The following is a reading list with some comments.
My early training was in the area of physical organic chemistry. I went on to study enzyme mechanisms. Mechanisms people, the products of such training, try to give a molecular level accounting of chemical reactions. That is, they give detailed accounts of how molecules are transformed during chemical reactions. Knowledge of mechanisms is used widely in science and technology, drug design being among the most recent and exciting areas. A person trained in mechanisms, when confronted with a learning problem, would ask questions about how molecules in organisms store and retrieve information. When you understand this context, you understand much of this reading list. Example discussion of chemical reaction mechanism.
The bottom line of this reading is that:
Neurological Basis for Learning. It is widely believed that learning results from neurological change.
In a remarkable experiment by Schab, a large group of students was regrouped into a 'chocolate' and a 'camphor' group. The first group studied in a room pervaded by chocolate odor, and the other group in a camphor room. The groups were tested, half chocolates in a camphor room, and vice versa. Outcome: chocolates test better in chocolate; camphors test better in camphor.
If you look for some rationale for this in terms of content cues, good luck. In a neurological model, however, the outcome is predicted. Test success depends upon what ends up being transmitted to the four million (or so) neurons connected to the learner's muscles. Having as many sensory inputs lined up the same way they were during learning as during testing increases the likelihood that a question (with its features that are likely to lead to an appropriate output) will make it through the neurological processing system to an accepted response.
While there are many books and papers I can recommend in this area, the one that that puts it together best for a teacher is Rethinking Innateness (Elman et al). Learning results in neurological change; without neurological change, there is no learning. Lest you think that no issues remain, read Marcus.
Teaching is neuron modification technology; a teacher is a neuron modification technologist. In spite of that, my sense is that there currently is no way to make progress in teaching at this time by focusing narrowly on neurons. While learning may result from synaptic change, there's no rational way to study that change directly in a fashion that helps teaching.
What adopting a neurological model does direct us to do is to:
Feedback
Performance-based feedback is the most powerful tool available for teaching.
As a result of my collaborations with Schraw, I've focused on performance-based feedback and modeling as vehicles for analyzing most successful instruction. My arguments about feedback are these. If you believe in a neural learning scheme, then you must provide feedback. We train neural nets in computers and in people the same way -- we give them feedback. This is consistent with first language learning, and many, many other observations.
How do humans learn? A really interesting book that all parents of young children and newborns should read is the Scientist in the Crib. There is substantial controversy about how language is learned. A view supported by data and measurement is set forth by Moerk in The Guided Acquisition of First Language Skills. Earlier books by Moerk set forth the data leading to this book.
It's not just a matter of providing the learner with sensory inputs. It's a matter of eliciting some output from the learner, and then giving sensory inputs to him or her based upon that output.
Feedback to learners can come from tests, participation in small groups, working in two's or three's on a lecture question, etc.
The current literature advocates engaging students in active learning. I don't know -- explicitly -- what active learning is. Like others, I do think I know it when I see it.
Performance-related feedback is a lot simpler to detect; you really do know it when you see it! Most education research fails to focus on feedback. Even when people talk about feedback, they focus on the trees rather than the forest. In most experimental research in education, there is a difference in the amount of feedback received by one group relative to another. Learning nearly always is most efficient for that group receiving the most feedback.
Many things are explained by this model. Obviously we can explain the stages that were described by Piaget. Nearly all of Vygotsky, for example, is explainable in terms of feedback.
Almost all inquiry strategies engage the learners in receiving more feedback than controls. (Overall, inquiry has a lousy record when you look for impact several semesters out. When you think you've found it, you most often have found evidence for the power of teaching some self-regulation strategy.) The literature in guided inquiry essentialy suggests places in which feedback to,learners would be most useful, and ways to provide such feedback.
Feedback often looks a lot like what a behaviorist would call reinforcement or punishment. During training, a computer 'neural net' is told yes (1) or no (0). One can associate value with feedback -- yes (good) and no (bad), yes (correct) and no (incorrect), or yes ( reward) and no (punishment).
A pidgin is an elementary language the emerges when two folks with different languages try to have a conversation. Feedback often imples language. Suppose your wanted to educate some animal, say, a porpoise. Just how would you talk to a porpoise? My students use English; porpoises make squeaks, some of which our ears cannot detect -- and maybe those squeaks are porpoise-ese. A remarkable paper on training porpoises (Pryor et al.) has provoked me to ask that question. One day, some porpoise trainers decided to reward their porpoises for behaviors seen just once that day but not been seen before during that same day. In other words, the game was 'show me something today that you haven't already shown me today.' So, the first time the porpoise jumped through a hoop, it got a 'reward' (click sound, followed by fish). The second time, no 'reward.' The porpoises caught on quickly, and, by the end of the day, the trainers had seen the porpoises do things they had no idea porpoises could do! How would you talk to a porpoise? I suspect that the porpoise activity described in this paper is much more a matter of feedback and much less a matter of reward than one might be led to believe by a classical behaviorist. Maybe the fish really isn't a reward; maybe it's pidgin porpoise, just a "1" or a "0."
