Forms attached to a whiteboard, surrounded by information about how they are used

Michigan School of Dentistry X-Ray Information Flow



  • Assess the information flow and usage of the Michigan School of Dentistry’s current practice with film X-Rays.
  • Provide guidance for the transition to digital X-Rays.

Dates: September 2011 – December 2011 (4 months)

Target Audiences: Administration of the School of Dentistry; Dental students

My Role: User Researcher


  • Learned the ins-and-outs of contextual interviewing, ethnography, information modeling, and project management
  • Impressed the client with our understanding of their processes and the thoughtfulness of our recommendations


This was grad school, so we got put into teams. And, of course, the first order of business was to name ourselves.

A logo that says "4 out of 5 dentists" feature tooth-shaped characters, one of which is blacked out
It’s so dumb but in the best way.

Now that we had done the hard work, it was time to start evaluating current practice at the School of Dentistry. At the time, the University of Michigan’s Dental School was still exclusively using film x-rays. They knew they had to upgrade, lest they risk falling behind other dental schools and having their reputation besmirched by matriculating Millennials and rivals alike, but they had no idea how the new equipment would affect their process – or how it should. We started by doing a series of interviews as well as ethnographic observations of the waiting room, student clinics, and the x-ray room.

A spreadsheet with each row pertaining to an individual interview, showing interviewer, interviewee, and date

We complied over 300 affinity notes from these interviews, coding them by User ID as well as tags indicating different sorts of notes. These, along with a collection of photographs that have been lost to time and the dark corners of the Internet, gave us the data we needed to model this information in a number of ways.

A list of notes from interviews, details uinimportant
Click through if you love pages and pages of raw data

We modeled the data in 4 -5 different ways. included here the physical model, showing the physical flow of people and information through the clinics, an artifact model that tied the interview data to the physical artifacts and forms people used everyday, and an affinity diagram that helped us understand needs and patterns.

A model of the flow of information regarding x-rays in the dental school
Physical model. Apologies it’s a bit cut off, this is the best shot I have.
Forms attached to a whiteboard, surrounded by information about how they are used
Artifact model
A white board with multicolored sticky notes, forming a hierarchy of information
Partway through the affinity diagram
A series of post it notes on a wall
One close up shot of the final diagram…
Multicolored sticky notes on a wall, displaying clusters and hierarchies of information
… and another.

The ultimate outcome was a report for the School of Dentistry. This is phenomenally unsexy, I know, but reports with well-research and well-presented recommendations are an essential, if unsexy, part of software design and development. Read an executive summary of the report, if you like.