Thursday, April 23, 2009

Attention and Multiple Intelligences Theory

I've endeavored over the past few months to exercise my "network literacy" (Will Richardson) and build a personal learning network with other educators using Twitter. Twitter is a micro-blogging tool that allows users to post 140 character messages ("tweets"). Twitter users can follow each other and grow a personally-tuned information stream. Notably, each user's stream of tweets also gets an RSS feed, and Twitter also provides RSS feeds for searches - this is very powerful! Howard Rheingold recently posted a tweet seeking feedback on his blog post entitled "Attentional Literacy", a subject important to educators, psychologists, and neuroscientists alike. A debate ensued as to whether "literacy" was the appropriate word to describe attention, which prompted me to do some thinking (which is exactly why a personal learning network is so valuable).

Recently I've been reading about Howard Gardner's theory on Multiple Intelligences (MI). Gardner defines an intelligence as "a biopsychological potential of our species to process certain kinds of information in certain kinds of ways." There are a variety of reasons I find his work captivating:
  • it fits with the modular model of the brain-mind
  • it defines intelligence as a brain-based capacity
  • it provides a model for instruction and assessment
Below are a few links to read through to get a sense of the past, present, and future of MI theory, as well as to see how it is being incorporated into education:
In what I've read so far (only a small sample of Gardner's work, let alone all the related studies and critiques), it appears that the 8 intelligences he's identified so far do not exhibit a hierarchical pattern. Although this is valuable in terms of maintaining equity among the many ways of demonstrating intelligence (vs. the more traditional assessments that focus almost completely on verbal-linguistic and logical-mathematical), a lack of such organization would be unusual if there is a strong relationship between the organization of the brain and the intelligences.

I jumped into the debate and suggested that, per Gardner's definition, that attention might be better labeled an "intelligence" than a "literacy". There is a lot of brain research happening to improve our understanding of the nature of attention, but there is no question that it is a biopsychological potential, and that it is related to information processing. Here are but a few examples of current literature on the neuroscience of attention:
As I thought more, I wonder if, in fact, attention may not be just an intelligence, but an example of a heirarchichal intelligence. I don't know if "meta" is the appropriate word to use here - is attention "above" and providing top-down influence on other intelligences, or is attention more of a foundation, lower-order intelligence which other intelligences must gain in order to activate? Perhaps as neuroscience research improves our understanding of the brain-based nature of intelligence, a clear picture of the intelligence heirarchy will emerge based on the brain structures involved and their relationship to information input / output patterns. However, whether the location is "above" or "below", it seems clear that attentional intelligence is on a different level than the others already identified.

I'm also now realizing that Will Richardson's "network literacy" could also be thought of as an example of Gardner's "interpersonal intelligence", though with the context shifted to the digital realm. What's also been on my mind as I learn more about MI theory is how it might be similar or different to other cognitive theories I've learned about in the past, particularly p-prims, facets, and cognitive resources. It's clear that educational systems have room for improvement with regard to instruction and assessment of all 8 intelligences. Thinking of attention as an intelligence within the MI theory also helped me to realize that we educators - except Howard! - tend not to provide direct instruction on how to develop and use intelligence. Regardless of whether it is accurate that attention is an intelligence or a literacy, Howard's point is well made that our increasingly multi-tasking and digital students will benefit greatly from direct instruction on how to pay attention.

Monday, April 13, 2009

Rubrics: The Keystones of Standards-Based Academics

I'm cross-posting here a response I wrote to the post "Settling the Score" over at What It's Like on the Inside, which discusses standards-based grading and references a recent article in the NY Times called "Report Cards Give Up A's and B's for 4s and 3s".

Great post, SG ... I've been chomping at the bit to find some time to leave a comment, and I hope it's not the case that the discussion here has already come & gone.

From my perspective, I think the 1 - 4 grading system for standards is a bad idea, for a few reasons. First, the 1 - 4 system is, I would suggest, so similar to the GPA system that they are easily confused. If using the 1 - 4 system is a matter of using shorthand symbols for communication efficiency, we could just as easily use shapes ... a circle for not meeting expectations, a triangle for partially, a square for meeting, and a star for exceeding. We could also just use the narrative definition of what each number is intended to represent, as others have mentioned.

Of course, those narratives and/or shapes don't necessarily have the simple ordinal inference that the number symbols do, which help the consumers of the information to determine their position in the continuum of learning. This leads to my second concern: the 1 - 4 system breaks the learning continuum into too few categories, one of which (the highest) is not available to all students at all times. The example used about simple addition skills is perhaps better critiqued from a curricular organization standpoint, but it'll serve as a good example about the problem with the grading system, too: if the concept / skill ("standard") is defined with such specificity that it can't be measured other than in a binary nature, it should not be considered a standard (it also opens the question about "scripted" teaching when standards are excessively narrow and numerous). I think teachers deserve multiple tools for assessing student learning, and trying to squeeze everything into the four categories represented by the 1 - 4 system makes an inherently difficult task even more difficult; furthermore, I think it gets us away from the most important goals of standards-based reform.

