Hi Anthony,
You have made an excellent point. If one does not know the context of the data (hence why I called it orphaned data), one should be wary of particular "models" to fit to that data and the assumptions used for that model, since for example I have shown your children's feet fit a perfect cubic equation (with the assumption of x being a whole integer), based upon the original data set. Though, one can fit a quartic equation through the complete data set perfectly (another lesson in using unsubstantiated polynomial fits: the larger the degree of polynomial, the better the fit!)
And also this shows the importance of complete data. Only by knowing child X's shoe size does it make it plane that a particular model (the cubic) wont fit.
Excellent exercise!
Jonny
The path to good scholarship is paved with imagined patterns. - David M Raup