How good can we expect a forage or feed test to be?

Figure 1 shows that laboratories with standardized instruments reading the same samples have errors averaging 0.1 to 0.4 percent, depending on the variable measured.

Figure 2 shows that NIRS consortium laboratories tend to have errors of less than 1% for all constituents measured which is considerably less than laboratories nationally outside the consortium (NFTA).

Why do samples results vary between laboratories?

· The largest variation arises from taking the sample, no matter how carefully split. When comparing laboratories always retrieve the sample from one and send it to the next.

· Wet chemistry techniques vary among laboratories outside the Consortium.

· NIR results are often based on wet chemistry of an individual laboratory and will reflect any wet chemistry differences.