RAA2: Adaptive Agents Monitor and Increase Students’ Engagement

Szafir, D., & Mutlu, B. (2012). Pay attention!: designing adaptive agents that monitor and improve user engagement. Proceedings of the SIGCHI conference on Human factors in computing systems, CHI  ’12 (pp. 11–20). New York, NY, USA: ACM. doi:10.1145/2207676.2207679
Purpose of the Research:
The authors of this paper set to answer the question: How to design computer-based educational tools that monitor student attention and employ attention-inducing strategies to improve learning in the way human teachers do?
The authors design a human like social robot as an instructor. This robot can monitor students’ attention level by detecting their unconscious brain signals. They use a wireless electroencephalography (EEG) headset to detect the students’ brain signals. This method is often used in Brain-Computer Interfaces (BCI). When detecting a drop in engagement level, the robot then will recapture the students’ attention level using verbal or nonverbal cues (immediacy cues).
The authors then propose two hypotheses: (1) The system can increase students’ attention, and therefor increase learning performance (specific to this experiment, increase students’ recall of narrative stories); (2) The system can increase students’ motivation and rapport to the robotic instructor. The first hypothese is developed based on the arousal-attention theory, and the second is developed according to motivational theory.
The authors then test the two hypotheses with 30 college students (15 male, 15 female).
Main Findings: 
Experiments with the 30 college students suggest that the first hypothesis is fully supported, while for the second hypothesis, only female students demonstrate increased motivation and rapport to the robot. The authors then discuss the possible reasons for the result, for example, the robot is a small stature with a child-like voice, which may have made it more difficult for males to connect with the robot than females.
(1) This paper has nice review of educational psychology theories related to students emotion, attention, and motivation, and shows how these theories can be used to improve computer based education systems. These are useful information for my own research in engineering education.
(2) The paper explains the difference between active Brain-Computer Interfaces (BCI) and passive BCI. In active BCI, the users have to consciously control their brain activities in order to control the computers. This usually requires extensive tuning and training on both the parts of the system and the users, thus is difficult to become generalizable. In passive BCI, the systems detect users’ conscious and unconscious brain signals, and combine these with other user inputs such as voice and gestures to determine the actions. I am thus seeing the future of user interface a combination of BCI and NUI (natural user interface), where the users use both brain activities and natural body movements, tactile, and verbal actions to interact with the computers.

Reading reflection: Attention

I will mention why I put this video here at the end of this post.

I briefly summarize a few papers and add some of my thoughts following each two papers.

01. Is Google Making Us Stupid? What the Internet is doing to our Brain,  by Nicholas Carr

Claim (argument): The Internet is changing our way of processing information to become shallow and make us less capable for concentration and contemplation.

Evidence: Start with personal experience of not being able to concentrate on reading and writing–>Expand to friends, acquaintances (literacy types)–>Bloggers he followed–>Scholarly studies that the web is changing the way we read and think ( simple, clear and well-structured way of generalizing and developing the argument)

02. Distracted: The Erosion of Attention and the Coming Dark Age, by Maggie Jackson

Claim (argument): Modern technologies is eroding our attention, thus the ability of deep, sustained focus and analytical reasoning as a society. We are facing a real risk of social decline if we do not nurture our attention.

Evidence: parallel to history; ADD; busyness, multitasking; empirical studies about kids multitasking, low patience, and lack of analytical reasoning ability; Power Point; Theories about attention and it’s importance, etc.

My thought on the topic of the two papers above:

I think maybe we can make an analogy of the society growing to a human child growing. My mum always tells me how she was astonished about and even “admirable” to me about how much I could concentrate on things when I was around 1-3 years old. When I started to learn how to use  scissors, one night, I was using the scissors to cut papers and I refused to go to bed, when my mum woke up from her sleep, I was still cutting papers maybe already for 2-3 hours. I was so concentrated and forget about the time, or I think I didn’t really have the concept of time, all things in my world was scissors and papers.  Another time, I would sit on the floor and play lego (the Chinese version of similar building block toy as lego) for 2 hours nonstop.  However, as I grow up, I start to be aware of time, I start to need to process more and more information, and my world is not that pure anymore. I feel I could never go back to that kind of mental status anymore even when I try my best to concentrate. When I spend too much time tackling some hard problems I get panic about time, because I have so many other things to do, so I start to look for some easier solutions, then I lose the spirit of never giving up, drilling on the hard part, and deeply polishing my analytical reasoning ability. This is almost inevitable as a part of growing up. Most adults’ worlds are much more complicated than children’s. We have to consciously process lots of information, and we need to worry about lots of things. The growing process is also the process that we go out of the ego stage (Piaget developmental theories) and start to be aware of the outside world, the cost is that we are inevitable exposed to more information, and maybe distracted by this information.

