Thursday I met with Shigeki Amitami, who is a post-doctoral research assistant in the Creativity and Cognition Studios. I was very fortunate to meet with him for over an hour - and he provided me a very illuminating insight, not only to his own research, but regarding the more general area of PhD-level research.

His research question - 

How can the computer be used as a creativity support tool?

How to design such tools, and what methods to take?

* Constructive approach - building software tools that are free for people to intuitively gain control and mastery over.

For his masters degree, he did a similar type of research, towards a support system for musical compositions. He worked in collaboration with a company on commercial exhibition installations.

At CCS he has been working on a system for making video sequences and generative systems.

* He is using a cognitive science approach borrowed from Sloboda (1985) and Tanaka (1990). Sloboda is a music psychologist. 

* Shigeki is also very keen in using protocol analysis. He showed me a book that is pretty much the bible on the topic by Ericsson and Simon (MIT press). I want to see if our library has this.

He discovered from his masters research, that company's often used really generalized forms and questionnaires to gain feedback form users/visitors that would pretty much arrive at 'did you like the exhibition / did you not like the exhibiton?'This information is obviously not very good for any kind of iterative and evolutionary improvement building for proceeding exhibitions. 

Instead, using protocol analysis, Shigeki is able to fashion a framework towards iterative designs that build upon the user feedback. This system follows the user and also incorporates video, to assess their thoughts as they are experiencing the design. Apparently there are two camps for protocol analyses, or two branches of approaches (often people use both) - where you have the viewer think aloud during their experience, and another where you ask them to recount from memory immediately following the experience. 

We then had a great conversation about having general/precise research questions, and strategies for making sure that at any point in the dissertation, one is able to thread back to the main question and there will be an obvious link. I drew up a diagram of this process.

A more provocative methodology, the Knowledge Nebula Crystallizer, is an amazing concept of designed emergence applied towards experimental epistemology. This concept was devised by Shigeki's previous supervisor in Japan.

Shigeki also suggest I speak with Zafer Bilda, another member of CCS who is something of an expert of Protocol Analysis.