM compatible), regular mail (include diskette), or by fax. The maximum
extension of the works must be 6 pages long (single spaced). A brief (1 page)
curriculum vitae must be attached, as well.

Please direct correspondence to:

Walter Salazar, Ph.D.
Escuela de Educación Física y Deportes
Universidad de Costa Rica
Sabanilla de Montes de Oca,
San José, Costa Rica

Fax #: + 506 225-0749

E-mail: wsalazar@...

Mail can also be sent to:

José Moncada, M.Sc.
Escuela de Educación Física y Deportes
Universidad de Costa Rica
Sabanilla de Montes de Oca,
San José, Costa Rica

Fax #: + 506 225-0749

E-mail: jmoncada@...

Please include with your mail your full name, pone/fax number, and e-mail
address. If your proposal is accepted, an official notice will be send to you
within two weeks after we receive your mail.

Costs

The cost of the Congress is US $ 75.00. Participants will receive a
participation certificate and the official publication of abstracts.

Registration

Registration will take place at the School of Physical Education. Registration
on-site will be allowed for foreigners. However, all other participants (i.e.,
Costa Rican) must register before July 7th.

Further information

If you need any further assistance, please do not hesitate to contact us.
Please address your inquiries to Walter Salazar or Jose Moncada.

School of Physical Education and Sports
Phone: + 506 207-3270
Phone: + 506 207-3269

In addition to lots of other informative comments, Will Hopkins wrote:

But, you say to yourself, there's no way in the world that smoking is
going to reduce lung cancer.  Apparently that means you can do a
one-tailed test, so you do it and get a p value of 0.03. (The p value
for a one-tailed test is half that of a two-tailed test.)  It's now
statistically significant, but how do you interpret that?  I'm not
absolutely sure, but I think you still have to conclude that there's
a 3% chance that smoking reduces lung cancer.  But, but... you can't
conclude that, because you just assumed that there was no way that
smoking could reduce lung cancer.  So it doesn't make sense.

Lets see if I can shed some light on this. I believe the correct
interpretation of a p-value of 0.03 for a 1-tailed test is that there is a
probability of 3% that the observed value or one more extreme was found
solely due to chance assuming that the null hypothesis is true. In this case
the hypothesis being tested is that smoking INCREASES the risk of lung
cancer. Therefore, a decrease in lung cancer risk would result in a failure
to reject the null hypothesis. This is not a nonsensical result, simply a
null study. A p-value of 0.03 with a 1-tailed test does not make any
statement about the probability of a true result being in the opposite
direction, because the opposite direction is assumed to be impossible. A
2-tailed test would have a different hypothesis: smoking is associated with
a significant DIFFERENCE in the risk of lung cancer. This hypothesis makes
no supposition regarding the direction of the association, whereas the
1-tailed hypothesis is only interested in an increase in lung cancer risk.
For what its worth, thats my interpretation of the differences in
interpretation of a 1- v. 2-tailed hypothesis test.

I agree wholeheartedly that confidence intervals greatly increase the
information provided with most statistical analyses, however, I'm not sure
how meaningful they are when a one-tailed test is being performed. I believe
that in a 1-tailed test, most of the relevant information is available in
the effect estimate and the p-value. The effect estimate gives you the
difference from the null and the p-value gives a measure of the strength of
association. Shouldn't this be enough to draw a conclusion from a 1-tailed
test?

By the way, thanks to all for your examples. I'll be sure to use one. I'm
sure you'll all see where if the editor doesn't bury it.

Regards,

Stephen Lyman, Ph.D.
Director of Epidemiology & Clinical Research
American Sports Medicine Institute
1313 13th Street South
Birmingham, Alabama 35205
Office: 205-918-2132
Fax: 205-918-0800
e-mail: StephenL@...
web: www.asmi.org