We held the kick-off meeting for Round 9 of the instructor training course this Wednesday (April 23). More than 40 people attended, and we got through quite a lot of introductory material on teaching as a performance art, the importance of feedback and reflection, the differences between novices, competent practitioners, and experts, and a handful of other topics.
The notes from the Etherpad are included below. For our next meeting on Wednesday, May 7, please do the following:
- If you haven’t already written a short biography for yourself, please log in to this blog and do so.
- Read the first two chapters of How Learning Works.
- Take a look at the concept map in the back of How Learning Works, and at some of these examples from previous cohorts.
- Choose a topic that you think you could teach to novices in five minutes or less. (Please choose something related to Software Carpentry in some way: it doesn’t have to be one of our core topics, but it should be something you’d plausibly teach in a bootcamp.)
- Draw a concept map for it and post that to this blog in the categories “Concept Map” and “Round 99/1″ by Wednesday, April 30.
- Comment on at least two of the concept maps posted by other people. Try to give substantive feedback: not, “This is great!”, but, “I wouldn’t have thought of the connection you’ve shown between A and B,” or, “I think P and Q ought to be part of this map.”
We’ll meet at the same times (10:00, 14:00, and 19:00 Eastern) on May 7 — I look forward to chatting with you all then.
Round 9.0: Introduction
Wednesday, April 23
Attendees (19:00 Eastern)
- John Mosher (Tulsa, OK Tulsa Community College)
- Greg Wilson (Mozilla, Toronto)
- Tim McNamara (New Zealand eScience Infrastructure, New Zealand)
- Matthew Dimmock (Monash University, Australia)
- Devasena Inupakutika (University of Southampton, SSI, United Kingdom)
- Wesley GOI (National University of Singapore, Singapore)
- Ahmadou Dicko (University of Dakar, Dakar, Senegal)
- Dana Bauer (Rackspace, Philadelphia, US)
- Simon Michnowicz (Australia)
- Dan Warren (Macquarie University, Sydney Australia)
- Jeremiah Lant (Louisville, Kentucky, US)
- Jacob Levernier (University of Oregon, Eugene, Oregon, US)
- Yu-Ching Shih (Taipei, Taiwan)
Attendees (14:00 Eastern)
- Greg Wilson (Mozilla, Toronto)
- Michael Schliephake (Stockholm)
- Bror Jonsson (Princeton University, Currently Cape Town)
- Victoria Offord (Royal Veterinary College, London)
- Dav Clark (UC Berkeley D-Lab, Berkeley, CA)
- Genevieve Smith (UT Austin, TX)
- Jon Duncan (UNC Chapel Hill, NC)
- Chandler Wilkerson (Rice U, TX)
- Russell Alleen-Willems (Diachronic Design, Seattle, WA)
Attendees (10:00 Eastern)
- Greg Wilson (Mozilla, Toronto)
- Padraic Stack (Dublin, Ireland)
- Hsingtzu Wu (JAEA, Japan)
- Scott Burns (Vanderbilt University, TN)
- Jonathan Frederic (Calpoly SLO, CA)
- Aur Saraf (Tel Aviv, Israel)
- Graham Etherington (The Sainsbury Lab, Norwich, United Kingdom)
- Dan MacLean (The Sainsbury Lab, Norwich, UK)
- Mark Stillwell (Cranfield University, UK)
- Alex Simperler (NSCCS/Imperial College London,UK)
- Stefan Pfenninger (Imperial College London, UK)
- Christian Jacobs (Imperial College London, UK)
- Timothy Warren (University of Washington)
- Mark Wilber (UC Santa Barbara, USA)
- Jeff Hollister (US EPA, Rhode Island, USA)
- Huayan Gao (CUHK, Hong Kong)
- Shyam Rallapalli (The Sainsbury Lab, Norwich, UK)
- Sandra Garza (UANL, Monterrey, México)
- Dureid El-Moghraby (University of Leeds, UK)
- Catalina Anghel (OICR, Toronto, ON Canada)
Please take notes here
- We are teaching “free-range adult learners” as opposed to “battery farms” ie. not in a classroom (farm), not a corporate training and not in a seat [ref pyladies]
- If you haven’t watched it, Gregs talk from PyCon is great http://pyvideo.org/video/2649/software-carpentry-lessons-learned +2
- Ed Psych is the study of how people learn (since 1950ies). We know a lot about it but it’s not reflected in public knowledge. Most teachers are uninformed
- Educational Psych.’s findings haven’t made it into lay language as much as medial findings. But there’s a large literature to draw from.