As a society, we are growing like a human child. As we growing, we inevitably need more information and communication with others. Modern technologies esp. the web went a long way to promote this growing by allowing easy information diffusion and communication among people, so we become aware of way more things than before. Also, inevitably, we get distracted by these things, but this is part of growing up. We are getting more mature, we need to learn how to deal with this and cultivate our lost attention.

If this analogy is appropriate, is the society going to die some day just like a human being is going to die eventually? Do modern technologies somehow speed the dying of the society at the same time it speeds the growing of the society? How old is the society now? Is death the dark age Jackson predicted?

05. Self-interruption on the Computer, by Jing Jin and Laura Dabbish

Research goal: Why do people interrupt themselves on the computer and switch to doing something else? What do internally generated interruptions look like in practice? What are their potential negative and positive side-effects?

Methods: Shadowing observations of 13 participants doing their normal work tasks on computers followed by 30min to 1 hour retrospective interviews; grounded theory; theoretical sampling to only analyze the internal interruption data; verified using independent coders

Main findings: A typology of self-interruption on computer. Seven types of self-interruptions were identified: adjustment, break, inquiry, recollection, routine, trigger, and wait.

06. Multitasking and Monotasking, by Dario Salvucci and Peter Bogunovich

Research goal: To test a claim that when users are alerted to interruptions at points of higher mental workload, they delay processing of the interruption until they have reached a point of lower metal workload.

Methods: 20 users participated the study; Participants have to do a mail task on customer support, while answering interrupting chat messages. The chat prompt was generated at a pseudo-random point: the system tracked the user’s events, after one of eight different events, triggered a chat prompt 50-200ms after the event. Users can choose to defer the chat. System recorded all the transitions and researchers analyze the transitions between events.

Main findings: Verified the claim. The experiment also helps to clarify one source of mental workload, namely the problem state—temporary information needed for task processing.

My thought on the topic of the two papers above:

The second paper above mainly talks about external interruptions, but it also says in the discussion section that “we would suspect (though further research would need to be confirm) that this ability also generalizes to user self-interruptions and discretionary multitasking”. I highly doubt this assumption/hypothesis. Users maybe capable to deal with external interruptions when they are inconsistent with the users working mode, however, internal interruptions are more difficult to control.

07. The Laptop and the Lecture: the Effects of Multitasking in Learning Environments, by Helene Hembrooke and Geri Gay

Research goal: To test two hypothese:

H1: Students in the open laptop condition would perform significantly poorer on immediate measures of memory for the lecture material.

H2: The memory decrement observed would not result from the relatedness of the content viewed in the secondary task (laptop use) to the primary task (lecture information). In other words, content relevance would not contribute significantly to the variance observed in the main effect above.

Methods: Two groups of students (22 in each group). One group was encouraged to use laptop during the lecture, while the other group was not allowed to use laptop. All the students were tested about the content of the lecture immediately after the lecture. ANOVA analysis.   

Main findings: The two hypotheses were proved, but this is only for immediate test of memory after the lecture. For the overall performance, the average score of using laptops in class is B+, which is very good. This is largely because the structure of the class was nontraditional, highly interactive and dynamic. Had the class been more traditional and grades determined by conventional test of memory the outcome of students who use laptops may have been different.

09. Caught in the Web: university student use of Web resources, by Yu-Mei Wang and Marge Artero

Research goal: 1. How do students use web resources (academic vs. non-academic); 2. What are the students online search behaviors? 3. What are the students’ perceptions of web resources? 4. Do students evaluate information on the web? 5. What types of training do students perceive they need to successfully utilize web resources?

Methods:700 questionnaires about information literacy skills. SPSS.

Main findings: the findings of this study show that there is an urgent need for students to develop information literacy skills and apply these skills in the electronic information environment.

My thought on the topic of the two papers above:

Whether or not to use internet in classroom has been a huge debt. Some think technology will fail education, as it can hugely distract students. Most say technology will save education. The question needs to be discussed is not whether or not to use, but how to use. How to guide the students to properly use, and how to change the structure of the class and the assessment methods to be more supportive for using web in class. There is a digital ethnography video provoking lots of thoughts on this. I loved this video a lot. I put this video at the beginning of this post to call for your “attention”.