Human Learning
From CS260SP09
Human Learning
Readings
How Children Learn Chapter 4 from "How People Learn," Brown et al., National Academy Press
Discussions
Please post your critiques/commments on the required readings below. To do that, first login by using your user name and password, then click the "edit" tab on the top part of this page (between the "discussion" page and the "history" page), New to wikis? Read the Wiki editing guide. . Hint - Please put a whole line == ~~~~ == (literally) at the beginning of your submitted critique, so the wiki system will index, sign and date your submission automatically.
KetrinaYim 18:50, 25 January 2009 (UTC)
Reading about how learning is affected by adult-child interactions reminds me of Jean Anyon's "Social Class and the Hidden Curriculum of Work" (Journal of Education, Vol. 162, no. 1, Fall 1980). In this essay, Anyon describes a small study in which five elementary schools in New Jersey are classified according to their locations (city or suburb) and the socioeconomic status of the students' parents. There were four categories, working-class, middle-class, affluent professional, and executive elite, and the differences in how the students were taught to learn and think, regardless of subject, were striking. In working-class schools, teachers gave children instructions and direct orders and showed them only one way to the right answer. Middle class schools encouraged a little creativity in solving problems, but were mostly geared towards finding correct answers and absorbing facts. Affluent professional schools provided more creative activity, teaching children to express and apply ideas and concepts. Finally, executive elite schools encouraged reasoning about problems, allowing children to develop analytical intellectual skills.
While this cannot be generalized to apply to all schools, it is as if the educators foresaw what careers the children would end up doing and prepared them by establishing the appropriate mindset. Their assumptions seem to be that blue-collar workers follow orders, middle-class workers use knowledge, professionals develop ideas, and executives apply reasoning. Thus the children in these schools are taught such assumptions through the above-mentioned "hidden curricula", in an attempt to shape how they learn and solve problems later in life.
Nicholas Kong 02:33, 26 January 2009 (UTC)
@Ketrina The study you reference is certainly interesting. However, I think that the resources available, particularly the number of teachers per student, may play a bigger role in explaining the different pedagogical techniques. Teaching students to exhibit creative activity is more time consuming than teaching students codified methods, as I would hypothesize that more student-teacher interaction is necessary in the former case. I have not read the paper, but I am guessing that the more affluent the school, the lower the student/teacher ratio.
While reading the chapter I was particularly impressed by the methodological techniques that were developed to test cognition in young infants, and further the remarkable concepts that infants already have. The fact that infants of 6-months already have the concept of number was surprising. An additional takeaway from the chapter was that many of the ways in which adults solve problems (such as forming schemas and generalizing, or using categorization techniques for recall) manifested themselves more explicitly in children (e.g. the Big Bird example). Thus, studying how children learn has ramifications for adult learning, and by extension to UI design. By discovering, for example, how humans use schemas, we may be able to leverage existing schemas for a more "intuitive" interface.
James Andrews 08:46, 26 January 2009 (UTC)
About Ketrina's study, I wonder if there is also some difference in the teachers at those schools -- it would be interesting to see, for example, if teachers tend to teach at schools similar to the ones where they were taught. It also seems possible that more qualified teachers would be attracted to higher-status schools.
About the reading -- I'm a bit uncomfortable with the use of habituation in studies to demonstrate conceptual understanding, where to me it seems to only demonstrate a more-shallow understanding of what events are common or uncommon. For example, the chapter points to studies that show infants look longer at "impossible" events, such as a floating box, and then says this indicates they understand the underlying concept of why that event is impossible. But the infant could just as well only recognize that a floating box is a 'less common' event than a non-floating box, without needing to form any explicit notion of the underlying concept of support. This could help explain the apparent disparity mentioned in the chapter, where habituation studies indicated an understanding of the need for point of contact well before children could really use that understanding to accomplish goals.
