Experimental methods are used across science including, most relevent for HCI, psychology. Experimental methods usually try to vary a single variable or small number of variables at a time keeping all other things constant, for example, looking at how human reaction times depend on the pitch of a sound. Typically there are a small number of independent variables the things being deliberately changed, several dependent variables, the things being measured, and everything else being kept constant.
In HCI (and other areas such as nutrition), it can be hard to vary one thing without changing others at the same time, for example it is hard to have exactly the same interface and only modify consistency. Perfect control is also difficult, as human participants come with a wide range of experience, personality and other factors and you cannot test multiple conditions on the same participant without potential order effects. Often conceptual conditions (such as consistency) are embodied in specific systems, so it can be hard to disentangle whether results are due to the condition or the specific instance in a piece of software.
Used in glossary entries: consistency, reaction times
