A collection in the GLOBE system consists of a user-defined group of GLOBE cases (case studies). Users can select cases to add to a collection based on:

• a common research theme, such as cases that represent case studies of agricultural intensification (see this example);
• similarity to a reference case, which be generated automatically with a similarity analysis;
• or random selection from the GLOBE database.

Collections are a means to represent, characterize, and analysis the set of case studies used in a meta-analysis. Sets of case studies can be individually georeferenced, their metadata entered into the GLOBE case format, and visualized using the GLOBE system.

Several GLOBE collections representing highly-cited meta-analyses of land change were created by the GLOBE team. Some examples:

• van Vliet et al. (2012)
• Keys and McConnell (2005)
• Turner, Hanham, and Portararo (1977).

Collections are also the basis of a GLOBE representativeness analysis. Representativeness assesses the degree to which a given collection of cases represents an unbiased sample of a specified global extent with respect to a set of global variables selected by the user. Conducting a representativeness analysis allows the user to rapidly identify and potentially remedy geographic biases in the selected cases that might influence a meta-analysis. An example of a representativeness analysis conducted on a collection of cases from a meta-analysis of swidden agricultural change can be found here.