Attending Susan Stafford, Jenny Orr, Judy Cushing, Margaret Burnett, Mike Bailey, Patrick Wingo, Nik Molnar-Stevenson, Chris Schultz,
Introductions
Objective–>For Margaret to look at vis software and prioritize changes to user interface
Comment 1: Margaret’s current research question: Gender Mag (how problem solving software–> when it’s not genera inclusive;)
Judy’s commentary on context: for experts.
–> usefulness of visualizations are important,
–> collaborators want to inherit the software
–> software will live in the public domain, we want to make this software more usable by folks, in general
–> our collaborators are thinking about making the visualizations, themselves, accessible by their non-scientist stakeholders, as well as their own scientist collaborators
–> scientists want to run the models and have the visualizations show in real time what’s happening in the models
Margaret’s questions: the collaborators are scientists (they are not here though), are they here?
Characterized example user (Dominique)
—–> Dominique will inherit this; integration of domain sciences; terrestrial ecosystems, etc.
—–> Questions about Dominique–> what did she do when she didn’t know? She asked (i.e., is she a tinkerer or tinker?)
—–> Active User (learning about content, not about the actual tool)
—> asks about risk aversion and Dominique
—> asks about problem solving (information gathering style, what is it? –> depth first information processor; comprehensive information style first (gather information about a lot hypothesis)
IMPORTANCE of thinking about users in particular…
Introduction to demo for the ViSTAS development
–> JUDY: 3 separate applications (VISTAS, taking files in and visualizing them, versus the vis engine and put it into another program)
2D information on a 3D surface (the topography can be readily seen); for stakeholders–> recognizing the place
Jenny’s demo: VALCEX
–> wind data; wind speed and direction at various places above the point
–> the advantage of the web is that it’s all accessible to folks
–> points out problems in interface; rotate, zoom in, etc.
–> shows the actual two stations being visualized, loads data; (Speeding up the loading is an issue).
Margaret asks: “If you have a collaborator who hasn’t used the tool, what button will they push? Will they know what to push?
Jenny replies: “No, there needs to be some direction and explanation from an expert user; feedback is important
Margaret asks: “How are you recording these issues?”
NIK: source code depository and issue tracker (GIRA);
Margaret: questions: why is start from the beginning the third thing in the list?
questions: what types of questions are the users trying to answer? (e.g., change over time, direction, pattern detection)
Recommendations on VALCEX
–> SEQUENCE THE STEPS WITH NUMBERS AND EXPLANATIONS AND BUTTONS
–> arrows need to be seen (group discusses arrows issues)
–> fixing the blue axes to a different color or outline them
–> for active learners, it hooks them to get to the task at hand
–> Jenny requests a time line (change the log so that it’s more visible)
–> recommend a way to create an anchor of what’s been seen before, instead of need to “memorize” what was seen before.
–> keep using the “active learner” as an example user, that will help to test potential changes
Nik’s Demo: Stand-alone VISTAS application
–> general enough, that it has some flexibility; but it’s a bit more complex because one has to configure vis that one wants to see
–> (documentation or training is necessary)
–> concept of a project is a way to structure
(Nik walks through and volleys questions)
Recommendations on stand-alone VISTAS application
–> visualization could exist without project (start without having a project)
–> bypass thinking about organizing files, and the program can automate the organization
–> change to ‘ADD DATA’ (instead of add data to project)
–> Jenny suggests–opening screen with 3 or 4 steps walking the user through the process; splash screen with “open recent projects”
–> questions whether user needs to know the difference between the types of data necessary for building the visualization; recommends doing a 1a, 1b approach on splash page so that the user doesn’t have to wonder what to load next (check them off till all the data is in there).
–> Margaret draws on the board linear steps and buttons; with potential for moving throughout the steps iteratively; (using back, ‘more data’, and next)
–> questions (once vis is on the screen) — think out loud : does the vis that pops up answer the question I’m asking?
–> inaccuracy versus exaggerated (to be able to see, needs to be exaggerated; Margaret doesn’t like this)
–> needs a way to show overview or composite of all data so that memorizing is not necessary (e.g., volume extrusion or line graph)
–> Margaret (asks “what just changed?”) states: “a lot times, people ask ‘what just changed?'” People need some sort of feedback (transitions, needing to see what happened before–eye to see what’s changing)
Susan summarizes: some of the things that need to be fixed are consistent among all the programs;
Margaret: “if we can think about the problem these folks want to solve, that’s what tells us ‘when we’re there.'”
Patrick’s Demo: ENVISION
description of ENVISION USE–> stakeholder engagement; communication tool
–> difficulty with white on light blue
Recommendations about ENVISION
–> WHO WE CARE ABOUT: the scientist and the stakeholders
–> what’s the use? discovery or engagement?
–> rename “path preset” to something like fly through (Margaret questions naming)
–> How does the user know what he/she is looking at? (disorientation of user; the legend needs to be built into the visualization; ON THE LIST)
–> Questions use of the word “tree”
–> Susan: using this to communicate with another audience; questions –> scientific audience (how does this work?)
–>
JUDY: problem: how to integrate into an existing package the features that VISTAS has (overlaying the 2D data onto a 3D visualization); manipulate by a scientist, etc.
“The UI task is to add the visualizing with the technical interface of ENVISION”