American Antiquity • Vol/Iss. 73(4) • Cambridge University Press • • Published In • Pages: 599-625 •
By Read, Dwight
The emphasis of a subsistence strategy on either land or aquatic animals will be associated with implement diversity.
Variables proxying subsistence strategy: percentage of aquatic animals in diet (AQU), percentage of land animals in diet (LAN), percentage of aquatic and land animals in diet (TAA). Variables proxying implement diversity: total number of complex subsistants (CSTS), total number of complex subsistant types (NCT), total number of subsistants (STS). Out of four regression models run on each diversity variable (for a total of twelve models), AQU is included only in a backward stepwise regression against CSTS, LAN is only included in the best subset for CSTS, and TAA is included in the backward stepwise against CSTS and in the Lasso regression against NCT. Once interaction effects are considered, these variables are ruled out as having a significant effect on implement diversity.
|Stepwise regression, best model analysis, Lasso regression, linear regression||Not supported (3 out of 12)||UNKNOWN||UNKNOWN||UNKNOWN|
|Variable Name||Variable Type||OCM Term(s)|
|Total number of complex subsistants||Dependent||Tools And Appliances|
|Total number of complex subsistant types||Dependent||Tools And Appliances|
|Total number of subsistants||Dependent||Tools And Appliances|
|Percentage of aquatic and land animals in diet||Independent||Food Quest|
|Percentage of land animals in diet||Independent||Food Quest|
|Percentage of aquatic animals in diet||Independent||Food Quest|