Investigations are, at heart, re-representations of complex phenomena in places and at scales that allow us to study them. Scientists have to decide what aspects of the phenomenon to represent and how to represent them, even before they are entirely sure what is important.

Developing an investigation involves deciding what to compare and for what reason. This includes identifying and controlling variables, considering whether a comparison will provide information, and considering scale (e.g. how much to vary a factor to decide if it makes a difference).