- Instructional Design is our second subject
- Bottom-up vs. Top-down approaches to teaching & learning re: instructional design. Must experiment with both approaches to optimize next lesson.
- “How Learning Works” (2010) is a really good summary of what research has found to date. http://www.wiley.com/WileyCDA/WileyTitle/productCd-0470484101.html
- Instructional design (2nd stage, ie. how to teach and design effective materials based on psychology) plausible methods, varies across age eg. competing sch of thoughts (applies for the english lang systems) 1. PHONICS [dominant sch of thought; better for kids] VS TOPDOWN (incentivised to go on to read eg. Dr. Seuss) which is feedback driven.
- Pedagogical Content Knowledge is our third subject. It means how to teach our specific topic. Computer education lags here 2-3 decades behind physics and math. (i.e. how to teach `for` loops to people who have never programmed before.)
- One of the things SC does is accumulate programming PCK (Pedagogical Content Knowledge — the “bag of tricks” used for teaching).
- For example: Use etherpad as teaching tool (have one instructor teaching, and one watching the etherpad chat. The chat and collaborative notes can both help to indicate misunderstandings among students quickly). Must be fabulous for webinars!
- Computing education is lagging here, compared to, e.g., teaching physics.
- Single most important thing for us to learn- Teaching is a *Performance art*. Are they intrinsically motivated to learn it. Primary JOB as a teacher is make them excited, hold their attention , make them interested in any topic that you’re teaching and WANT to learn. Can be done in different media (borrowed from performing arts domain), best done live (which is why SC runs 2d workshops).
- “You are not your learners.” Just like developers/programmers are not their own users.
- Improvement in performance requires feedback
- Example of US Teachers getting, on average, 4 hours a year, Finnish teachers get 3 per *week* (Scary!).
- Intrinsic motivation = single best predictor of success
- teacher: make students want to learn, by making them believe they can
- for example problem based learning to get them engaged
- Three stage model (from Patricia’s five stage model) with respect to learning any skill (started with Patricia Brenner on Nurses, 1980s):
- Novices: ask questions that don’t make sense. They don’t know what they don’t know. They need the correct foundation before proceeding, otherwise the imparted knowledge falls into an unknown and potentially incorrect mental model.They haven’t a working model. In our shell tutorial, we teach 15 unix command in 3 hrs — and teach key concepts, e.g. what is a file, what is a folder, what is a command … Likely to reinforce wrong ideas if you start without the right map
- The idea here is not to teach the commands themselves (e.g., `whoami`) as the final goal — it’s to use the commands to show larger principles.
- Too much information too quickly can actually end up reinforcing incorrect understandings.
- RECOMMENDATION: Give novices a good conceptual model.
- Competent: can accomplish routine tasks most of the time. Workable model, can identify which rules apply to a particular situation. Let’s get scientists to be competent practitioners. Debugging is generally more difficult for competent practitioners.
- Sparsely connected mental map
- SWC’s goal is to move people from Novice to Competent
- Three stages of feedback:
- first: we get feedback (e.g. fifth grade teacher’s comments on an essay)
- second: we give feedback (e.g. analyze a theme in a novel)
- third: turn the critical eye on yourself (distinguishing expert)
- Expert: someone who can look at themselves like others do (context: performance). Critique is a practice of this (code review, book reviews). At some point a switch flips & then critiques turn inward, unconsciously. This feedback loop* accelerates improvement. Experts tend to be much better at debugging than competent practitioners. They have more (correct) connections between knowledge structures. Experience alone (as well as number of facts memorized) does not make an expert, they need to routinely reflect on themselves (i.e., using their ability to critique and give feedback to others, and applying that to themselves). The expert is willing to take feedback on board.
- “There aren’t experts who don’t reflect.”
- Teaching will improve with feedback, consciously or unconsciously
- Average US teacher has 4 hours of another teacher in classroom.
- Finland — 3h/week
- Never hesitate to sacrifice truth for clarity
- Prompt for feedback in your classes. Applies to any teaching context.