I hold the view that you must be able to separate feedback from reinforcement in your thinking. We neither reward nor punish neural net computers; we just give them feedback. If you are going to be an effective teacher, you need to be able to detect instances of feedback. You need to know the many ways in which you can elicit performances and provide feedback, testing being just one of these.
There is a key difference between this model and a behaviorist model. In behaviorism, you must see the impact of learning in behavior, so you must detect a difference in what the 4 x 10E6 neurons connected to muscles do to decide that there has been learning. There are 10E10 to 10E11 (100 billion) neurons in a human brain. Each neuron is connected to 100-10,000 other neurons through connections called synapses. One can have a hell of a lot of neural change (learning) before anything shows up in a motor neuron. There are papers that relate behaviorist notions to neuronal notions. Brembs et al. describe operant reward learning in Aplysia. Shidara & Richmond show changes in neuronal output based upon reward expectancy.
Modeling. What should be learned? What does it look like to work through a process? Parents, teachers, and others model performances, and these modeled performances can be learned. Modeling is not the same as most lecturing. Most lecturing seeks to transmit knowledge, something done efficiently by written material or a Web site.
Explicit instruction involves telling students. Many strategies advocate having students discover. In general, telling is much more efficient than discovery. The most efficient strategies are the ones that combine telling with numerous opportunities for feedback and application. Even when transfer to a situation where creativity and innovation are sought, prior direct instruction with lots of feedback works best. While not a widely held view in the science education community, this is the outcome you discover when you read very extensively in the literature. Direct instruction can be interesting, exciting, and fun as well as being productive. If you translate what this says to mean that lecturing with 2-4 tests and a final exam constitutes effective instruction, you've missed the point entirely.
Self-regulation. Skinnerian behaviorism has fallen into disrepute.
As a teacher, I was never 100% sure about was was a reward and what was a punishment. For example, is having those students with the highest scores on a test stand during class a reward or a punishment?
A couple of years ago, while teaching Curriculum and Instruction 859 (Instructional Message Design), I gave a Web reading assignment on cognitive load and also a copy of a reprint to each of 20 students. Each student got a different reprint, but we covered about 2/3rds of the citations in the general assignment. I read all of the papers over the Thanksgiving holiday. When class met, a student asked if I had read all the papers. When I said that I had, she said, 'you need to get a life.' Problem is, the most enjoyable thing I did over that weekend was to read those papers. I had a sense that many of my questions had been answered at the end of my reading; I was able to give some clear direction to several of my doctoral students. If there were some reward connected to that extensive reading activity on my part, it was internalized. In fact, after she asked me that question, I thought about it for some time before I decided that I really had enjoyed reading all those papers enormously and not just a little bit.
These days I describe essentially all such events -- such as my involvement with those readings -- using the jargon of self-regulation.
Piaget; the neoPiagetian era. For modern teachers, especially science teachers, and understanding of the ideas and impact of Jean Piaget is important. Piagetian notions imply that humans have stages of development. Piagetians see times when big changes or jumps from one stage to another stage emerge. Piagetians believe that teaching something before its developmental time is a waste. anyone who has raised children can relate to these ideas.
Let's look at this more closely, however. When you train a neural net, you provide linear feedback: a training example, a response from the computer, and then a 0 or a 1 based upon that response. The input is quit linear. The output is not at all linear; it is characterized by performance bursts. In fact, while each enural net trains differently, a study of numerous such nets tends to show exampls of every pattern of intellecutal development shown in small children. The first day a child rolls over, or the first time s/he takes a step is described as entering a new stage. If you focus on the output, you definitely see stages. One day, no steps; next day, steps, and next month, toddles. If you focus on the input, however, what you see is linear. It is both possible and productive to reinterpret the observations made by Paiaget in terms of other models.
My experience is that you make a big mistake waiting around for development. This is more a matter of performance expectation than exposure. It does not mean that 3-month-old children can be taught to talk or walk! What it does mean is that some learning -- neural modification -- will be taking place before you are likely to see a big change in performance.
Sometimes there are accommodations. Baby talk, for example, seems to emphasize the vowel sounds of the child's language. If you skip baby talk, then maybe you hurt rather than help a child's development. For an explicit example of a developmental difference, see Schlaggar et al.
Becoming an Expert. It takes thousands of hours to become an 'expert.' Ericsson has discussed many factors that contribute to achieving expertise, deliberate practice being key among these. Experts access libraries. Experts usually have a key mentor or coach who helps them. It's one thing to train employees to fill out a form correctly; it's another to become an expert musician or surgeon or engineer.
