Comparing Standard Errors, t-Statistics, P-Values, Logistic Curve Fitting Exercise

Question

I have two growth curve data sets, A (Martians) and B (Venusians). Data point sets of age (0 (birth) - 250 months, X axis) against height (0 - 200 centimeters, Y axis). The first set (A) contains 67 X Y point pairs, the second set (B) contains 27 point pairs. I have fit both data sets to my favorite version of the Logistic Equation using NonlinearModelFit. NonlinearModelFit returns estimates for my two independent variables: Increment, (N0), and Time Coefficient (k). Then following I invoked "ParameterTable" calculating: (1) Standard Errors (2) t-Statistics and (3) P-Values for both of the curve fitting exercises, Martians and Venusians. Of the three Parameters: Standard Errors, t-Statistics, and P-Values, which parameter indicates a better fit to an energy conservative logistic equilibrium? Standard Errors on the calculated Time Coefficients (k)? t-Statistics on the calcuated Time Coefficents (k)? My question, is growth on Mars more of an energy conservative mechanical process than growth on Venus? Are data sets with different numbers of point pairs directly comparable on Standard Errors, t-Statistics and P-Values?