KARMA is a quantitative model which can be used to predict a user's acquisition, retention, maintenance, and transfer of system knowledge. In KARMA, three stages over the course of an operator's model (see Schumacher & Czerwinski, in press) are characterized: pretheoretic - the understanding of system structure centers on the retrieval of superficially similar instances in memory; experiential - some understanding of causal system relationships emerges, as does some reliance upon structural system information. Also, abstraction begins to occur across different, but similar systems; and expert - at this stage, the user makes abstractions across various system representations stored in memory; full reliance upon structural system knowledge emerges. Currently, a programmed simulation is used to compare our theoretical predictions and assumptions with experimental data. The simulation is running in Scheme (a dialect of LISP).
/lp/association-for-computing-machinery/karma-knowledge-acquisition-retention-and-maintenance-analysis-FWg56EGcZ6