technology to support complex learning

Editor's Notes

• #8: Will now present some examples in the context of my work– will talk about technology
• #9: ----- Meeting Notes (9/5/12 17:17) -----STELLAR was an attempt to support problem-based learning in teacher education through the uses of technology.
• #17: - analyze only steps 3 to 6 of the activity for each group, the line counts begin at different numbers for each group because Group 2 engaged more with the STELLAR tools in the earlier phases of the activity than did Group 1.The vertical axis shows the categories of tool hits, discourse codes, and speakers. The horizontal axis shows the number of tool-related events, either a log entry or a discourse turn. The bottom seven categories represent tool hits by any member of the group. The top six or seven categories represent the speakers.Group 1- weaker group, in Group 1, the facilitators were involved early and fairly frequently, and asked most of the explanatory and metacognitive questions. Ann and Fauna seemed to dominate the discourse though other students contributed. Task talk continued throughout duration of activity as they struggled to figure out what they were doingAnother aspect of how the CORDTRA diagrams help us distinguish the collaborative activity between two groups is by showing the overall relation between the discourse and the tool use. In Group 1, the students initially viewed the video and the Knowledge Web but after about line 650, none of the group members used these two resources until the very end of the collaborative phase at line 1000. The content of their online postings were intermixed conceptual, social, task and tool-related talk throughout their work on the problem. There was some discussion of tools as a problem midway through the discussion. ----- Meeting Notes (9/5/12 18:03) -----To better understand the variability, we created visual representation in which we examined the relation between discourse features and log data from using the various tools.CORDTRA-- builds on work of Rose Luckin-- Chronologically ordered representation of discourse and tool related activity all on a single timeline.
• #18: Group 2In Group 2, the facilitators joined in much later and made infrequent contributions in the form of questions. This is because in this group the students themselves asked a number of questions throughout the duration of the problem. The group participated fairly evenly except for Matt who chimed in late in the group’s work. His contribution built on one of the other student’s ideas and was grounded in personal experience. His later contributions offered an important new idea for an activity that was grounded in psychological theory. What is interesting in Group 2 is that they went back to the video and the Knowledge Web at intervals throughout the discussion (e.g., around lines 850-925, 1050-1100, 1275). It appears that group members did this following several explanation questions as Figure 7 demonstrates (for example at about line 1050). This suggests that Group 2 was using the resources of the video and the Knowledge Web in a purposeful manner. That is, they seemed to be using the STELLAR resources to answer the questions at hand. They were bringing together the problems of practice and conceptual ideas repeatedly. We hypothesize that it is the repeated meshing of conceptual ideas from the Knowledge Web with the perceptual ideas from the video, which is a precondition for transfer (see Derry, 2006). Our current analysis, however, did not test to see whether transfer in fact occurs.The CORDTRA demonstrates several other distinctions between the two groups with regards to timing. Group 2 engaged in group monitoring throughout the activity whereas Group 1 did not engage in this kind of monitoring until relatively late in the activity. This was generally a request for feedback from the rest of the group. Inspection of the data shows that it was the facilitator doing most of the group monitoring in Group 1 whereas in Group 2, it was the students who were the ones monitoring themselves. The facilitators adapted their support to the needs of the group and thus differentially engaged with each group as needed. ----- Meeting Notes (9/5/12 18:03) -----There are other distinctions I could mention, but from a methodological perspective, this allowed us to see how the different groups engaged in their problem solving task.
• #19: Participants viewed a brief video in which high school students learned about electricity, electrical circuits, and how a light bulb works. Before viewing the video, they received a brief written explanation describing how the video clip illustrated a problem the teacher had: that even top students were maintaining their pre-course misconceptions about electricity after instruction. The video explained how the teacher had spent a month covering advanced topics in electricity and provided hands-on experience designed to reinforce those concepts and illustrate how electricity enabled a light bulb to work. Video also showed an interview with a good student before and after instruction and demonstrated that she maintained the same misconceptions following instruction.They had thirty minutes to answer the following four questions: (1) How do you know that the student failed to learn? (2) Why did the student fail to learn? (3) What recommendations would you make to the teacher to help him improve his teaching? and (4) What else do you need to know to better understand the teaching-learning situation? What additional questions would you ask?
