Home : : Research : : Direct Manipulation 
Direct manipulation - giving a testable definition
By Fred Voorhorst (Zürich, 2000).
Direct manipulation is an ill-defined concept. It belongs to the category of quality, intuitiveness, ease of use and so forth. The typical 'every body knows what it is when they see it' type of criteria. Because it is ill defined every body feels competent to formulate an opinion about it, and those who actually are competent have no real measuring stick to help guide development.
On this page I will present some examples showing how concepts like direct manipulation and ease of use have been driven more from a technical point of view and less from a user's. At the end I propose a measurable definition for ease of use and direct manipulation.
Qualification of the user interface is driven by hardware development. At the time of a keyboard based systems were common, the mouse-based pointer was considered 'direct manipulation'. The pen swiftly followed this, when pen based systems became fashionable. At the moment the virtual reality systems are topic for research and are the modern synonyms for ease of use and direct manipulation.
Although technology improved, the interface did not. The interface seems to be designed to fit the technology rather than the task or user. For example, Virtual reality systems, with the ability to react to any possible user action, still ask the user to 'push a button'. Research recently caught up to the fact that humans actually have two hands, and hopefully soon will find out these hands is skilled in detailed manipulation.
To improve usability and direct manipulation you need to measure that what hinders these: control over the user interface. The more time and effort the user has to invest in controlling the interface the less time and effort the user can invest in completing the task.
I therefore propose to use 'time spent in controlling the interface' as measure for usability. Usability (for a specific user) is optimal when the task is completed without ever having to find a button, search for a function or even having to look at how the mouse is oriented.
Consequence of this definition is that on the one side usability becomes measurable, and on the other side that it becomes decoupled from the hardware implementation; instead it is coupled to the relation between task at hand, the user's skill and the hardware. A keyboard becomes perfectly usable to a professional typist having to type out the minutes on a Dictaphone. A mouse does not.
Kees Overbeeke - Interaction design & Affordances
Tom Djajadiningrat - Tangible interaction
Stephan Wensveen - Emotional design
John Flach - Ecological system design and cognitive engineering
Douglas Adams
Caroline Hummels - POP, for insiders, expert in Wizard of Oz studies
Bill buxton
Jan Koenderink
Bill Gaver
www.brusino.com