Reflection-in-Action allows designers to act as researchers in a given context by developing new problem spaces and by adapting both
ends and means. Donald Schön treated the design process as "a reflective conversation" with existing situations which (as Herbert Simon
had proposed in The Sciences of the Artificial) ought to be changed into preferred ones. The interplay between "changing things
and understanding them" does not invoke a self-referential system where practitioners freely interpret the objects they actuate. A
robust reflective practice would be based on appropriate move- and hypothesis-testing in order to dis-/confirm predictions and intended
Iterative design methods can have a generative component, where concept solutions are formulated and artefacts made, followed by an evaluative phase where outcomes are checked against goals, requirements, and criteria. The system may also warrant a redescription before the next design iteration proceeds.