Benefits of Ecological Interfaces Under Equivalent Sensor Sets

Critics of Ecological Interface Design (EID) laboratory studies have argued that the experiments are confounded by non-equivalent sensor sets reflected in the contrasting interfaces. Whereas some EID proponents have responded by emphasizing theoretical justifications for the interfaces employed, the critique is grounded in practical challenges that endure in human-machine interface research. We construct an experiment from data that includes an interface that removes the sensor set confound. The results are surprising. Adding additional sensor information to the data-impoverished non-EID interface results in poorer task performance and no improvements in control stability. Participants using the EID interface exhibited superior task performance and control stability to those in two non-EID conditions. This result speaks to the generalizability of EID research findings to industry applications.