Open Web Mapping: How do we teach this stuff?

Today I gave a talk at the NACIS conference about redesigning the lab curriculum for Geography 575, UW-Madison’s Interactive Cartography and Geovisualization course, to more effectively teach the new Open Web Platform mapping tools that are now the industry standard for web mapping. People seemed to like it.

But the 20-minute (or, in my case, 23-minute…oops) conference talk/slideshow format presents a tradeoff between conveying lots of necessary information and slides that support the talk without textsploding the visual cortex. After paring down to the bare bones of my story, I was left with a number of unavoidably text-heavy slides showing our 2014 curriculum outline and comparing it to the revised sequence for the next iterations of the course. All I could do was point to and say, “don’t read this now, here are the key bits to notice…” People who didn’t see the talk can go back and read the content, but they won’t get the key take-aways I presented verbally (Why is this web mapping workflow thing so important it gets 7 slides? Why are those curriculum topics highlighted in purple?).

So, to fill in the missing pieces of the online slideshow without too much extra labor on my part, here is the outline of my talk. I am in the process of submitting a more detailed journal paper and will post the citation here if it ever gets published.

Open Web Mapping: How do we teach this stuff? NACIS 2015 talk

  1. Title slide
    1. Intro self
    2. Talk based on my experience as a curriculum planner and TA for G575
    3. 4-credit course on cartographic UI/UX design taught by Rob Roth
    4. 2 hr per week lab component
    5. Will talk about
      1. Desired learning outcomes for course
      2. Curriculum redesign to match new teaching technology stack
      3. Evaluation process
      4. What comes next
  1. Desired Learning Outcomes
    1. First two outcomes corresponded to what were initially three lab assignments: animation, sequencing and retrieve interactions, and geovisualization
      1. Eventually merged first two lab assignments into one Leaflet lab without animation requirement
    2. Third outcome informed by final project—collaborative start-to-finish work experience with real world scenario chosen by students.
    3. Fourth and fifth outcomes “soft” outcomes that come with learning tools
      1. Translate across technologies
      2. 4: necessary cognitive development to apply technologies
      3. 5: best practices for design & development informed by research
  1. Technology stack I
    1. Taught with Flash until 2012, then moved to open web mapping technologies as Flash was dumped as industry standard
    2. Flash: encapsulated IDE with seamless integration of design software and well-featured native scripting language
  1. Technology stack II
    1. Open web platform technology stack is bigger and more unwieldy
    2. Job no longer to teach students 1 or 2 pieces of software to make great maps; is to teach integration of dozens of platforms, languages, libraries, frameworks to make great maps
    3. Major unanticipated consequences for teaching techniques
    4. Didn’t put much thought into restructuring curriculum initially; kept everything but lab technologies the same
    5. Results in 2013 were very mixed
      1. Award-winning maps
      2. Many students struggled, especially with learning D3
  1. Web Mapping Workflow I
    1. Rich dissertation: idea of Web Mapping Workflow to describe process of creating a web map on Open Web Platform
    2. Ideally, Workflow should inform scope and sequence of our teaching
      1. Scope: what is taught
      2. Sequence: in what order we teach it
    3. My concept of workflow is adapted, not identical to Rich’s
    4. First step in workflow is to design web map.
      1. Mostly teach this in G572, Graphic Design for Cartography; but included as part of final project
  1. Web Mapping Workflow II
    1. Second step is to set up a development environment—akin to a workshop with hammer, saw, drill, etc.
    2. Initially treated this topic lightly, but adding more structure to it
  1. Web Mapping Workflow III
    1. Third step is to find and format data
    2. Always takes longer than expected—often most time-consuming stage
    3. Proved to be tricky for students; required more attention
  1. Web Mapping Workflow IV
    1. Fourth step is creating basic Markup that forms backbone of the web page
    2. Can be seen as the space for cartographic representation—where the actual elements of the web map exist
  1. Web Mapping Workflow V
    1. Fifth step is script
    2. Fourth and fifth steps highly integrated; dynamic markup created by the script
    3. Script is necessary for adding cartographic interaction
  1. Web Mapping Workflow VI
    1. Sixth step is fine-tuning
    2. Could include debugging, but more usability evaluation and feedback
    3. Haven’t had time to teach usability evaluation; do now emphasize debugging more
  1. Web Mapping Workflow VII
    1. Final step is Deployment
    2. This has always been included in course, but relatively minor