Anuj 08:57, 26 January 2009 (UTC)
@Ketrina Though the study you talked about definitely sounds interesting to me, I would like to look at the rationale behind the classification of schools, closely. After going through the reading I am less sure if children at this place in New Jersey pursue careers that they were capable of, or are forced to pursue some career, because that's all they were trained for. The social interactions with the people around, and the information exchange with them lays the foundation to how a child thinks. If he/she is trained to be something, it is possible that he/she habituates to the same and pursues the same.
Moreover, the reading asserts the belief that to be able to design for humans, we need to understand the background they come from. The cognitive process of a human being is heavily influenced by his childhood and also by the various adaptations made while growing up (which in turn are a result of the environment). Therefore, "one size fits all" approach does not work well as far as designing for human subjects is concerned. Designs should be contextual for the user, and should bear an impression of the background of the user.
David (Tavi) Nathanson 10:58, 26 January 2009 (UTC)
@James You make a very interesting point when you say that an infant's successful differentiation of "impossible" events from typical events does not necessarily indicate a true understanding of support. Rather, the infant might simply look longer at the event that he/she is not used to seeing! However, upon thinking further about it, wouldn't the infant need to understand support in order to *know* that the "impossible" event is unusual? In other words, if you compare the image sequences of one possible event and impossible event (Figure 4.1), you'll notice that the image sequences are very similar: both have two boxes that go through very similar motions relative to each other, and the key difference is really the placement of the hand. Would an infant really be able to distinguish these two sequences without an understanding of support?
On a different note, I am intrigued by the fact that children use diverse strategies to add numbers (i.e. they might count from 1, they might use their memories, etc.). This nicely parallels the fact that people often interact with user interfaces in diverse ways: they might perform very different actions to reach the same goal. In both cases, when one has various different strategies for accomplishing tasks, he/she can choose a strategy that makes sense for the task at hand.
Mao Ye 12:03, 26 January 2009 (UTC)
Basically, I feel psychological studies are not very convincing and reliable: limited by the "intelligences" of the researchers, the explanations are often more than farfetched. For instance, in the experiment about animate and inanimate objects (Page 88), one alternative explanation other than "Biological Causality" may be: animate objects can provide more subtle details that can attract infants' attentions and then the time for an infant to keep being not "perturbed" may be proportional to the amount of the attracting details the objects can provide, which has nothing to do with "biological stuff".
The common mistake for most of infants/children learning studies conducted by adults, in my humble opinion, is that adult researchers tend to explain results from adults' angles of view. It is possible that infants/children mimic adults' words/speeches/behaviors with no understanding all. For them, shuffling words/phrases to create some "meaningful" sentences, composing poses to show some "significant" actions, which mean "understanding" for those adult researchers, is just as simple and unconscious as randomly playing puzzles, which might mean nothing about "understanding".
Kant and some others believe, knowledge cannot be learned, instead, it COULD BE discovered. One day a year ago, I talked to my 2-years-old child about "death", saying "Death means somebody can no longer be seen. Now, you are too young to understand this, but someday you will." Surprisingly, just one week later, when watching an old lady's head "cut-off" in a comedy on computer, the child shouted "the old woman dies". Given the fact the child had never seen/heard of dead things, such a learning process cannot be explained by any modern psychological theories. I would like to believe the child was discovering the concept.
The "knowledge discovery" theory could be potentially supported by a preliminary scientific research that finds memories may be stored on human's DNAs (http://www.newscientist.com/article/mg20026845.000-memories-may-be-stored-on-your-dna.html).
Anyway, I suppose we are in the early stage of learning "learning".
James O'Shea 17:27, 26 January 2009 (UTC)
I think David brings up an important point when he mentioned that children will use different strategies for relatively simple tasks such as addition. This is certainly true for adults as well, yet the strategies may be even more diverse depending on the complexity of the task. Related to this is the fact that prior experience can play such a major role in learning. For children, it may be relatively safe to assume the nature of their prior experiences, but this would be difficult to do for an adult given they've had a lifetime of experiences on which they may be basing their learning and actions. I think one of the challenges for designing effective computer systems is having a proper understanding of the users' prior experiences and how they may shape their approach to the UI.
