Monday, April 25, 2016

A Practical Solution to "What We Say/What They Hear"

“In art intentions are not sufficient and, as we say in Spanish, love must be proved by deeds and not by reasons. What one does is what counts and not what one had the intention of doing.”  -- Pablo Picasso

It's my belief that teachers at all levels have good intentions and genuinely want their students to learn.  What this post is about is the notion that instructor intentions are not sufficient, and getting students to do what is needed for authentic learning is what counts.

Background
The starting point is for the ideas I want to get across from a the Launchings Blog of the MAA.  David Bressoud does a wonderful job of describing a recent paper by Kristen Lew, Tim Fukawa-Connelly, Juan Pablo Mejia-Ramos, and Keith Weber.

Below are links to Bressoud's twin posts.  They are worth the time!
"What We Say/What They Hear I"
"What We Say/What They Hear II"

The short version is presented in this diagram

















Despite multiple passes through the proof and explanations, students in the study have a difficult time pulling out the instructor's intended messages in the proof and the instructor's spoken comments.  At first glance this makes sense, as the "information transfer" model doesn't work so well when the goal is "developing critical thinking."

As mentioned by Bressoud, Annie and John Selden and others have documented the difficulties students have with analyzing, proving, generalizing, packing/unpacking statements, etc. Learning higher mathematics is challenging, and most math majors struggle with learning proof.

Bressoud suggests options like flipped classrooms and clickers can work, but he notes that such methods have high initial investment in time and requisite knowledge and skill.  While these methods are learnable, many instructors may be in situations, where implementing them is not practicably feasible.  Other options are needed.

Another Option Exists!
Now to the main point of this post. There exists an "easy entry, high upside" method to help students come away with the intended messages.  Put simply, instructors can take their list of intended messages and turn them into math tasks. These tasks can be deployed via small group work, homework, etc.

Let's take a closer look. The intended messages of the instructor in the study are
  1. Cauchy sequences can be thought of as sequences that “bunch up”
  2. One can prove a sequence with an unknown limit converges by showing it is Cauchy
  3. This proof shows how one sets up a proof that a sequence is Cauchy
  4. The triangle inequality is useful in proving series in absolute value formulae are small
  5. The geometric series formula is part of the mathematical toolbox that can be used to keep some desired quantities small
      Each of these five points can be turned into "after the proof" problems.  They can be reformulated as
      1. Explain using sentences and diagrams why Cauchy sequences "bunch up."
      2. True or False and Explain:  One can prove a sequence with an unknown limit converges by showing it is Cauchy.
      3. In the proof, find the part that proves the sequence is Cauchy.
      4. In the proof, find the part where the triangle inequality is used, and then identify a general strategy based on this specific instance that you can put in your mathematical toolbox.
      5. Explain how the geometric series is part of a mathematical toolbox to keep some desired quantities small.
      One way to deploy these in the classroom is to present the proof on the board or pass out a handout with the proof, and then have students work through some of the tasks in pairs and then share (i.e. Think-Pair-Share).  If time is an issue, doing one or two and assign the rest for homework.  Alternatively instructors could ask students to read the proof before class, and the instructor could highlight the proof and spend more time on student-centered tasks in class.

      Several advantages exist with this option compared to flipped classes or using clickers.  The first advantage is that it requires the least experience and least amount of pre-class planning or classroom equipment.  One could be in classrooms like I sometimes teach in with no technology, old boards, and desks built 50 years ago.

      Another advantage is that it does not require deep knowledge of common misconceptions.  Instead, the instructor asks students to explain their thinking (in one way or another) and that's how the instructor gets insights into student thinking.  That is, use an activity and gather formative assessment.

      A third advantage of this method is that it does not deviate from the conventional class prep process used by most instructors. Preparing presentations of a proof with explanations is something that instructors have done many times. The method presented here tweaks the process by transforming the intended messages that instructors would normally say to students into concrete mathematical tasks for students to work on. This change is practically feasible and doesn't require a significant alteration of an instructor's workflow.  It's also a step towards active, student-centered teaching and can be built upon over time into forms of teaching that more deeply engages students, such as IBL.