• #22: Group A- difficultydemonstrated by the need for the teaching assistant to frequently intervene and facilitate this group’s workGroup B- Both groups engaged with the concept of transfer and they showed this in how they used the technology
• #23: All students discussed in P1, number decreased in subsequent problems. All the students in this group used transfer in their independent proposals but did not use it as an idea in commenting on other proposals, suggesting that the members of this group used the concept in parallel without real knowledge building Jenny quote 1: This excerpt shows that the student brought in ideas about transfer through direct quotes that define the concept. This is much like knowledge telling (Bereiter & Scardamalia, 1987), as Jenny shared relatively unprocessed information with her group. Although this group continued to use the concept in independent proposals, it later became encapsulated as part of the group’s language and no longer required detailed explanation.
• #24: illustrates a more fluid use of the concept of transfer without needing to provide all the definitions, suggesting that the group members had achieved a shared understanding. This use of conceptual ideas was seen to an even greater degree in Group B.
• #25: , similar to Group A, the group members began by incorporating direct quotes as they shared information, as this excerpt from Kathy shows. But unlike the other group– she more carefully mapped the definition to the evidence in the video.
• #30: Preserves video and metadata by converting to digital formatOffers a research and teaching portalProvides analytic tools, customized ontologies, and personal spaceSupports individual work and group collaboration
• #31: Different methods help deal with the complexity of theoretical perspectives
• #32: Same pre and post tests as in other enactments– but here our research questions focused on how people used the VMC resources
• #37: Posted Assignment This week's assignment for online work involves a video and two readings, with threaded discussion, that follows class work on problem solving for the Ankur's Challenge task.  The following questions are intended to guide discussion in your small groups (and will also be posted in the introduction to group discussion threads).
• #38: ACROSS GROUPS, THE PERCENT OF POSTED DISCUSSIONS RELATING VIDEOS TO OWN AND OTHERS’ PROBLEM SOLVING WAS STATISTICALLY SIGNIFICANT.THE PERCENT OF POSTED DISCUSSIONS ABOUT ENJOYMENT ACROSS GROUPS WAS STATISTICALLY SIGNIFICANT.GROUPS A & D MADE LESS REFERENCE TO EARLY INTERVENTIONS AND TO PRACTICE .
• #39: TABLE 2 SHOWS THAT FOR GROUP C, THE PERCENTAGE OF SS WHO RELATED VIDEOS TO ASSIGNED READINGS WAS STATISTICALLY SIGNIFICANT,
• #40: Answered this with qualitative analysis
• #42: Purposes include:purpose (e.g., research, professional development)We also wanted, for classes, to better understand design features
• #44: Part of goal was to get sense of terrain.
• #45: Used this to look at high and lesser quality analytics to help future instructors better design tasks and evaluate analytics. Help us in discerning additional criteria that distinguished high and low quality
• #46: This was derived from researcher-supplied themes– will be interesting to apply to the raw data.
• #48: Talk a little bit about move
• #49: Embedded in problem- design aquarium ,video of eutrophication of pond,
• #50: Describe the screen – variables, controls, graph, outputs, fish… controls: sunlight, nutrient runoffDescribe the screen and the purpose of this simulation. At this point they have already identified inverse relationships between O2 and CO2. That when the O2 is down, more fish die. The teacher has just reminded the whole class that the goal of the activity is to model how the fish died in the pond—to reproduce the conditions in the pond that resulted in the fish dying. They have been modeling a system that keeps the fish alive or to die of old age. Why is the O2 decreasing?----- [NEXT SLIDE]
• #51: Move from SBF CMP– need to explain this.
• #52: Note that a great deal of this work is in progress.
