First, one important difference is the biochemical processivity, determined by comparison of steady state ATP turnover and microtubule-induced ADP release [34]. From the steady state ATPase parameters it follows that kcat/K0.5,Mt of the Head3/Neck1 chimera is 23.2/0.2 µM−1·s−1 = 116.0 µM−1·s−1 (Fig 3A and Table 1). The fast phase of the microtubule-induced mant-ADP release rate in stopped-flow assays showed kmax/K0.5,Mt = 52.5±s−1/1.7 µM = 30.9 µM−1·s−1 (Figure 3B). The slow phase is probably due to background drift. The ratio of these values suggests a biochemical processivity of kbi,ratio = 116.0/30.9 = 3.8, indicating that Head3/Neck1 remains attached to the microtubule filament for 3–4 catalytic cycles. This is more than the NcKin3 reference that detaches after one ATP turnover [16].
http://plosone.org/article/info:doi/10.1371/journal.pone.0004612#article1.body1.sec2.sec4.p1

Important elelements of error analysis are missing from this section and reference 16 (Adio et al. 2006) that make it difficult to assess the statistical significance of the apparent increase in the biochemical processivity.

First, uncertainties are not consistently reported in one location. But it does appear that all measurements with uncertainties can be obtained by collecting from Table 1, Figure 3, and this paragraph.

More importantly, the nature of the uncertainties are not specified: are they standard error of the mean? Assuming that they are, and that they represent 68% confidence intervals, then it is important to discuss how significant the increase is compared with the biochemical processivity measured in ref. 16. Especially since the uncertainty from ref. 16 is not clear, but likely very high.

As a clear example, the steady-state K0.5,MT is reported at 0.2 +/- 0.12 micromolar. If this represents a mean +/- standard error of the mean then the resulting 68% confidence interval for biochemical processivity would at best span a range of a factor of 4 (from 2.4 to 9.5). (The assumption of normally distributed error, is probably not correct, though.)

It's tough to say what the uncertainty of the processivity in ref. 16 is (the uncertainties are quite large). But it's likely that the 68% confidence intervals from both references overlap and thus it cannot be strongly said that the processivity of the Head3/Neck1 increased significantly.

It's also possible I am misinterpreting the meaning of their uncertainties...but this would need to be clarified by the authors. For example, if they reported standard deviation as apposed to standard error of the mean.