2/2 Elizabeth-diSessa (2) and Neressian

In the Neressian article, all 8 practices, from asking questions to constructing explanations, were used.  This is so mainly because it was a novel study that was conducted and there was a lot of room for investigation, explanation, and revision.  First, they started out with a question regarding the communication between neurons when learning takes place.  Then they developed the dish-model system and carried out a four year investigation.  Throughout those four years data was continually collected and additional computational models were created.  This was evident with the “bursting” component of their study, which prevented the detection of learning, which was critical for data to be collected.  Thus, the scientists had to come up with a way to prevent bursting (designing solutions and constructing explanations) and included one researcher creating a computational model of the system to promote progress and understanding of the neurons and their interaction.  Throughout the whole process, positive and productive argumentation took place from the obstacles they hit to the data that was collected.  While this study was new in its field, the information collected from this study has made a huge innovative step towards further understanding neurons and their interaction when learning. 

In Chapter 2, diSessa focuses on computational models as ways for the average student to grasp and understand concepts that “geniuses like Galileo” even struggled to understand (34). This aligns itself with Neressian and her article, where the researchers continually made computational models to further understand what was going on in their experiment, especially when no study like their’s has been done before.  Both authors point towards the importance of computational models as a means to observe phenomena that could not otherwise or more easily be observed.  This was evident in the diSessa article where the students could manipulate the program by changing certain variables and see how, as a result, the motion was affected.  Through the use of this program, students slowly began to build upon their own models of motion and acceleration, just like Neressian and her fellow researchers did during their own study, where they continually changed and added to their model as they came across certain obstacles.       

2/2 Jenna - The Individual and the Tool

In her article, Nersessian describes how neural engineers used multiple models to understand neural networks in order to build a neural system that could learn. Modeling is at the heart of her argument, and it is by far the most prominent NGSS practice in the piece. However, because she is looking at a community of scientists, we also see many of the other NGSS practices at work. We read of their questions and the process of defining their problem space. We read how the scientists synthesizing work from a number of disciplines to construct explanations and design solutions. This integration of domain enabled them to access their problem space and define the parameters of their models. Finally, by carrying out investigations with the models, the scientists were able to generate novel concepts about their field.

Nersessian's account of modeling states that "novel scientific concepts arise from the interplay of attempts to solve specific problems, use of conceptual, material, and analytical resources provided by the problem situation, and often through model-based reasoning processes" (2012, p. 1). This is consistent with diSessa's ideas about computational literacy, because he views this as an essential tool for scientific practices. Computational literacy (modeling through programming) requires conceptual analysis, but supports discovery through lower levels of abstraction and the ability to view the system moment by moment. Through programming, the learner builds a system from the inside out, and sees how the underlying principles fit together like puzzle pieces to produce the observed phenomenon. The versatility of this literacy empowers the individual to participate in many tasks (beyond modeling) within the scientific community. Furthermore, the tasks do not have to produce novel concepts for the entire community; the learner is equipped to inquire about the world and construct concepts that are interesting and novel to him- or herself.

2/2 - Kim - Decoding diSessa

From the NGSS, these core ideas for K-12 science instruction feature prominently in Nersessian:

1.        Have broad importance across multiple sciences or engineering disci-plines or be a key organizing principle of a single discipline.

2.        Provide a key tool for understanding or investigating more complex ideas and solving problems.

3.        Relate to the interests and life experiences of students or be connected to societal or personal concerns that require scientific or technological knowledge.

4.        Be teachable and learnable over multiple grades at increasing levels of depth and sophistication. That is, the idea can be made accessible to younger students but is broad enough to sustain continued investigation over years.