On a separate note, I found the section on guided learning very interesting. This seems like an area in which technology could play a major role in helping to direct learning or help users understand a computer system. Story-telling is considered such an important tool in learning, and I think this framework could probably be adapted for different UI or tutorial designs.
Yuta Morimoto 17:54, 26 January 2009 (UTC)
From readings, I think the process of infants and children is not simple and very unclear. Although the different stages of learning process indicate determined process of ther learning. the order of the process does not possesses reasonable explanation. So, after an infant learned certain thing, what is the next? Seemingly, infants and children gradually obtain higher, abstract and metagognical skill. But the order of the process is still unclear in some sense.
The reading has pretty much insight on learning process of a child. One of the most interesiting thing I encountered is children's learning process. In early age, children interact physical object to understand the world around them especially when understanding how objects interact and what natural behaviors is. From UI perspective, they are constructing conceptual model: their concepts are possibly some mistake. I think the important thing we can apply to more elderly people is that it seems that basic methodology of learning in adults is not still changed even if they use many sophisticated ways to learn.
Seth Horrigan 20:47, 26 January 2009 (UTC)
Wow, Seth sure was great today, wasn't he? Oh, wait, these things aren't anonymous. Dang!
As I was reading through this chapter and chapter 1 of the NRC report, I was strongly reminded of John Holt's book "How Children Fail". In essence, it is a long series of case studies describing specific students as they develop meta-strategies, intelligence theories, socio-cultural interaction, and in some cases are helped by the educational system, but more often their desire to learn and their learning capacity is hindered or crushed by the institutional schools. In the end, Holt realized that institutional schools are not focused on teaching children but on continuing to do what they are comfortable with. New teachers will enter the system, full of enthusiasm and good ideas, but over time they are forced to either conform to the standard or are forced away. The overwhelming emphasis on summative evaluations and preparing students to perform well on standardized tests drives out any chance for spontaneous or learner-motivated inquiry and replaces it with curriculum designed to fit students to an ideal knowledge of a set of facts. Also, as John mentioned, the emphasis on rote memorization over understanding of concepts (as exemplified by the disapproval of "cheat sheets") leads to worthless batches of unconnected, unmotivated knowledge that either lies dormant or is abandoned.
John also mentioned that thanks to the experimentation and structure in good educational research, alternatives like home-schooling are becoming increasingly viable and even superior options. Two books and twenty years late, Holt came to the conclusion that home-schooling is the way to go, but obviously this requires more of the parents' time, and if done incorrectly has the potential to produce students no better than traditional educational institutes produce.
Kenrick Kin 23:12, 26 January 2009 (UTC)
Late comment. Our discussion today reminded me of this article: Inside the Savant Mind: Tips for Thinking from an Extraordinary Thinker.
Tammet holds the European record for reciting the first 22514 decimal points of Pi. His description of how he treats words and numbers is an interesting one. He describes them as having form, color, texture, etc, and thinks of them as "friends." He is considered an autistic savant so perhaps his mental development was different than for most people. Are there other creative ways to teach children that may better match his style of thinking? Or is he such an exception that methods of learning cannot be transferred. In any case, this certainly calls into play the topic of Multiple Intelligences. Also, similarly to how we tried to segment a string of letters into meaningful groups, Tammet chunks groups of numbers into meaningful visual images.
I agree with James O that guided learning definitely seems like an area where technology can have a great impact. Guided learning principles can be used in generating effective interactive tutorials. "Known answer" questions or asking a subject to perform the action can be applied to confirm a subject's understanding of the task before moving to the next step. This is apparent in in-game video game tutorials where the player is often asked to repeat the task he/she was just taught. On a side note, I thought the discussion on the cultural differences on the role of questioning is interesting and I wonder if one method is more effective than another.