- Expert’s mental graph consists of more densely connected facts (higher degree per node) VS sparsely connected ones (novice, competent)
- Novices: ask questions that don’t make sense. They don’t know what they don’t know. They need the correct foundation before proceeding, otherwise the imparted knowledge falls into an unknown and potentially incorrect mental model.They haven’t a working model. In our shell tutorial, we teach 15 unix command in 3 hrs — and teach key concepts, e.g. what is a file, what is a folder, what is a command … Likely to reinforce wrong ideas if you start without the right map
- http://teaching.software-carpentry.org/category/concept-map/
- http://teaching.software-carpentry.org/2014/01/22/concept-map-unix-find-command/
- work with another instructor and solicit feedback from that person — Software Carpentry incorporates this, and it’s a good idea for all teaching.
- Difference between top and bottom map illustration at the link: the bottom one includes descriptions for each link between nodes.
- Concept map: picture of your mental map of a topic including blobs (simple ideas) connected to others through a labeled arrow (defining the relationship).
- potentially advantageous when:
- you’ve not idea of the order of points when jotting down notes
- co-teaching (ie. to check the co-teacher’s concept map with yours)
- With 20+ people in a room, all familiar with an idea, each may have a completely different concept map of that idea if he or she draws it out.
- WHY? to give them a reference to check their mental maps against.
- Or to check learners’ understanding of a concept you’ve just explained, as feedback on your teaching. (This does require that students’ believe that this is feedback for teaching, rather than some sort of exam).
- Give students a concept map before class and they will doodle on it during class.
- The more polished something is, the less good feedback you’ll get on it (if you sketch something out in front of someone — even if you’ve actually practiced the sketch beforehand — people will be more willing to comment honestly on it, thinking, “Well, he/she can just make another one — that sketch didn’t take any effort.”). Thus, it’s recommended that we not use Concept Mapping tools, vs. just hand-drawing and snapping a picture/scan.
- Human Brain Working memory [Memory] vs Long term memory [HDD] in terms of the the recall speed
- 7+2: eg. recalling telephone example 7digits (everyone’s fine) ~12 (most failing)
- Certain numbers can be “chunked” down and though of as one item/number, though (like one’s local area code). Later, we’ll talk about how this is directly relevant to programming.
- potentially advantageous when:
- We typically are good at giving out facts, what we need to improve is our ability to explain how facts are linked to one another
- Three good reasons to use them:
- Allows you to avoid getting locked into a single lesson plan
- The more links there are in your head to a fact, the more likely you are to remember the fact. An isolated fact gets forgotten.
- The fewer hops you have to traverse in your brain-graph, the faster you’ll solve the problem. The more hops, the higher chance you’ll forget in the middle or make a mistake or…
- These check understanding not only between teachers & learners but also teachers & teachers. Verify the teachers are speaking the same language.
- Some more reasons:
- Plan lessons with them.
- Communicate with co-trainers.
- grow a mental model for people on a flip chart/white board —> accelerate learning
- let participants draw them as they learn (uncommon in the english education system)
- checksum: have the students do them and give them to you to check they’re on the same page with you. More useful feedback than asking what they didn’t understand.
- Easy to share with neighboring students in a class
- Ask students for their concept map during a break, you can see the holes (/wholes?)
- [Students must not think this is a test]
- How to use concept maps:
- “Is it simple enough that you can teach it in five minutes or less?”
- If you have ~6 bubbles, you’re doing fine. 20 is way off — they won’t fit in your working memory. Concept of seven plus or minus two as limit on e.g. length of phone numbers (note, Vogel claims capacity is 3-4 “chunks” — Greg gives example of an area code as one memory slot / item / chunk).
- The more polished your work, the less honest the feedback you’ll receive. <— YES
Exercise
- For feedback: what makes sense? what is unclear? what would you do differently?
- Prefer a topic in the core SC material we all share so you’ll get more feedback.
- Read chapters 1 and 2 of “How Learning Works” by Ambrose et al.
- Go low tech — just a whiteboard/paper and pen will do. (The more polished something is, the less honest the feedback will be)
- Place the concept map blog posts under the categories ‘Concept Map’ and ‘Round 09/1′.