In Elements of Programming Interviews in Python by Aziz, Lee and Prakash, they state on page 41:
Insertion into a full array can be handled by resizing, i.e., allocating a new array with additional memory and copying over the entries from the original array. This increases the worst-case time of insertion, but if the new array has, for example, a constant factor larger than the original array, the average time for insertion is constant since resizing is infrequent.
I grasp the concept of amortization that seems to be implied here, yet they seem to imply that in other cases, a newly allocated array could possess a constant factor smaller than the original array. Is that so? What does "constant factor" mean in this particular context? I'm having trouble understanding what's being said here.