Nersessian wants conceptual innovation to transfer across time periods and methods of analysis because that is how it has been changing and advancing in present-day science.  I believe that diSessa’s ideas about computational literacy fit into this idea of transferability extremely well.  DiSessa asks us to assume that the future of computer programming will be accessible to learn for elementary-aged students.  He also emphasizes how programming, and his tick model, is the best way for students to learn about motion.  To him, the inferences students would learn from algebra and calculus are not enough because they are not synthetic, one cannot “experience” it.  I think on this point and a few others, my beliefs are in contention with diSessa because learning the basics of algebra and calculus are how students would begin to even understand what the computer is even synthesizing for them in the first place.  Maybe I do not “get it” because I learned the “old-fashioned” way, and maybe my understanding of motion would be far superior had I learned it via programming and the tick method.  I feel like this is an important consideration that diSessa has glossed over, but I do not think anybody could know for certain which method is best.

2/2 Laura: Adding Computation to the Toolkit

         The practices presented in the NGSS standards are intended to represent the cycle of authentic practice in contemporary science laboratories.  In Nercessian’s article, we see each of the NGSS practices at work in the neural engineering lab studied, which confirms these practices as authentic and also as valuable to discovery.  While each of the steps is included in Nercessian’s depiction of the lab, she focuses on modeling, interpreting data, computational thinking, and communication most prominently.  Each of these practices supports her thesis that discovery is a dynamic process, as the lab went through multiple revisions of their model, changed their interpretation of burst patterns, and utilized computational thinking to better more effectively abstract and see the significant inputs to the system, all while communicating and arguing for different hypotheses, ultimately producing a discovery that was the product of many brains working and communicating together. 

            diSessa argues that computational literacy can provide the literate with a toolset that can facilitate more advanced results.  In Nercessian’s sample lab, D11 exemplified diSessa’s computational literacy by creating a computer model to more effectively abstract and understand what was going on in the physical model.  Before the computational model, the lab understood the bursts as noise, but when the bursts also occurred on the computational model, it was clear that they were not random and instead were the result of some combination of conditions inherent to the system.  Because computational modeling allowed for the more precise isolation of stimulants, D11 and company were able to make greater strides with their physical model and ultimately make a discovery about learning that may not have been possible without the computational model.  I think it is also important to note that computational thinking does not act in isolation, but rather supports the process of modeling and is just one of many tools scientists can, and should, use to get the best understanding of what is happening in the actual system. 

2/2 Joey: Discovery with diSessa

I would say that in Nersessian’s paper every single one of these practices were prominently involved throughout the research.  The researchers had a question about neurons and defined a problem to solve.  How can we use technology to build and better understand “a living neural network”?  They developed multiple models (conceptual, physical, computational) to help represent the phenomena and system in question.  They planned and carried out many investigations, analyzed the data and made tweaks to fix problems.  Math and computational thinking were involved in the construction of models and designing solutions.  The researchers were constantly constructing explanations and designing solutions.  Along the way they would use data and other resources to argue using evidence.  As many different researchers came together to complete the solution; they obtained, evaluated, and communicated information to each other along the way.  Also, by having this research published, this information was communicated to me as well.

diSessa describes many ideas about computational literacy and many aspects line up with Nersessian’s account of scientific modeling. Just as Nersessian documents the use of computational models, diSessa feels these models can be extremely useful as well. diSessa also mentions, “Programing turns analysis into experience and allows a connection between analytic forms and their experiential implications that algebra and even calculus can’t touch” (diSessa pg. 40).  Just as Nersessian notes how these models can be used to observe and investigate phenomena the human eye can’t clearly see, diSessa notes the same advantages.  diSessa also points out how computational literacy involves scientific inquiry and leads to discovery.  In both cases computational literacy is viewed as something that will allow one to investigate a phenomena, create a representation, develop a model, gather data, make revisions, argue through evidence, and communicate findings to others.  diSessa really sees this new type of literacy as the future and a truly unique opportunity for the community to grow.

diSessa clears up some of the questions I had previously about to what extent teachers would have to be familiar with computers and new language to be successful at teaching this new literacy.  When diSessa writes, “I am taking computer programming languages to be a material form for a hypothetical new literacy, and I’m assuming that programming is within the grasp of elementary school students” (diSessa pg. 34), it becomes clear that I need not worry of being fluent in binary code and some of the more advanced computer languages.  Also, after seeing how each student in our class came up with a unique way to represent uniform acceleration, I more clearly see how the possibilities for creativity can be endless.  