Expertise almost always involves something that is implicit rather than explicit. Successful teachers often are good at making things explicit.
Cognitive Load. Many of my students are engaged in creating multimedia materials. John Sweller has been a principal worker in developing the notion of cognitive load. You might say this is a formal way of thinking about the old teaching adage, KISS (keep it simple, stupid). To understand this work, you need to read about 40 or so papers. Cooper provides a good introduction. (I have loaner copies of many of Sweller's papers, and papers of other workers in the field.)
Interactive, Compensatory Model of Learning. Schraw developed ICML as a part of his teaching while at Nebraska. There is so much about teaching and learning out there that one easily can become confused. Obviously, intelligence and motivation both count. The questions are, in what ways do they count, and how much do they count.
Tutoring. Tutoring is the gold standard in instruction. Mastery learning is good. Keller Plan (PSI) courses generally have had good to excellent outcomes.
Training researchers. On the one hand, the data seems to support the notion that explicit training rather than discovery training leads to better performance in situations demanding innovation. On the other hand, I can't really imagine creating researchers without having some kind of practice and feedback. This is an area where I digress from my own system. I think that a research experience is a really good thing to include in teaching programs -- at all levels, but certainly beginning in college. To do this properly requires enormous resources, so it is something that needs to be done well.
Bloom, B. (1983). Human Characteristics and School Learning. New York, NY, McGraw Hill.
Bloom, B. S. (1984). "The Two Sigma Problem: The Search for Methods of Groups Instruction as Effective as One-to-One Tutoring." Educational Researcher 13: 4-16.
Brembs, B., et al., Operant reward learning in aplysia: Neuronal correlates and mechanisms. Science, 2002. 296: 1706-1709.
Brooks, D. W.; Schraw, G., & Crippen, K. K. (2002) Performance-related feedback: The hallmark of efficient instruction. http://dwb.unl.edu/Edit/P-RFeedback-4REFs.pdf
Cody's Science Education Zone. Guided Inquiry. http://tlc.ousd.k12.ca.us/~acody/5c.html Accessed 6/02.
Cooper, G. (1998). "Research into cognitive load theory and instructional design at UNSW."
Elman, J.L., et al., Rethinking Innateness : A Connectionist Perspective on Development. 1996, Cambridge, MA: MIT Press. 447 pp.
Ericsson, K. A., R. T. Krampe, et al. (1993). "The Role of Deliberate Practice in the Acquisition of Expert Performance." Psychological Rev. 100(3): 363-406.
Gopnik, A., A.N. Meltzoff, and P.K. Kuhl, The Scientist in the Crib. 2001, New York, NY: Perennial.
Herron, J. D. (1975). "Piaget for chemists. Explaining what 'good' students cannot understand." J. Chem. Educ. 52: 146.
Kulik, J. A., C. C. Kulik, et al. (1979). "A meta-analysis of outcome studies of Keller’s personalized system of instruction." American Psychologist 34: 307-318; Kulik, C. C. and J. A. Kulik (1987). "Mastery testing and student learning: A meta-analysis." J.Educational Technology Systems 15: 325-345; Kulik, J. A., C. C. Kulik, et al. (1990). "Effectiveness of mastery learning programs: A meta-analysis." Rev. Educ. Rsch. 60: 265-299.
Marcus, G. F. (1998). "Can connectionism save constructivism?" Cognition and Instruction 66: 153-182.
Moerk, E.L., The guided acquisition of first language skills. 2000, Stamford, CT: Ablex Pub.
NWREL, Science Inquiry Model, http://www.nwrel.org/msec/science_inq/
Pryor, K. W., R. Haag, et al. (1969). "The creative porpoise: Training for novel behavior." Journal of the Experimental Analysis of Behavior 12: 653-661.
Schab, F. R. (1990). "Odors and the remembrance of things past." J. Experimental Psychology 16(4): 648-655.
Schlaggar, B.L., et al., Functional neuroanatomical differences between adults and school-age children in the processing of single words. Science, 2002. 296: p. 1476-1479.
Schraw, G. and D.W. Brooks (2000). "Helping Students Self-Regulate in Math and Sciences Courses: Improving the Will and the Skill." http://dwb.unl.edu/Chau/SR/Self_Reg.html
Schraw, G. and D.W. Brooks (2000). "Improving College Teaching Using an Interactive, Compensatory Model of Learning." http://dwb.unl.edu/Chau/CompMod.html
Shidara, M. and B.J. Richmond, Anterior cingulate: Single neuronal signals related to degree of reward expectancy. Science, 2002. 296: 1709-1711.
Vygotsky, Vygotsky Resources, http://www.kolar.org/vygotsky/,