Mental Models for GIMP

From GIMP GUI Redesign
Jump to: navigation, search

Executive Summary

MMG (Mental Models for GIMP) is a user research project for GIMP, the free software image manipulation program.

It aims at better understanding professionals and aficionados in this field by eliciting their context, tasks and philosophies in order to improve the overall usability of GIMP.

Research will be carried out following Mental Models, combining qualitative interviews with content analysis and a graphical format for reporting, as proposed by Indy Young, creater of this method.

Results will provide guidance for interaction architecture and software development.


  • Provide further information and understanding in the following areas:
    • users ("who?")
    • tasks ("what?", "how?")
    • context ("where, when, why?")
  • Facilitate empathy for users
    • How do users tick?
    • What are their motivators or pressures?
    • What contributes to their satisfaction?

Proposed Method

  • Mental Models by Indi Young
    • qualitative method
    • based on open interviews (no questionnaires)
  • Further information on this method is available, see:
    • The book (rosenfeld media) Mental Models
    • excerpts available for online reading via paperc (German website, free account required, book in English)
    • The author explaining the method in this Video
    • diagrams from the book on flickr

Course of Actions

  1. define audience segments based on product vision
  2. set scope of the interviews
  3. recruit interview partners, see recruiting channels and information for interview partners
  4. conduct interviews, see interview partners
    • at least 4 interviews per audience segment
    • focus: tasks, behaviours and philosophies of users
  5. transcribe interviews
  6. analyze results
    • distillation of tasks, behaviours and philosophies into patterns
    • create graphic of patterns in a mental model
    • oppose mental model vs. existing functionality
  7. make use
    • gap analysis (planned and existing functionality vs. users' behavior, needs and tasks)
    • adjust user scenarios / audience segments
    • derive stragety for development
    • prioritize features (importance vs. feasability)


  • The product vision is used as a starting point to find the right partners (interviewees) for the mental models interviews.
  • Finding the right interview partners is pivotal. Recruiting must be carefully prepared and carried out.
  • Recruitment channels can be, but are not limited to the following:
    • GIMP developer mailing list
    • GIMP users mailing list
    • blogs, forums, wikis on GIMP
    • blogs, forums, wikis on high-level photo manipulation

Ideas for Conducting the Method

  • Embed an intern during a Season of Usability on OpenUsability.
    • Valuable help can be provided by the intern before, during and after research, e.g.
      • brainstorm ideas for optimal recruiting
      • recruit interview partners
      • conduct interviews / parallel transcription while interviewing
      • transcribe and analyze interviews
      • help build the visualization
  • Inquire / brainstorm how the research process and it's results can be made transparent for developers / maintainers / the community.


  • Mental Model Diagram
    • one diagram per audience segment
    • tasks organized in "Mental Spaces" of users
    • justified audience segments based on tasks
    • gap analysis
    • functionality overview will be confronted with users' perspective
  • A primer describing the target audience of GIMP can be derived (e.g. 20-40 page pdf about users, tasks, context; for both new and seasoned developers or interaction / usability contributors)

pros and cons

  • (+) pros
    • fundamental research, broad focus, high-level
    • detailed analysis of users and their tasks
    • research is built on existing findings
    • verification of user scenarios / audience segments
    • based on tasks and behavior
    • adaptable / scalable method
    • supports strategic vision and long-term development
    • revision and clarification of user scenarios
  • (-) cons
    • fundamental research, broad focus, high-level
    • no immediate improvements
    • complex method
    • not all existing findings are considered