    3. Used to be in-house, now must be off-site
  1. 2014 Curriculum Sequence
    1. Based on 2013 experience, designed a new curriculum sequence
    2. Heavy direct instruction of key concepts first couple weeks
    3. Workflow stages shown by letters
    4. Workflow stages not taught sequentially, but progressively more integrated over time
  1. How well did it work?
    1. Two reasons for wanting to assess the course:
      1. Did we meet the desired learning outcomes?
      2. Where were the sticking points? Threshold concepts (Bampton)
    2. Used four tools:
      1. Entrance survey to find out where students were at
      2. Instructor logs for qualitative observations
      3. Student extra credit assignments narrating stumbling blocks and aha moments
      4. Exit survey—gave most of quantitative data on student learning
  1. Entrance Survey
    1. Weighted toward the bottom of familiarity with open web technologies
    2. Students rated themselves even less familiar initially in exit survey
    3. Course teaches computer science concepts to cartographers and design majors—would be a very different class if we were teaching cartography to computer scientists
  1. Instructor Logs
    1. Critical practice for reflective teaching—highly recommended
    2. Had to teach to different learning speeds for different groups of students
    3. Some difficulties we didn’t think students would experience
    4. Teaching D3 the most successful outcome of course in comparison to 2013
      1. Most likely due to teaching data first and highly structured D3 lessons
  1. Student Feedback
    1. Helped highlight some misconceptions students held, such as underestimation of time, and threshold concepts, such as object-oriented programming
    2. Online examples could be both helpful and troublesome depending on how well explained and how close they fit lab scenarios
    3. Students displayed increasing understanding both of open web concepts and of what they did not yet know
    4. Filled with evidence for new computational thinking skills (quote)
  1. Exit survey
    1. Self-reported expertise with tools we taught increased from low to moderate, statistically significant
      1. No increase in open web tools we didn’t teach
    2. Asked students to rate overall emotional experience; average increased with each assignment
      1. No very negative responses to D3 lab
  1. Learning outcomes
    1. Which of the desired learning outcomes were demonstrated?
      1. Leaflet—vast majority of students got passing grade, rated their knowledge of Leaflet higher than beginning
      2. D3—same plus positive, confidence-building experience
      3. Final projects—all passing; a few students completed on their own; some professional quality
      4. Cognitive development—demonstrated in feedback, logs, exit survey
      5. Concept integration—harder to say; students rated concept transfer highly in exit survey but relied on lecture material little
        1. Room for improvement
  1. Topic sequence I
    1. In exit survey, asked students to reorder any topics they thought were out of order in course sequence
    2. Boxplots show ranges, quartiles, medians; circles show actual topic location in sequence
  1. Topic sequence II
    1. Most topics had median close to actual location
    2. Notice “using developer tools” and “GitHub” had extremely low median compared to actual location in sequence
      1. Foundational threshold concepts—should have come earlier in sequence
  1. Topic sequence comparison I
    1. We will be re-teaching 575 in-house this Spring, online in Spring 2017
    2. My summer job was to write lab modules for the online version
    3. I reworked module order based on assessment results
    4. Will also be used as basis of topic order for this spring’s residency course lab
  1. Topic sequence comparison II
    1. This shows where some of the key threshold concepts were in 2014 iteration
    2. Each topic expanded into multiple topics with more structure
    3. Some non-helpful topics eliminated
    4. Advantage of online: students work at own pace. Disadvantage: harder to give individualized assistance or review remedial concepts
      1. Entire online curriculum more structured by necessity; need not be for residency course
  1. Topic sequence comparison III
    1. This slide shows the topics students identified as coming too late in the sequence
    2. Developer Tools moved to Module 2
    3. GitHub mostly moved to Module 1, but separated between development platform aspects of Git and and deployment aspects of
      1. Will use it as main platform for project storage, versioning, and grading
    4. No final project in online course due to limitations of collaboration over vast distances
  1. Thank you
    1. This has been a bit about our experiences redesigning our G575 lab curriculum to better teach new open web platform technologies
    2. Github URL is source for published versions of the course lab assignments
    3. Student projects
    4. Happy to take questions.
  2. Bonus slides!!! Check out my students’ work!