2/9-Elizabeth-a lot of levels of learning

The Wilensky articles touched upon many of the previous article’s and the practices of NGSS’s main ideas, especially the process of asking questions, investigating, creating models, collecting and analyzing data, then going back and fixing models.  All of the authors point towards the important computational models as a source of scientific learning.  

Main ideas:

·      Disciplinary learning should be an active study where students constantly observe, investigate, question, and argue their findings, not readily accepting ideas and theories.

·      Computational modeling is a way to bridge the gap between the classroom style learning of biology and the research style learning of biology.

o   Allows students to use their prior knowledge and personal experiencesàmakes science more interesting and relatable

o   Can grow/build from simple models to complex models

·      All science is interconnected and relatedàit is important for students to recognize this.

I, as I’m sure a lot of people, could relate to the opening paragraphs of the “Thinking Like a Wolf…” article by Wilensky, where he discussed the differences between the approaches to the classroom education of biology and that of biological research.  In high school and especially in college, we are taught solely to memorize facts and structures, rarely allowed to stop and think about the concepts as related and under the same branch called biology.  As a logical result, it makes the process of understanding biology much harder that it has to be.  Furthermore, when Wilensky stated that in contrast, researchers approach all data and “facts” as dynamic and ever-changing, never readily accepting it without further testing through modeling.  This active quality of research biology that is absent from classroom biology is critical to understanding what biology is and that it is not a stagnant field of study.  As we touched upon this last week, if high school biology curriculum were to change to incorporate computational modeling and provide students to be active agents within the field, would it lead them to be more or less prepared for college?  While the goal of high school is to graduate students and not necessarily to prepare them for college, as some may not go, how would it prepare those who are going to college?    

One question I have is what would this look like for a chemistry unit?  Wilensky talked about biology computational models but other topics may not have as many or as interesting computational models.  Many chemistry computational models focused on kinetics.  When I used Netlogo last semester and was putting together my final project, I had trouble finding an exciting model that the students could use.  I felt as though the models could only be used by upperclassmen as they were complex and not as hands on as OneTurtleJar.  Furthermore, I felt as though there was less you could manipulate with the models and they therefore would not be a major focus within a lesson.  While I say this, I would like to learn more about it and find better ways to incorporate it into a chemistry classroom.

2/9 - Kim K - Leveling Up Bonus! +10 Computer Skills

A major theme I’ve noticed is about investigating how complex phenomena can arise from simple components and simple interactions.  Most of the time, there is a lot more behind phenomena than we might think or initially observe, and one will only figure that out after deeper investigation into trying to explain the processes behind them.  Investigations are, of course, key processes to scientific inquiry, and what teachers should expect their students to learn to do naturally.

I am very excited to work with StarLogo because the possibilities for modeling different phenomena open up so much.  In our last class, I had really wanted to see how two turtle agents could be programmed to interact with each other with regards to planets’ motions within our solar system.  There are so many other uses I can think of now too.  However I did appreciate the introduction to Logo programming with NetLogo, where we could only manipulate one turtle agent.  This allowed me to get a sense of what the programming itself does without worrying about levels, and I think if I were to use Logo in the classroom, I would introduce my students to it in a similar fashion.  Asking my students to think of modeling a concept/phenomena with only one agent will allow them to realize on their own how powerful manipulating multiple agents could be.

I think that the GasLab shows how crucial programming can be for experimental labs that are not easily reproducible outside of the computer-modeling environment.  This also corresponds with their statement, “the StarLogo modeling language enables much younger and less mathematically knowledgeable students to have access to explanations that connect the micro- and macro-levels of phenomena.” This definitely supports diSessa’s discourse we read last week and is probably Wilensky and Reisman’s basis for their Thinking Like a Wolf paper where they seek to remove the barriers of formal mathematical requirements so that students can experience meaningful engagement.  Wilensky and Resiman would also agree with diSessa’s discussion about tool-rich cultures, “the way that we see the world is greatly influenced by the tools that we have at our disposal.” Which leads to my question: How do we address the major challenge of developing a better understanding of when to use which approach and why?

Modeling Problem:

How many people does it take to initiate a “human wave” at a game?  I’ve noticed that it sometimes takes a few tries to get started, so can one model the influences of this?  There are probably different levels contributing to this, such as how exciting the game is and other crowd participation that might be encouraged throughout the game.