Priyanka Reddy 05:41, 27 January 2009 (UTC)
Like James A, I'm not sure I agree that habituation is an adequate representation of a child's conceptual understanding. Instead, it seems to demonstrate how well acquainted the child is with the situation in the picture. ie. the child stares longer at the floating box because he's never seen it before. However, I wonder what would happen if that same picture were shown to him multiple times a day for many days. Would he become habituated and thus stop staring for so long at the picture? Or would he still believe that objects need support to prevent them from falling and deem the picture to be impossible? I'm leaning towards the former idea because although the child has never seen a floating box, is that reason enough for him to believe it's impossible? In my opinion, the child would need be able to try the scenario out for himself and then discover, through demonstration, that it is impossible. This experiment reminds me of the stories we hear about children who, after watching people fly in movies and video games, attempt to fly. If children really understood the physical concept of support, I would think they wouldn't believe they can fly.
On a different note, I found the idea that children and adults have the same M-space, but that adults are able to overcome general limitations through the use of strategies really interesting. It seems like a good explanation for why 2 people of the same age, and therefore the same M-space, learn at different rates. If the M-space were the determining factor, it seems as though the 2 people should learn at the rate; however, if learning depended on the individual's ability to overcome general limitations, it could explain why 1 person would be able to learn faster.
Brian Tran 06:50, 27 January 2009 (UTC)
I agree with James O on how designing effective user interfaces relies on understanding their prior experience. People act based on experiences that affect their expectations of end results of different actions. Becoming familiar with these experiences is crucial to understanding what people expect to happen in applications and thus make those applications intuitive. A very easy example of this is how many people learn new languages. Upon hearing a phrase in the language, they will break the phrase apart into separate words, translate each word to their native tongue, put the translated words back together, and then process the meaning of the phrase. They decide how to respond in their native tongue and then translate the phrase into the new language before speaking. It is clear that people who have that new language as their native tongue do not learn new words in the language by translation to another language. Thus, people around the world learn English, Spanish, or any other language through a very different thought process.
The article’s reference to zone of proximal development reminds me very much of how our schools work. We expand our knowledge by gradually building on top of things we already know. Schools will first introduce addition, and then introduce subtraction as the opposite of addition. Multiplication is then introduced as doing addition multiple times. We are taught these particular algorithms of doing addition, subtraction, and multiplication in American elementary schools because they rely on previously taught algorithms.
Simon Tan 03:30, 29 January 2009 (UTC)
@Kenric: I've heard of Tammet and the "savant mind" before, and that article you linked definitely suggests that his mind develops much differently from the minds of most people. The way he thinks about numbers and words is uniquely his own; he finds it surprising that no one else thinks the way he does, as we are surprised at his abilities. I don't think that applying our way of teaching (i.e. the U.S. public school system way) works for people like Tammet, as it assumes a certain way of thinking about numbers and words. While it may not be worth it to come up with a creative way to teach children with a mental model like Tammet's (as there are so few cases like his), the issue of teaching children with a wide variety of learning styles (or multiple intelligences) is one that hasn't been really addressed in mainstream schooling.
As I was going through the reading, I wondered about how development varies between different cultures and hoped the authors would address it in some form. It was interesting to see them capture this in the section on cultural variations in communication, especially in the study comparing middle-class white families and working-class African-American families. I had no idea that the role of questioning could be so different between communities, so much so that assumed methods of teaching could work with one group of students and not another. The teachers' quotes on the African-American students were surprising: "They don't seem to be able to answer even the simplest questions." If a teacher got this type of response from her students and didn't know any better, perhaps their natural assumption would be that the students were underperforming. With reports of whole schools that are "underperforming" and happen to be predominately composed of one race, perhaps a mismatch of teaching and learning styles are to blame. Why have I never heard of this possibility before?
