| http://www.w3.org/ns/prov#value | - For example, one can check how robust the trend is to such things as changing the start or end years by one or two years, changing which data set you use (GISS or HADCRUT, for example), or eliminating one year of data. (For example, if your entire conclusion changes if you eliminate the year 1998, then the conclusion is probably not very robust!) I have done these sort of checks for the GISS and H
|
