Break Through Tech Branding Assistant - Proscribed Answers

Break Through Tech Brand Assistant - Open Answers

Educational data mining/Learning Analytics and its relation to intelligent computer-assisted instruction

-
If the authors from the Corbett & Anderson’s article read the paper on Explanatory Learner Models back in 1995 when their article was originally published, do you think this may have influenced their research (why or why not)? 

In any industry, 24 years is an exceptionally long period of time. In technology, 24 years is practically another era. For this and other reasons, I do believe that had Corbett & Anderson (1995) had access to the article on explanatory learner models by Rosé & McLaughlin (2019), their research and subsequent article may have trended in a very different direction. 

In their article, Corbett & Anderson (1995) focused on a model that described "students' changing knowledge state during skill acquisition" (p. 253). It is worth noting that at the time, Corbett & Anderson (1995) were also iterating and expanding on the status quo idea of "mastery learning" and were attempting to "bring a cognitive model of skill acquisition to bear" (p.254). They found that their model was successful at predicting test performance and was able to "enabl[e] most students to reach a high level of task performance" (Corbett & Anderson, 1995, p. 276). One can see that in 1995, research was mostly focused on the predictive power of models for learning, without much thought being given to "why" or "how" learning was taking place. 

Jump ahead 24 years and the field has shifted. Rosé & McLaughlin (2019) present a "view of learning analytics" called explanatory learner models, "the goal of which is to enable insight-driven use of such analytics in technology-enhanced education" (p. 2943). Specifically, they define an "explanatory learning model as a model for which insight about learners, the learning process, or the instructional context can be derived from interpretation of the structure of parameter estimates of learned models" (Rosé & McLaughlin, 2019, p. 2944).

If influenced, what changes may you see in Corbett & Anderson’s research approach (e.g., perspectives, agenda, methods) moving forward?

Some of the main takeaways from the Rosé & McLaughlin (2019) article were on the importance of transferability, the extent to which the model can be employed in an application. If Corbett & Anderson (1995) had access to Rosé & McLaughlin's (2019) research before they began their own study, they may have intentionally added another test condition that examined how the model performed in diverse environments. Specifically, Corbett & Anderson (1995) may have tried to employ a development methodology that allows for redesign of the algorithm and accounts for "model and application error" within "authethentic use cases" (Rosé & McLaughlin, 2019, p. 2955).

Additionally, Rosé & McLaughlin (2019) "suggest that learning engineering teams should put more emphasis on the interpretability and actionability of educational data mining and learning analytic efforts, to produce more explanatory models" (p. 2955). By trying to understand why the model "achieves better predictive accuracy than alternatives," Corbett & Anderson (1995) could have gathered more "insight into how students succeed or struggle to learn relevant material" (Rosé & McLaughlin, 2019, p. 2955). This type of data loop could then be used to redesign the course materials and practices and assess subsequent student learning outcomes. 

References

Corbett, A. T., & Anderson, J. R. (1995). Knowledge tracing: Modeling the acquisition of procedural 

    knowledge. User modeling and user-adapted interaction, 4(4), 253–278. 

Rosé, C. P., McLaughlin, E. A., Liu, R., & Koedinger, K. R. (2019). Explanatory learner models: Why 

    machine learning (alone) is not the answer. British Journal of Educational Technology, 50(6), 2943-

    2958. 

Comments

  1. Sandra OkitaNovember 2, 2020 at 3:35 AM

    Great reflection. You make a great point of how in 1995, not much thought was given to the "why" or "how" learning was taking place. I appreciate how you clearly state the takeaway points from the Rosé & McLaughlin article that are most relevant to the Corbett & Anderson article, and how explanatory models may have changed their research approach.

    ReplyDelete

Post a Comment

Popular posts from this blog

Final Project Brainstorming

-
1) Out of all the different pedagogical agents and intelligent computer-assisted instructional systems introduced in this week and past readings, which agent/system did you find most interesting and why?  Although not really a pedagogical agent, ELIZA is a computer program that really stands out to me. I especially liked being able to try out ELIZA myself to get a sense of how it apply Rogerian theory to build its responses. As someone with a background in mental health, I enjoyed how the creators used the reflection, validation, and summary-based structure of Carl Roger's client-centered therapy to inform ELIZA's responses.  One of the pedagogical agents that really stood out to me was the teachable agent  Betty's Brain . I like the idea of students learning by teaching an agent. I think that idea is based on solid pedagogical theory and is a idea that many teachers employ even without using agents, for example when students are tasked with tutoring other individuals. I...
Read more

Human Modeling Approaches vs the Learner-centered Approach

-
1) How are the two agents similar and different in their a) design of the agent and b) interaction with the learner? Betty's Brain The teachable agent examined by Schwartz et al. (2009), Betty's Brain, allows students to apply interactive metacognition, where the act of teaching Betty helps learners strengthen their own metacognitive skills and results in better learning outcomes.  Betty's Brain was "designed for knowledge domains where qualitative causal chains are a useful structural abstraction (e.g., the life sciences) (Schwartz et al., 2009, p. 5). When interacting with Betty, students teach her by "creating a concept map of nodes connected by qualitative causal links" (Schwartz et al., 2009, p. 5). Betty answers and asks questions based on her understanding of the concepts. She can also animate her reasoning process, which enables users to track her reasoning and "remediate" her knowledge if necessary. (Schwartz et al., 2009, p. 5). Features...
Read more

Animated Instruction and Social Metaphors

-
Image
Artifacts at Various Stages of the Uncanny Valley Curve For each of these either pedagogical agents or intelligent computer-assisted instructional systems, I will endeavor to answer the following questions: How the student will interact with the agent/system (and why)? What the agent/system may be capable of when it comes to evaluating/assessing student's learning (and why)?  For what kind of learning would such agent/system be good for (and why)? Not being familiar with the majority of these agents before this exercise, I'll do my best to answer the questions based on my first impressions.  Hal Since Hal does not resemble a human physically, but does seem to have speech capabilities, I would recommend it for tasks that don't require a high level of empathy with the agent. Although there are examples/exceptions to the rule that a human form will result in a higher level of trust, such as in the film Her (2013), I believe that Hal will mostly be useful for thi...
Read more