The role of indelcost: OM, LCS and Hamming

How do you parameterise Optimal Matching analysis of lifecourse sequence data? A lot of ink has been spilt on how to build the substitution matrix, defining the differences between the states through which sequences move (the cost of substituting elements), but what about the role of indelcost, the cost of inserting or deleting elements? Here is a quick demonstration (using SADI, in Stata) that shows that below a certain threshold OM becomes LCS, the longest-common-subsequence distance measure, that above another threshold (less well-defined) it reverts to Hamming distance, and that in between there is a minimum meaningful indelcost.…