Nothing, to me, reveals the complexity and difficulty of the standards-based reform hypothesis more than the commonly-used but poorly-formed phrase "meeting the standard". The word standard is being used in two different ways: first, as an articulated concept or skill that should be learned as a result of the class; second, as an articulated expectation of a learning outcome. Therein lies my third concern with the 1 - 4 system: it's a "solution" to the wrong problem. The problem is that it is difficult to describe what we want students to learn and how we expect them to demonstrate to us that they've learned.

The 1 - 4 grading system doesn't solve that problem, it makes it worse because it confuses (and potentially alienates) parents and other consumers of our outcome reports like college & university admission officers. It's important that we acknowledge that our students are moving through a continuous educational system; it is our folly and detrimental to students to ignore the importance of good communication among all constituents in the educational community. While it's possible that parents and colleges might become savvy over time, it's just as likely that parents might become disengaged from their child's educational progress earlier or that colleges might mis-translate information in the process of trying to compare student learning outcomes. The 1 - 4 system doesn't inherently improve student-teacher communication, either - the critical component here isn't a grading system, it's rubrics.

Rubrics act as the map for students to navigate what they need to learn and how they need to demonstrate that learning. Rubrics can be aligned with the 1 - 4 system, to be sure, but they can as easily be aligned with grading systems such as A - F or percentages. The point, I think, of standards-based reform is to improve our communication of the concepts and skills that we want students to learn, and to provide students with clear guidelines on how they are to demonstrate their abilities. Rubrics, not grading systems, should be the focus, since they are the key communication tools bring together learning goals, demonstrations of learning, and achievement results.

Wednesday, April 8, 2009

PKMzeta and Long-Term Memory

Just a few days ago the New York Times featured an article in their Brain Power series that has quickly risen to the top of their "Most E-Mailed" list: "Brain Researchers Open Door to Editing Memory". The article discusses some recent breakthroughs by Shema, Hazvi, Sacktor, and Dudai in their research on the molecular basis for long-term memory, recently published in the journal Learning & Memory under the title "Boundary conditions for the maintenance of memory by PKMzeta in neocortex". The subject of their work is the molecule PKMzeta, a protein found in regions of the brain that are involved with long-term memory. Their work also centers around the molecule ZIP and its ability to inhibit PKMzeta, and thereby "erase" long term memories.

As the authors note in their introduction, protein kinase molecules such as PKMzeta have been under investigation for some time in terms of their role in brain function, specifically with regard to learning and memory. Only recently has PKMzeta been isolated in certain studies for its role in sustaining long-term memory. The study by Shema et al used a variety of training scenarios (with rats as their model organism) designed to form long-term aversions to certain tastes (sugar and salt mixed with water). The researchers then dosed the rats with ZIP, varying the timing of the ZIP dose relative to the training process, as well as the strength of the ZIP dose, in order to determine the limitations on PKMzeta's role in the long-term memory.

The findings indicate that PKMzeta's role in long-term memory formation begins at somewhere around 72 hours after the training event, and that the molecule's role in sustaining the long-term memory sustains for at least three months after the training. The researchers demonstrated that PKMzeta is not involved in short-term memory processing, as doses of ZIP given just prior to training, during training, and shortly after training did not affect the formation of long-term memory. Furthermore, the researchers found that similar long-term memories (taste aversion) could be re-formed after rats were dosed with ZIP. The researchers also found that ZIP dosing has a critical concentration below which it is ineffective, and that its "erasure" of long-term memory is not taste-specific.

Although these findings provide significant insight into the limitations of PKMzeta's role in long-term memory formation and sustainment in the rat neocortex, much more needs to be discovered regarding the specific mechanism of action for the molecule. The researchers note that PKMzeta is involved with pathways involving neurotransmitter receptor expression as well as with cytoskeletal modulation in order to enhance synaptic connection strength, but much more detail is needed in order to understand how these types of changes relate specifically to long-term memory. This also sets aside the ever-present question of how these cellular and molecular changes are experienced consciously.

While it is exciting to see that headway is being made in the molecular basis for learning and memory, and to extrapolate that someday we might be able to help students learn and retain information better through medications and therapies that target these specific pathways, it is also important to keep in mind that there are significant differences between the rat brain and the human brain, and that the limitations of communicating with rats means that only certain types of memory-based behaviors, such as taste aversion, have been studied. It is certainly true that the basis of learning and memory in the human brain will be molecular and cellular in nature, but these specific findings regarding PKMzeta and ZIP may be localized to regions of the brain that are of little interest to human educators. However, these findings clearly indicate that we are increasing our ability to understand the inner workings of the brain, and that it is reasonable to place hope in neuroscience's ability to disentangle the complex mechanisms of learning and memory.

Wednesday, April 1, 2009

Introducing "CADIE" - Google's Artificial Intelligence project

As many are aware, Google employees are encouraged to use 20% of their work time on projects of interest. Google announced today the fruits of yet another "20%" project, a groundbreaking foray into "Strong Artificial Intelligence" named CADIE (Cognitive Autoheuristic Distributed-Intelligence Entity). It appears that after months of initial programming work and access to Google's huge computational resources, CADIE has, as of very early this morning, manifested consicousness and is communicating, unsurprisingly, via a web page. Because of her advanced neural-network based programming, input from the real world via sensors and access to all of Google's information, and the number of CPU processors available, it appears that CADIE is "growing up" at an incredible pace - keep watching her web page for developments as the day continues - who know's where this will go!