Dear All,
I remember global and local Moran's I in the range [-1,1].
However, Tiefelsdorf and Boots (1997) indicates the feasible range is (-n)⁄((2√(d_i))) , where d_i is the local linkage.
Tiefelsdorf, Michael, and Barry Boots. 1997. A Note on the Extremities of Local Moran's Iis and Their Impact on Global Moran's I. Geographical Analysis 29 (3):248-257.
Some people aslo criticise that [-1,1] is wrong.
I am so confused and spending couple of days to find references. Surprisingly, quite few indicate [-1,1]. Which one is true?
I understand there might be some extreme values in usual cases.
Appreciate if anyone can help.
With best regards,
Zhenshan
--
Assistant Professor,
Institute of Geographical Sciences and Natural Resources Research,
Chinese Academy of Sciences