      Instructors have good intentions and intended messages. I claim that it is how the intended messages are deployed in class that can be addressed in practical and effective ways. We can turn those amazing insights into amazing learning experiences, and "let problems do the talking!"

      Friday, April 15, 2016

      IBL Workshop Model and Real-World Results (Wonkish)

      This post is about professional development, and is intended for those interested in faculty professional development.  There may be some applicability to K-12 PD, but I'm not promising that. Presented here is a model that I argue can be used outside of Mathematics in higher ed. My experience in K-12 PD tells me that changes can be made for K-12, but that there are other aspects (such as working with school districts, parents, testing, etc.) that add more layers that need to be carefully and thoughtfully handled (that may require much more resources).

      This post is also from the perspective of people on the front lines. Our team works with teachers, and we are teachers.  Our work is grounded in the hard efforts of teachers who have thought carefully about the issues.

      The evaluators for the project I've been working on are Dr. Sandra Laursen and Chuck Hayward, at Ethnography and Evaluation Research, CU Boulder.   We have run a set of IBL workshops (funded by NSF SPIGOT) in 2013-2015, and have some results to share.

      The main results:
      • Total number of participants is 138
      • ~ 75% implementation rate
      • 61% of the participants are in the first 5 years of their teaching careers.
      • In just the first year after the workshop, participants taught 180+ classes to 4600+ students (in the real world)!
      • Ongoing support via Email mentoring is a critical feature
      • An inclusive or "Big Tent" definition of IBL is critical for allowing participants to find their own way, suitable for their situation.
      The short version is that uptake is happening at a high rate!  This is an encouraging sign, as it provides evidence that effective PD can make a difference.

      How we got to these results is a very, very large undertaking spread over about a decade of work by a team of people.  I'll briefly highlight the main facets in the rest of this post.  An embedded slideshow appears farther down this post with more details and diagrams.

      First, we identified major obstacles that faculty face when learning to implement IBL methods.   If an instructor has not had much or any firsthand experience with IBL, it is hard to convey through discussions what an IBL class is like and how to pull it off. Additionally, there are skills, practices, curricula, and assessment to consider, all of which are different or changes.   Hence the need for a highly specialized professional development experience that directly addresses the challenges and needs new IBL instructors deal with.

      Obstacles or challenges include lack of experience with IBL,  the need for instructors to learn IBL teaching skills, the Math Ed Knowledge Gap, IBL course materials, assessment in IBL courses, and organizing the structure (i.e. syllabus) of an IBL course. With a list of obstacles in hand, much effort went into designing workshop components to direct address and reduce the identified barriers.

      It it emphasized that the workshop is designed as a single composition, where all the parts of the workshop are interconnected and the main goal is getting participants to a point where they can teach an IBL class successfully.  It is definitely NOT about trotting out our favorite activities.  Indeed, teaching is a system, and teaching is a cultural activity. Hence, the workshop is about providing a broad IBL framework that participants can adapt to their situation. The main focal points are (a) IBL teaching skills and practices, (b) understanding the evidence for IBL and how students learn, (c) linked teaching choices (assessment, problem posing, and content), and (d) building a practically feasible (to the participant) course.

      Another important feature of our work is data driven decision making. The staff used evaluation data and research to inform decisions about all aspects of running and designing workshops, from recruitment, building workshops materials, adapting sessions to better meet participant needs, and also to main components that are successful.  It's through this iterative, self-assessment process that the workshop model has improved over time.  Sandra Laursen and Chuck Hayward head up the evaluation and research efforts for our projects, and their regular insights helps use make small and longer-term positive changes.

      Our current NSF funded project, PRODUCT, has the main goal of expanding the profession's capacity to offer a version of the 4-Day IBL Workshop and developing short workshops that can be "sent" around the country and increase awareness of IBL.  The goal is to build the PD network up to a point where a much larger capacity PD exists and a larger variety of workshops (weeklong and short workshops).

      More details are in the slides below.


      OR click this link to Slides About the IBL Workshop Model

      Edit: Click here to link to the evaluation report.