Radu Luchian(ov): Pastime: Commentary
On Expertise

Notes around my Master Thesis
WARNING: area under construction; wear yer protective gear and stay alert

During the design part of my experimental work I always reach the point where I have to decide what the target population is for the experiment and how I'm going to sample it. Most people simply skip this step outright or pick the handy student population (with disastrous results for the external validity of the results), on the premise that "we're all created equal." Well, yeah, it is morally and ethically correct to assume that from a social point of view we should have equal rights. But that does not mean that we are born and develop the same way, nor that our behavior (social, cultural, emotional, rational, creative, etc.), is the same from one day to the next, let alone across people.

To lay some of the terminology I'm using: observed behavior can be formalized in an action-drama paradigm. There is an action performed that involves actors and props (with a blurry distinction between the latter two, distinction made through the definition of the specific action). The actors are the entities (people, animals, animate natural objects or artifacts, energy fields etc.), who can determine the course of the action by performing parts of it or reacting to parts of it. The props can be any entities whatsoever that the actors can use (manipulate, reshape, combine, extract, etc). Specific actions can be formalized by specifying how the actors are using the props. Directed behavior (which is an inferential interpretation of observed behavior, not necessarily a correct interpretation!!), furthers the definition of an action to include a goal that the action is intended to achieve. (Philosophers tend to get lost in the woods after this point so I won't continue :)

Now, we suppose (through generalization from introspection - ouch!), that actions are guided by the knowledge gathered within the actors (in the actor's mind, reflexes, conditioning, database, algorythm, mechanism, etc.) Knowledge is knowledge (that is useful, leading to the goal of the action) only if it applies to the current situation (incidentally retrieval from memory is a topic still mysterious in CogSci.) So we functionally split up knowledge in domains of applicability (in short, domains.)

As far as I can distinguish now, we can split actors into novices (people who are in the process of learning the action) and experts (people who are confident that they have internalized the essentials of the action). We can further refine the expert category into generalists (people who perform a large number of actions spanning diverse domains) and specialists (people who tend to repeat the same few actions very often and ). Now, their needs and resulting strategies are different and that should be reflected in any good interface:

  • novices need simple interfaces, they need lots of documentation readily accessible
  • generalists need hints in the interface to allow them to stay situated and orient themselves in the current context
  • specialists need as transparent an interface as possible

    Two examples:

  • [ Home ][ Current Projects ][ Portfolio ][ Pastime ][ Pseudonyms ][ CV ]