Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

What is "User Modeling"

User Modeling - trying to make sense out of a large data set

During a Contextual Inquiry project your team will amass a large collection of raw data on your area of investigation.  But after you have all this data, what do you do with it?  How do you make sense of it? 

During User Modeling the Contextual Inquiry team gets together and reviews the data looking for patterns.  The outputs of this modeling will be the creation of personas, goals, tasks, and scenarios.  There are a number of approaches to doing User Modeling which are outlined below. 

This wiki page is meant to be a very practical, hands-on, how-to guide.

How To Do User Modeling

During your contextual inquiries you gathered a lot of data on your users and their use contexts.  This data is now stored somewhere and you need to analyse it and make sense of it.  Here are the steps you need to take to get this seemingly overwhelming task accomplished.

Pre-User Modeling 

  • All raw data from each CI is stored in a common location
  • The CI teams have met throughout the CI process to discuss high level findings.  You want to make sure going into your User Modeling exercise that your team is primed on the data.
  • High level summaries of the data have been created.
  • Your team has a good understanding of the design/problem space and what categories or questions you are trying to find answers to
  • You have gathered your team in a room with white boards, post-it notes, flip charts, all the data, coffee and lots of time
  • Have team members write their observations onto post-it notes before attending the User Modeling sessions - this will save lots of time

During User Modeling

  • Gather your team in a room with lots of coffee and food - you'll be there a while (smile)
  • Make sure the room has lots of open wall space
  • Realistically User Modeling will take a few days depending on how much data must be reviewed
  • Begin by reviewing the research goals the CIs are trying to address
  • If for example you had several "roles" of users you observed, focus on one role at a time, say "instructors".
    • Have each CI team give a brief overview of their high level findings
    • Next have the teams start to place their observations (written on post-it notes) on the walls in the room
    • After all teams' post-it note observations are on the walls, have the teams go around and try to move the post-its into "categories"
    • if an observation belongs in more than one category, have the person write it on another post-it note.  Don't remove from someone else's category
    • When all team members have done their "walk around" to organize post-it notes, you can move onto the next step
    • Now have team members try and name the categories they are seeing for each category.  A category can have more than one name.  This is where "patterns" begin to emerge.
  • What are the categories?  Are they a group of "Tasks" or a "User Scenario"?  Who is doing the task/user scenario?  These are your personas.
  • Create lots of personas at first and then look for things in common.  If two personas have similar goals and tasks then they can be collapsed into one persona.  You want the minimal number of personas needed to span the design space.  But to get there, its easiest to start with too many personas and reduce them down.

Post-User Modeling 

  • Share out with your community the Personas your team has created
  • Share out with your community any User Scenarios your team has created
  • Are there still any gaps of understanding you have?  Are further CIs needed?
  • No labels