Notes from Sessions

Thursday morning table talk report outs

Major points from each table:

  • Introducing the idea of Live Scripts and different ways they can and have been used in courses. Silvio has several examples of things he does including using an LeTeX editor. More notes in his workspace page.
  • First and last lightening talk were interesting as a pair. We were thinking about Live Scripts and using them in scaffolding different types of learning. How much information to give to students is an open question. We want people to decompose questions as they approach answering them – how to start them with easier questions. Can MATLAB grader be inside Live scripts? Can it check for aspects of a graph for us – like looking for correct axes?

    For a discipline specific course. Can students have appropriate usage of pseudo code – how to check this ourselves or using Grader within Live Scripts. Want students to have an intuitive understanding of specific code and know the limits of that method (Euler’s method). (MathWorkers have ideas on this – maybe talk with Jon and Evan).
  • We talked about using data focused from first year to graduate level. We want students to deal with real data. How to get that into MATLAB and how to visualize it and interpret it.  What does that whole journey look like for students and how do we think about it across classes. 
  • This group discussed how important final exams are and the value of a stressful, intense learning period for getting students to really work with the new information. This functions different for students than mid-term exams. They want more from Benjamin on the topic of his short talk. The lockdown of AI at Vanderbilt was interesting and they want to know more about how and why that is happening and how students and faculty are responding to that.  A question they have is: If you put information into an AI – where does it go? Want to expose students to all aspects of modeling including how to best use documentation and help functions. 
  • We discussed the overlap between programming course content. There is an opportunity to scaffold learning of programming across courses where they are being taught by the same person or by faculty who are able to work together on course design. A learning objective could be : take a real world problem and translate it to a logical sequence of algorithms. This would allow you to teach you both naming and new things about algorithms. 
  • The main topic here was learning goals for a programming course. Focus on two: 1)skews in programming – student can code these skills, functions graphs logic etc. 2)discipline focus for any specialty. Example: linear algebra this afternoon talk. Write own functions – writing matrices or inverse of matrices, finding areas under a curve etc. This will be discipline specific.
  • Focus on what resonated from the short talks – programming courses using GPT. How to learn to debug code, compare and contrast code. How to provide good GPT prompts. How to get code from English to other languages. How to use Live Script to its fullest function. How to use Grader to its best capacity. Outcomes: writing effective algorithms and fix and debug code. Discipline specific: how to replicate and modify code from simplified to more involved code. 
  • Using MATLAB as Tools in different disciplines. Engineering and Science. Most of us don’t want students to use AI, but this is the modern version of what a calculator used to be. It’s coming and we need to learn to use it and those who can’t are going to be left behind. Teach with it as a tool. Enhance our way forward incorporating AI to keep up with the rest of our peers. Difficult topics we are teaching could be made easier by AI (visualizing complicated equations for example). 

Learning Outcome Affinity group report out 

Individual group notes pages here.

  1. Using computation to learn about the discipline. Interpret data and draw conclusions. Compare outcomes of experimental results. explain why simulations are not always accurate. Recognize data patterns. 
    1. Many specific course outcomes such as: collect and analyze data, predict, design, interpret and compare, identify deviations and sources of error. 
  2. Good general outcomes :
    1. Be comfortable solving problems
    2. Cleaning data (manipulating data)
    3. Communicate understanding - technical report, pseudo code
    4. MATLAB is something not scary that they might use later in life or school. 
    5. Accreditation driven outcomes due to discipline requirements 
    6. Tech courses can include social or interpersonal goals - how to put these into learning objectives for a technical focused courses. 
  3.  Bloom's is a good framework 
    1. Progression of Blooms can work across course levels
    2. Emphasize approximation across Blooms
    3. Decompose and distill problems - implementing appropriate analysis
    4. Students will demonstrate .... (several examples- see group notes) 
  4. Motivating students to learn is an outcome
    1. Interpret data and graphs
    2. Evaluate, assess, interpret
    3. How to bring data into the classroom is a challenge for faculty
    4. Systems approach and understanding
    5. Decide if results they get are reasonable
    6. Bringing in understanding from other courses (stats, CS)
    7. Just being willing to engage with MATLAB to do things like data visualization 
  5. Compute accurately
    1. use real data, accurate, appropriate methods 
    2. efficiency in tool use
    3. ability to communicate with non-experts, also do this with visualization
    4. Some specific examples of these in a math context - see their notes page 
    5. aFor introductory courses
      1. Rewrite equations, setups, implement algorithms, use functions, demonstrate understanding of tools, numeric accuracy, 
      2. creating visualization for analysis and phenomenon identification
      3. be able to restate a problem
      4. pattern recognition, classical function recognition 
      5. computation tool to gain ....
    6. Majors - 
      1. no hard coding, all flexible variables
      2. be knowledgeable about error
      3. use programming to better understand what is going on in class
  6. Thinking about majors 
    1. Develop scripts for data visualization
    2. Describe real world with equations
    3. Model real world systems
    4. Break up systems into simpler steps
    5. Mirror Bloom's taxonomy - 
      1. Start with memorization then move on 
      2. work the same topic up through all the steps of Blooms
      3. end with a creative step
    6. Specific examples in the group notes 
  7. Split lower and upper levels
    1. Graphs tell a story - plot, fit, extract parameters, understand relationship between parameters
    2. Tools to build models. Word problem to code that can run. Synthesize to code
    3. Executive function tasks, professional tasks. How to grade this. Using understandable names, making code readable to humans, communicating with other people and faculty
    4. Try to focus on one thing at a time in assignments - not doing everything all at once.