In lieu of an abstract, here is a brief excerpt of the content:

  • Errata

Volume 80, Number 4 (August 2008), pp. 377391, " Surname Analysis in Biological Anthropology: Alpine Populations in the 17th and 18th Centuries," by M. Prost, G. Boëtsch, M. Girotti, and E. Rabino-Massa

In the original Table 4, the values related to matrimonial isonymy (Fiso) and its components were outside the range of typical values because they were expressed as percentages. Because they were classically presented as frequencies, we have adjusted the table accordingly. As a consequence, all the values in the article must be considered according to the new table. Finally, we highlight that all calculations have been verified, and we confirm the accuracy of the data and all the conclusions achieved in the original paper.

The new Table 4 is as follows:


Click for larger view
View full resolution
Table 4.

Evolution of Matrimonial Isonymy (Fiso) and Its Components (Fr and Fn) in the Chiomonte and Vallouise Populations (CV%)

[End Page 501]

Volume 72, Number 3 (June 2000), pp. 447454, "A Look at the Inheritance of Height Using Regression Toward the Mean," by T. Clemons

[Colin Steer, writing on behalf of Tracy Clemons, submitted the following errata.—Editor]

I thought Dr. Clemons's article was an interesting application of adjusting for "measurement error" due to nonpaternity. However I think there is a typographical error in Eq. (13), which should read

assuming that the covariance between biological and nonbiological fathers σcf (using her notation) is 0. This was stated on p. 451 and is implied by Eq. (9). Of course, in one of the scenarios Dr. Clemons considered, where a = 1 - p, the revised equation and the published equations are equivalent.

I think it is also worth reemphasizing two common pitfalls for readers attempting to use such techniques. First, the observed covariance between parental heights is σxm, not σfm, although the two are related: σxm = aσfm. Second, as Eq. (12) tells us, we need to apply Λ-1 to the regression coefficients from OLS to obtain "unbiased" (or at least consistent) estimates of the parameters.

C. D. Steer
University of Bristol

...

pdf

Share