Originally Posted by Roody
Body tissue composition measures lack even the first qualification for being a valid measure. They are not reliable. That means they will not always give you the same value when you repeat the measurement with the same method. They will also not give you the same number if you use two different methods to measure the same person. Without reliability, you cannot have validity. OTOH, Weight measurements are highly reliable, and height measurements are almost totally reliable.
But you're confusing things here...
Sure, height and weight measurements are pretty tought to mess up, but the derived information isn't necessarily useful.
Think about it... What if we measured foot length and multiplied that with head circumference? Sure, you'd have a very repeatable measure but it would (probably) have no usefulness whatsoever!!!
On the other hand, a body fat measurement, even if off by as much as 2-3 percentage points between two measurements using the same method on the same person, it's still a very useful measure, since the healthy and unhealthy ranges overlap for individuals and are also fairly broad. If a an adult man measures as having 14% body fat, or 17% body fat, it still provides a good indication of healthy body fat levels.
A measure that has a high variability but a high degree of usefulness is much more valuable than a measure with negligible variability but almost no usefulness...