An Interaction Model for Resource Implement Complexity Based on Risk and Number of Annual Moves

Note

Variables proxying risk: growing season (GS), natural log of effective temperature (LET), natural log of above ground productivity (LNAGP). GS is considered by the authors to be the most direct measure of the impact of annual climatic variation. Variables proxying implement elaborateness: total number of technounits (TTS), average number of technounits per subsistant (TTS/STS), total number of technounits summed over the most complex implement per implement type (MXT), average number of complex technounits per complex subsistant (CTTS/CSTS). Out of four regression models run on each elaborateness variable (for a total of sixteen models), LNAGP is included in the backward stepwise model against TTS/STS, LET is included in none of the models, and GS is included in every model except for the backward stepwise model against TTS, the backward stepwise model against TTS/STS (but LNAGP is included), and the forward stepwise and lasso models against CTTS/CSTS (in which no variables were fit). As CTTS/CSTS did poorly in matching any variables, it was rejected by the authors as a response variable, so risk therefore seems to have a strong relationship with implement elaborateness.