2/2 Dan: diSessa (again)

The NGSS laid out 8 practices for K-12 science classrooms. While arguments could be made that both articles addressed all of these topics, some were more prominently featured than others. The Nersessian article focused mostly on Practice 2: Developing and using models. A key point in her article was how the use of a computational model of the synapses allowed the researches to measure and control critical variables with a much more precision. It also focused on the process of developing a model, highlighting that it is a continual process that can always be perfected and refined. Practice’s 4: Analyzing and interpreting data, and 5: Using mathematics and computational thinking, were also addressed in some detail. The team developed new ways of displaying their results, creating a visual of the network that could map they type of bursts and where they occurred over time. Using vectors, they were essentially able to plot the center of activity trajectory (CAT). These achievements required them to analyze the data they were collecting and use mathematical processes to communicate their results.

diSessa’s chapter’s focus on how computational literacy can enhance students ability to grapple with difficult and sometimes abstract concepts. He shows how a study of motion can be inaccessible to students without a strong foundation in algebra and calculus, but through computational modeling and programming, students can engage in discovery about motion at a much younger age. I see this idea aligning with Nersessian’s article that focuses on how we can use modeling, specifically computational modeling to learn about previously unreachable topics, in his case, the brain. While diSessa’s article made an argument for computational modeling and programming as an instructional practice and Nersessian was more detailed in outlining the steps involved in developing a model, both were able to show how these practices provide unique advantages. For example, diSessa highlights the way in which students can see how the up and down processes of a ball in flight are really the same thing through the s and a vector sliders. Nersessian explores a similar example in that the researchers were able to see how the bursts were not just random synapses firing, but that the could in fact track the movement of activity through the CAT. Both authors would agree that computational modeling and programming can offer unique insights that traditional mathematical methods cannot easily access.

2/2 Caitlin - Practice Modeling II

NGSS describes several scientific and engineering practices that every student should learn by twelfth grade in order to be literate in science. Some of these were more or less prominent in the Nersessian paper. A couple of these practices include the building of an experiment and collection and usage of data and observations. Creating an experiment is the first step in a long process for studying a scientific question, and a researcher needs to decide what data is relevant or important to answer the question that is asked. Nersessian described how the researchers in the case study used computational modeling to create a simulation model of a neural system. This supports diSessa’s idea that computational literacy should be something that anyone should be competent in. Modeling, and computational modeling, is important throughout the scientific and engineering processes.

According to NGSS, students should be able to create and use models to help them analyzing and interpret data and observations. In the Nersessian paper, the researchers used modeling, including computational modeling to help them analyze their data. It was shown that modeling also helped the researchers revise and improve their study. Computational modeling showed the researchers how an event that they originally thought was something to minimalize in their experiment was actually something that was important to their study. Modeling information observed from data collection can lead to findings that were not originally predicted, leading to adjustment and revision of the concepts or hypotheses. The computational modeling helped them understand what they were observing, which is what diSessa would want to see as a computational literacy skill. If the researchers were not computational literate, they would not have been able to improve their study, or their hypothesis.

According to NGSS, scientists use models and representations to explain, and defend, findings or design solutions, which is what the researchers in the Nersessian paper had to do throughout and after their experiment. DiSessa argues students can use computational models to create and organize information for themselves to see and make sense of concepts and that the computational models can act as an explanation for concepts. The researchers in the case study used the models they created for their data to explain, and defend, their findings to the scientific community. Nersessian and diSessa want students to be science and computational literate, so they are able to find understanding and create their own ideas.