Thoughts on Contextual Recommender Systems + Personalized Learning (Part 2): Affect on Learning

Continuing off of my previous post introducing contextual recommender systems for technology enhanced learning (TEL), I am interested in learning about various types of user contexts. Verbert et al. categorize these into the following:

• Basic personal information, which usually includes information such as a student’s name, contact information, affiliations, language capabilities, gender, age, education level, etc…;
• Knowledge, which can represent (based on existing works) the “prior knowledge levels of the learner” or the student’s current performance;
• Interests, which are basically a student’s interests or preferences, and can include a user’s search terms, tags, comments, and resources created, read, or rated;
• Learning goals can include both short-term goals, such as trying to solve a specific problem, and long-term goals, such as goals relating to a course or future plans;
• Learning and cognitive styles define a student’s “preferred way of learning presentation and cognitive processing,” for example, visual, textual, or auditory presentation;
• Affects, or emotions, can also be modeled, with research oftentimes referring to Russell’s “two-dimension ‘circumplex model of affect,’ where emotions are seen as combinations of arousal and valence”; and
• Background, which refers to a student’s previous experiences “outside the core domain of a specific system,” such as work experiences, religion, and culture [1].

For now, I will dig deeper into affect. Verbert et al. have stated that “in TEL, research on the influence of emotions on learning has also gained major interested in recent years,” so I am very curious to see what works already exists.

But first, how does affect affect learning?

George Mandler, in Affect and Learning: Causes and Consequences of Emotional Interactions (1989), has stated that “affect is the least investigated aspect of human problem solving, yet it is probably the most often mentioned as deserving further investigation.” He then states that contemporary cognitive theory would illustrate human beings as “passionless creatures” who can “think and act rationally and coolly” [2]. But that is not precisely the case. 

“In reality… human beings typically are frustrated, angry, joyous, delighted, intense, anxious, elated, and even fearful when dealing with complex problems… If cognitive psychology aspires to an understanding of human thought and action, it can ill afford to leave out their emotional aspects.”

– George Mandler, in Affect and Learning: Causes and Consequences of Emotional Interactions (1989)

We are often and easily driven by emotion. Emotion is “not only anecdotally and phenomenally part of human thought and action; there is now a burgeoning body of evidence that emotional states interact in important ways with traditional cognitive functions” [2]. For example, there is work showing how positive feelings can affect decision-making and influence problem-solving strategies [2]. And there is a relationship between emotions and learning [3].

“The extend to which emotional upsets can interfere with mental life is no news to teachers. Students who are anxious, angry, or depressed don’t learn; people who are caught in these states do not take in information efficiently or deal with it well.”

– Daniel Goleman, Emotional Intelligence

Kort et al. discuss the natural process of learning, which involves failure and “a host of associated affective responses,” and how educators of science, math, engineering, and technology (SMET) rarely provide this experience to students [3]. They state that when materials are presented to students, they are usually in a “polished form” that lacks these natural steps of “making mistakes” (and thus, feeling confused), recovering from those mistakes (and so, overcoming frustration), “deconstructing what went wrong” (e.g., “not becoming dispirited), and trying again or starting over (“with hope and enthusiasm”) [3]. Since students don’t often get this experience, when they do finally have to experience failure, it is likely for them to immediately jump to thoughts of “not being good enough” or that they simply “can’t do it” [3].

A four-quadrant learning spiral model has been proposed by Kort et al. in which “emotions change while the learner moves through quadrants and up the spiral” [4,5]. Here, the learning process is broken up into two axes labeled “learning” and “affect.” The horizontal learning axis ranges from “constructive learning” at the top to “un-learning” at the bottom. Then, the vertical affect axis runs from “negative” to “positive” affect from left to right. Various affects are then illustrated in different quadrants. For example, one may feel “puzzlement” or “confusion” as a negative emotion, yet this falls under constructive learning. This model relates phases of learning to emotions and is illustrated in Kort et al.’s work [4]. (It is interesting – go look!)

We must realize, then, that this cycle of learning is essential and that a student may feel many emotions throughout this process. The negative emotions aren’t necessarily bad, and “it is not simply the case that the positive emotions are the good ones” [4]. We should not try to keep the student in a single, positive quadrant. Instead, we should try to help them realize that this is a natural cycle of learning – that it is vital to be able to “keep orbiting the loop” and to know “how to propel themselves especially after a setback” [4].

The authors of this model state that it is “not intended to explain how learning works”, but is rather a beginning step intended to raise questions about “the role of emotions in learning” [4].

“We have only begun to explore what are the appropriate scaffolds for promoting learning. We have also much to learn on how computational and communication technologies can support teacher collaboration and professional development.”

– Eliot Soloway, Scaffolding Technology Tools to Promote Teaching and Learning in Science; quoted in [4]

Next, I will look more into what intelligent systems exist which take into account the learners’ affective states and how.

---

References

1. Verbert, Katrien, et al. “Context-aware recommender systems for learning: a survey and future challenges.” IEEE Transactions on Learning Technologies 5.4 (2012): 318-335.
2. Mandler, George. “Affect and learning: Causes and consequences of emotional interactions.” Affect and mathematical problem solving. Springer, New York, NY, 1989. 3-19.
3. Kort, Barry, Rob Reilly, and Rosalind W. Picard. “An affective model of interplay between emotions and learning: Reengineering educational pedagogy-building a learning companion.” Proceedings IEEE International Conference on Advanced Learning Technologies. IEEE, 2001.
4. Kort, Barry, Rob Reilly, and Rosalind W. Picard. “External representation of learning process and domain knowledge: Affective state as a determinate of its structure and function.” Workshop on Artificial Intelligence in Education (AI-ED 2001), San Antonio,(May 2001). 2001.
5. Craig, Scotty, et al. “Affect and learning: an exploratory look into the role of affect in learning with AutoTutor.” Journal of educational media 29.3 (2004): 241-250.