Within this paper, the term mutation rate refers to the parameter mutation rate rather then the effective mutation rate. When dealing with an alphabet size larger than 2, the typical technique is to replace a given symbol with a randomly selected symbol from the alphabet. This leaves open the possibility that the randomly selected letter will be the same as the original letter thus resulting in no actual change. The parameter mutation rate is probability that a given letter will be replaced in such a fashion. The effective mutation rate is the probability that the change will actually produce a different string.
Many evolutionary strategies or related algorithms make use of binary genomes. In this case, a mutation usually defined as a flipping a bit rather then changing a bit to a random value. In that case both the parameter mutation rate and effective mutation rate are the same. Thus, in most cases there is no need to distinguish between them.
However, some simulations make use of a non-binary alphabet. This includes WEASEL, Avida, and Ev. Although not explicitly stated, it is our understanding that all of these simulation use mutation rate as used in our paper. In our particular situation we are concerned with recreating the same algorithm as Dawkins originally used. As such we are interested in the mutation rate Dawkins would have selected to put into the program which would have been the parameter mutation rate. As such, the effective mutation rate is simply not as interesting.
