Experimental Error 3-1 Significant Figures

Experimental values for selected ionization constants are given in Table 23-1. Although some of these values, such as the constant for acetic acid, are reliable to at least two significant figures, keep in mind that others may be in error, some as much as tenfold. For example, the dissociation constant... 

The stretching of the N-H covalent bond in N-H—O systems, involving N-H bond lengths of 0.95-1.10 A, is not very significant (Figure 5), in comparison with the O-H—O system, 0.85-1.20 A (Figure 3) or in comparison with experimental error. The reason for the small variation in N-H bond lengths is probably that there are no or few exactly-matched structures reported in the CSD. We noted a difference... 

Perhaps the most common failing in reporting AS values is to list them with more significant figures than the experimental data justify. It is for example common, but hardly ever justified, to list AS values to 0.01 e.u. One can assess the problem by considering quantitatively the contribution to errors in the Arrhenius parameters.2 Considering... 

The primary experimentally determined data in Tables 1.2 and 1.3 are the isotopic ratios and volume fractions in air and the total mass of air these are tabulated with their reported uncertainties. Quantities derived from these are shown without error limits in Tables 1.2 and 1.3 for reasons cited earlier, they are stated to more significant figures than are justified by the precision of the primary data from which they are calculated. 

---