2/9 David Bergsmith Levels

            Both Wilensky articles this week used examples found in research to show how modeling can be used to develop understanding of science on small scales to further understanding on broad views. This speaks directly to Wilensky’s view of ‘levels’ and how to effectively develop knowledge at a simpler point so that students may have a better chance at comprehending and exploring topics at a higher level. Research in both articles used examples of students and how they created models to represent scientific phenomena. Students engaging in this practice developed questions and challenged their own reasoning. They were asked to hypothesize and revise their models at least once; creating a better understanding of the process they were asked to model.

Wilensky gave many different examples at how asking students to create models may develop their understanding of a process. Students created initial lines of code to begin a model and hypothesized what would happen. In all the examples, students were then asked to revise their models after they made observations that were not expected. This type of reasoning cannot only be used in computational models, but in all practices of science and engineering. On page 172 of the “Thinking like a Wolf,” article, Wilensky speaks about how science as trended to a prescribed procedure rather than reasoning from gathered evidence. In the articles, the students developed an understanding of the material not from following a set of rules, but rather engaging in observation and investigation.

The models that the students used also had characteristics of the practices from the NGSS. Specifically, I noticed the students asking questions and defining problems in their models. While the initial rules that students created for their models were good, after the first trial students became curious as to why the model did not run completely as expected. Students defined specifically what they were curious about so that they could change the code and the outcome of the model. Some students were asked to analyze and interpret data for their models. These are some of the specific practices from NGSS.

            After reading these two articles, I began to wonder how computational literate the students from Wilensky’s research were. Specifically, questions such as what training had these students had, was this training formal or informal and if it were formal was it an elective course or required, or even at school at all? On page 17 of the “Thinking in Levels,” article, Wilensky says that model use should be used in a pluralistic approach so that strengths and weaknesses are adhered to. What approaches could be used to help students who are alienated by traditional classroom mathematics? Models could scaffold knowledge of algebra, but also be uninteresting to a student feeling academically inadequate. What strategies could be used so that students who are uninviting to a program that uses formulas and variables so that they are interested and develop an understanding of tools such as algebra and concepts such as predator-prey relationships?

2/9 Joey: Thinking on a Different Level

I think Wilensky and Reisman hit the nail on the head when they commented, “In school settings, typical instruction emphasizes the memorization of classification schemas and established theories” (pg. 172).  I know I can remember classes where I had to memorize parts of a cell, or memorize a certain pathway/cycle.  Not only was this type of instruction monotonous at times, but it was also not very effective as I have forgotten many of the pathways.  Engaging in computational models where students get to explore and create their own theories and hypotheses appears to be much more exciting and effective.  The processes described also fit very nicely into the eight practices the NGSS framework recommends.  Students were able to ask questions about a phenomena, develop a model, plan and carry out an investigation, analyze and interpret the results, use computational thinking, construct explanations for results, engage in argumentation from evidence (other research papers comparing accuracy of model results) and obtain and communicate information to others.  Although Wilensky and Resnick focus more on thinking in different levels, their examples imply very similar ideas.  I really liked the Biology examples (Predator-Prey models, Fireflies, and Slime Mold).

I used a modeling program called Populus (I think), and we were able to mess around with the Lotka-Volterra model by changing the values of the variables in the equations.  Although I could see the effects over time by the graphs the program produced, it was more math oriented and less fun to interact with than the NetLogo model.  Perhaps the more mathematical model would be more appropriate for college students with some knowledge of calculus and the NetLogo model would be more appropriate for students with less background knowledge in calculus?  One of the affordances of the NetLogo model is its ability to explore complicated ideas with less intense math attached.

            I found myself thinking as I was reading about the flashing fireflies that I would probably start by using a more deterministic centralized mindset (where the leader gives deterministic orders to his or her followers), but this would have been the wrong approach.  I can’t help but wonder what the result would be if a student (or myself) took the time to create this model and work out all of the bugs, only to read later in a research article that their mechanism was completely wrong.  I would no doubt be frustrated and feel my time was wasted.  However, many revisions are usually needed in any model built from scratch and that is part of science. 

            A few questions I have going off of that thought is: At what point should students be able to look up/research information about the phenomena they are looking into?  Would having that information before creating the model take away from the inquiry process?  Or would having relevant background information aid in student thinking and help give some direction to the model they will build?

