From Data to Design (41216)

From Data to Design (4/12/16) Methods      Flow models (Beyer and Holtzblatt)      Critical Incident Analysis           great for designing systems that improve on existing solutions, incremental changes      Conversation Analysis/Linguistic Analysis  “I can’t talk now, I’m in a fitting room"      Fixed Coding (dedoose, Atlas.ti)            inter-rater reliability tends to be low; meanings of tags are different for different raters.       Affinity Analysis of Researcher Insights      Open/Axial Coding / Grounded Theory Analysis

Grounded Theory Analysis      comes from anthropology (Kawakita)      logical, linked category      “let’s look at what’s really going on, and see if we can build a theory from that."      Popularized in HCI by Beyer and Holtzblatt Three levels

1.  Items (post-its)
2. Patterns (groupings)
3. Constitution (story)

How to evaluate results      Establish trustworthiness (validity)           compare/contrast within team           triangulate with other data methods           dialectical interpretation           inspection of results by professional community      Establish usefulness            design ideas, guidelines, what your team thinks, what other teams think      Establish timeliness           insights per day spent (in industry, it’s typical to spend one week on generative research)