Reader Comments

Post a new comment on this article

Minor corrections in Supplemental Text S1 and Supplemental Table S10

Posted by nmyhrvold on 23 Dec 2013 at 22:30 GMT

In reviewing the published supplemental materials, I noticed three minor corrections that need to be made. Two corrections involve typographical errors in Supplemental Text S1, and a formula should be corrected in Supplemental Table S10.

None of these errors impact the analysis, the outcome, or the scientific understanding of the article; the first two are strictly typographical errors that appear only in the manuscript, and the third correction will increase the sum of squares across all of the regressions (original and new), but does not change the conclusion that my regressions have lower sum of squares error, and thus are better fits to the data than the previously published regressions.

The error in Supplemental Text S1 involves the depiction of the logistic equation used in my reanalysis of references 20 and 25. The logistic equation is given as having the form:

m(t) = a/(1-exp(-b(t-c))) + d

The correct functional form reported in the original publications and used in my reanalysis is:

m(t) = a/(1+exp(-b(t-c))) + d

This typo was depicted in the following functions:

mA(t) = a/(1-exp(-b(t-c))) + d;

mass = 5551/(1-exp(-0.57(t-16.1))) + 5;

mB(t) = a/(1-exp(-b(t-c))) + 5;

mA2(t) = apub/(1-exp(-b(t-c))) + d;

mB2(t) = apub/(1-exp(-b(t-c))) + 5;

mC(t) = a/(1-exp(-b(t-c))); and,

mC2(t) = apub/(1-exp(-b(t-c))) + d.


The correct functions that I used in my reanalysis are:

mA(t) = a/(1+exp(-b(t-c))) + d;

mass = 5551/(1+exp(-0.57(t-16.1))) + 5;

mB(t) = a/(1+exp(-b(t-c))) + 5;

mA2(t) = apub/(1+exp(-b(t-c))) + d;

mB2(t) = apub/(1+exp(-b(t-c))) + 5;

mC(t) = a/(1+exp(-b(t-c))); and,

mC2(t) = apub/(1+exp(-b(t-c))) + d.

Additionally, the following sentence in Supplemental Text S1 contains a typo:

“The daily growth rate implies a divisor of 375 days/year which gives an age under Wells data of 112.3 mya; under Tenchov it is 1.5 mya.”

It should read:

“The daily growth rate implies a divisor of 375 days/year which gives an age under Wells data of 112.3 mya; under Tenchov it is 1.5 bya.”

Finally, in Supplemental Table S10, the formulas that sum the square of errors for each regression (original and new), appear to exclude the square error for the first data point in each data set. The current formulas start with row 10, and should begin at row 9 in each tab of the table.

No competing interests declared.