• #54: Typically, we look at an aggregate pre to post test gain, but in complex domains, this is not sufficient- need to know how students learn– where they start, which aspects of systems are harder than others.Question 1 we are beginning to gain traction on– analysis is ongoing– question 2 is on the burner/Microgenetic analysis
• #55: Coded from Drawings taken over 4 time points in a year--
• #56: To support our analysis, we constructed a graph) that depicts the group means for each systems dimension in Classroom 1. Students appear to have steep growth towards higher levels of Biotic/Abiotic and macro/micro dimensions but relatively more steady growth towards higher levels of SBF and Coherence dimensions that converge by posttest. From the start, students represented structures in terms of their behaviors or functions and made connections between phenomena (Coherence). However, these structures were primarily macroscopic and biotic. Students made more complex connections between Biotic/Abiotic structures (AA1) and between Macro/Micro structures (AA2) before reaching higher levels of SBF or Coherence. Finally, the inclusion of extraneous structures peaked at AA1 and then declined until no drawings included extraneous structures by posttest.Of course, means mask some of the variability but we are seeing some consistency in B:A preceding M;M in general and that making both of these are important for a coherent understanding– represented by both the coherence measure and sBF score.  
• #60: We designed the interview to assess students’ generalization of mechanistic reasoning and to make sense of new problems related to aquatic ecosystems. In the first problem, students were shown a paper copy of their group-generated EMT model depicting factors that may have led to fish dying suddenly in a local pond. The students were then asked to label the model in terms of CMP and explain their reasoning. In the model (see Figure 1), the entire problem reflected the phenomena, the rectangular boxes represented components that were linked together by explanations of their mechanistic behavior. In the second problem, students were told that there has been a sudden increase in geese population around a lake that has resulted in changes to the aquatic ecosystem. They were shown three versions of EMT models (the first consisting only of components, the second had only mechanisms but no components, and the third consisted of numerous components, mechanisms connecting them and phenomena). Students were first asked to rank each model on a scale of one to three, with three being the most complete explanation about what happened to the lake ecosystem as a result of the overpopulation of geese. Next, students labeled the model they had ranked the highest and were asked to explain the criteria for making their selection.
• #61: To trace generalization of mechanistic reasoning from an AOT lens, we compared each student’s labeled model in task 1 with that in task 2. We identified items they labeled as identical in terms of CMP in both tasks and also kept track of areas where they exhibited differences (i.e., identified phenomena in one task as the entire model and in the other had it labeled as a mechanism).
• #62: As we compared students’ labeling of models, it was evident that a majority of students generalized the concept of components and phenomena. We observed that students would either circle the entire model or focus on the primary problem under investigation when asked to label phenomena. Additionally, almost all the students were quick to tag components as factors that may have led to the problem
• #63: Social coordinationBehavioral engagement refers to the degree of the group’s on-task behavior.Task engagement refers to the focus of engagement on efficient planning and what steps to take next to accomplish the task. Can also refer to a focus on manipulating the technology tool (but the rating does not reflect whether concepts or content represented).
• #64: Conceptual engagement is considered a continuum that ranges from content connections that are focused on the key question or task problem or relating to the real world/experiences to simple knowledge telling.  Note: The conceptual and consequential connections do not need to be accurate.To warrant a high rating, students need to go beyond referencing the larger question of why do fish die, for example, to making a connection with evidence or content.  Also, a connection to a simulation or piece of evidence is still a medium rating, unless connected to the larger question.Medium: Group discusses and identifies content relationships between visible and invisible components.Low: simply reading the hypermedia; only interpreting simulation data without content relation (nutrients go up).We are seeing trends that the simulations foster high conceptual consquential engagement, task engagement, and behavioral engagement– more variability with EMT modeling tool.
• #65: Technology can play important role, but is synergistic with creating engaging contexts that create need for knowledge, bringing hard and soft scaffolds together– thinking about how to distribute scaffolding– also need to examine learner activity