“…Where the population under consideration is small, measurement of its spatial concentration into relatively large units could well under-estimate the degree of segregation. Chicago, for example, had only 29,583 Blacks in 1880 (out of a total population of 1,692,793), and its 18 wards averaged nearly 30,000 residents, whereas the EDs averaged fewer than 1,500; at that date the index of dissimilarity at the ward scale was 0.648; at the ED scale it was 0.694 (Logan et al, 2015a(Logan et al, , 1070. A further problem with a small population is that there will almost certainly be few of its members in many of the spatial units, in which random allocation processes may well lead to over-inflation of the degree of segregation -as clearly demonstrated by Carrington and Troske (1997).…”