Merge Blocks

A merge block consists of a number of columns which have been merged together so that details have been lost of the entries in the individual columns but the total number of entries is still known.

A merge block might arise due to failure of some machine to operate properly at one or both extremes of operation, for example because it was being asked to do something it had not been designed to do. Alternatively, it might be part of a data-collection routine, where observers have noted instances of '6 or more'.

In practice, when data is collected in batches, by different researchers or for different purposes, it sometimes happens that some batches contain a merge block but some do not. If this happens then merge blocks can overlap the precisely-known columns, and might be better thought of as being blocks of missing data. Some machines might fail, some might not. Some observers might note '6 or more', some might record the detailed figures.

Although it might seem that a merge block is to do with the data, it is actually a modification of the underlying set and has to be treated as such. The existence of a merge block does not affect the degree: all of the above have a degree of 8.