            It would be interesting to try to model how the shapes and characteristics of different trees leaves effect how the trees respond to different climates. (sharp pointy needles vs. broad flat leaves, deciduous vs. non-deciduous, nutrient availability, tropics vs. temperate forest)

2/9 Caitlin - Computing Complexities

The two Wilensky readings are about how embedded computational modeling can help students learn about complex scientific systems and concepts. Many of the points made in the Wilensky papers correspond to the arguments made in the other papers, such as Nersessian, Shwarz and diSessa. One of them being that science students are owed a chance to learn the same tools that a scientist would be expected to have. For example, being able to observe, make theories and revise those theories, instead of simply learning what has already been learned. Wilensky also says that students who learn what doing science means, they will be able to appreciate the discipline more, whether or not they go into the science field as a career.

A major point emphasized in the studies was the idea that science concepts are usually complex and are inter-dependency and interconnectivity on other systems, and that these systems can be explored and learned through computational modeling. One of Wikensky’s studies used the term ‘levels’ to describe the interconnectivity between larger and smaller systems. Wilensky’s other study focused on inter-dependence between groups and individuals (wolf, grass and sheep, and fireflies coordinating). Students should have to be able to reason and think about connections between systems, as many concepts they will have to learn will be dependent on other concepts. 

The computational modeling helps students see how individuals, following certain rules, can create group (higher level) phenomena. In other word, the students can move to understanding the mechanics behind an event. When I had a chance to play around with a similar Netlogo wolf-sheep model, I was able to see more clearly what factors could affect how long the population could survive. I was able to come to the same conclusions as Talia; the grass, the number of organisms, and other mortal factors lead to how steady the populations will stay. This is similar to what diSessa was arguing about for having computational modeling in the classroom. Modeling can help students see how something works (the kinetics behind the kinetic formulas, or the cars in a traffic jam), instead of just knowing that it works.

I am curious as to how long the students had to complete their models and learning on the wolf-sheep and firefly models. In a course, I will only have so much time to give to a unit. Furthermore, I will still have to teach my students the material that will be on end of year tests. If the NGSS, or similar standards, is implemented, I might have more time to give my students the chance to experience more deep and constructive modeling. However, if they are not, then how can I cut down the time that such experiences would take? Would careful, but efficient scaffolding help? Or would I have to decide which concepts in population dynamics I would want my students to learn? Furthermore, how much scaffolding would I need for my students to reach the same reasoning and thinking skills the students in the Wilensky studies were able to do?

2/9 Steve: Modeling beyond one turtle

1.       Articles

a.       These articles made me think about how it is so important to consider different topics within a course like physics as requiring very different instructional methods.  Not all science topics can be done with modeling, but there are certainly a large number that would be so much better with this kind of agent based modeling.

b.      Obviously in a biology class these examples would be very helpful showing the questions teachers need to ask students as they test their models. Having good investigative questions for students engaged in modeling seems critical with these kinds of tasks. 

c.       The articles continue the theme of models that constantly adapt and change with new information and test results. Another continued focus is the importance of pushing students to evaluate their models over and over again.  Also stressed is how models are more than just showing other people how something works; they give the model-makers insight into how the system works as well.  In this case, the model-makers have to put themselves into the shoes of the wolf/sheep/slime and look at the system from that perspective. 

d.      Most of the authors so far seem mostly in agreement.  DiSessa so far has stressed a more general computational literacy and showed us some concepts that it would be useful for, while these pieces provide specific case studies of computational literacy being used to better understand a concept.  Nersessian’s work was similar to this week’s papers but in a non-educational context.  This week’s authors suggest that science is more like modeling, and Nersessian’s article gives anecdotal evidence of that claim.

e.       These pieces on agent based modeling made me very excited to get to model with more than one turtle.  The degree of complexity that we can see thanks to this kind of modeling while only knowing some very simple things is amazing.  I hope I get to apply it to some cool realms of science.  Playing with variable such as chance of reproduction seems like it could be a fun logical exercise as well.  I find computer science modeling like what we do in class to be so helpful for teaching how to look at things from other perspectives.  Even with just one turtle acting, I often catch myself thinking of turning right or turning left relative to my north rather than the turtle’s heading. 

2.      Questions

a.       How does assessment play a role in modeling exercises?  Is it mostly effort based?

b.      I would be curious to see what a great entry event into this wolf modeling activity looks like.  I can imagine showing a video about wolves and elk in Yellowstone.   Can anyone think of a good entry event for the slime modeling?

3.      Possible questions to model

a.       How do we get maple syrup?

b.      How does lightning decide which path to take?

c.       How does the golden ratio emerge from a spiral sea-shell?