] [Thread] [Date] [Author]

[SCHFUN(--nospam--at)aol.com]

One method that I have used is to extrapolate the results of testing on 
similar defects installed by the same contractor (or sub) directly from the 
results of the invasive testing.  For example, if shear walls are improperly 
installed in 50% of the invasive testing locations, I would argue that 50% of 
the un-tested locations are probably improperly installed.  The problem with 
this analyses is that you don't know which 50% are defective until you open 
them all up.  Therefore, it is common to include the costs of verifying the 
existence of the defect at every location, and then the cost of repairing the 
defect at 50% of the locations.  As you probably can guess, a large part of 
the cost of certain repairs is the cost of getting to the defect, and then 
doing the cosmetic repairs.

Another way to look at the problem is to see if particular subcontractors 
improperly did something, and then try to identify which of the units were 
built by that sub.  If the testing results warrant, you can reduce the 
extrapolation on the basis of who did the work.

Due to the complexities of construction, and the variations between 
construction practices between crews, (and on certain days of the week ?), I 
would be skeptical of any rigorous mathematical procedure to extrapolate 
invasive testing testing results.  It generally comes down to what makes the 
most common-sense, based on all the data.

Hope this helps.


Steve Helfrich