Inferring student attention with ASQ

Vasileios Triglianos, Cesare Pautasso, Alessandro Bozzon and Claudia Hauff

@asqjs

Computers are also good for programming

Computers are also good for typing answers

Computers are also good for drawing diagrams

Computers are also good for interactive questions

Heard you study the Web?

I made you a Web-based tool to learn about the Web, while practicing the Web without leaving the Web

nswers

lides

uestions

Turn devices from

sources of Distraction

to

affordances of Learning

Agenda

ASQ

- Motivation and Goals
- Events

Inferring Student Attention

- Current methods for measuring student attention
- Findings for student attention
- From events to higher level attention states
- Results

Goals

No Context Switch

Everyone answers

Q1. Do you teach any course involving programming?

Yes

No

Q2. What's your favourite programming language?

Q2. Favourite programming languages

Browser events

Event Name	Description
`windowfocus`	The browser window that displays the ASQ receives focus.
`questioninput`	Student input in a question.
`exercisesubmit`	A student submits the solution to an ASQ exercise.
`idle`	Student is inactive for 10 seconds.
`connected`	A student connects to the ASQ server.

tabhidden, tabvisible, windowblur, exercisefocus, exerciseblur, input, answersubmitted, disconnected

RQ1. To what extent can students’ attention be inferred from their interactions with a Web-based platform?

RQ2. Which type of interactions are most correlated with (in)attention?

Method	Invasive	Accurate	Privacy
EEG signals	✓	✓	✓
Facial expressions		✓
Eye gaze and skin conductance	✓	✓	✓
TAs Observing Students		✓
Note-taking analysis

State of the Art Findings

F1: Students' attention drops over the class period.
F2: Attention breaks occur regularly and increase in frequency as the class progresses.
F3: Attention rises when interactive elements are introduced.
F4: Immediately after interactive teaching elements, the level of distraction is lower than before the start of the interaction.

Web & Database Technology

- 2015/16 edition
- 1^st - 3^rd year Bachelor students
- 3x90min lecture sessions
- 111-187 students

Browser events for 2 students

From low-level events to attention states

State indicators

Name	Description
`exercise`	The current slide has an exercise.
`connected`	The student browser is connected to ASQ.
`focus`	The browser has focus on the tab or exercise related to the lecture.
`idle`	The viewer is idle.
`input`	An ASQ exercise HTML element receives input.
`submitted`	All the ASQ exercises of the slide have been submitted.

Connected-focus indicators for the whole class

F1: Students' attention drops over the class period.
F3: Attention rises when interactive elements are introduced.

connected-focus indicators for ER Conceptual Design lecture

connected-focus indicators for Web Security lecture

exercise	connected	focus	idle	input	submitted	Inferred Attention State
-	-	-	-	-	-	Disconnected
✔	-	-	-	-	-	Disconnected
✔	-	-	-	-	✔	Disconnected
-	✔	-	-	-	-	Distracted
-	✔	-	✔	-	-	Distracted
✔	✔	-	-	-	-	Searching for a solution
✔	✔	-	✔	-	-	Searching for a solution
-	✔	✔	-	-	-	Interacting with non-question slide
-	✔	✔	✔	-	-	Following
✔	✔	✔	-	-	-	Thinking
✔	✔	✔	✔	-	-	Thinking
✔	✔	-	-	-	✔	Bored
✔	✔	-	✔	-	✔	Bored
✔	✔	✔	-	-	✔	Waiting
✔	✔	✔	✔	-	✔	Waiting
✔	✔	✔	-	✔	-	Working
✔	✔	✔	-	✔	✔	Reworking answer

Attention State for 2 students

Attention State for the whole class

F1: Students' attention drops over the class period.
F3: Attention rises when interactive elements are introduced.
F4: Immediately after interactive teaching elements, the level of distraction is lower than before the start of the interaction.

Let's zoom in a question slide

Attention-focus variation

F2: Attention breaks occur regularly but do not increase in frequency as the class progresses.
F3: Attention rises when interactive elements are introduced.

Future work

Map ASQ-inferred state ('distracted') to actual state ('looking at the beamer', or 'taking notes')
Reduce 'bored' state
- Time to move on
- Further engage with self/peer assessment

Broadcast your slides and get Live feedback from your audience
Engage your audience by Embedding Questions in your HTML/PDF Slides
Scale to 100s of Students
Questions are plugins. You can write your own!
Turn Students’ Smartphones from Distractions into Learning Tools

http://asq.inf.usi.ch

We want to collaborate with you!

Vasileios Triglianos, Cesare Pautasso, Zhenfei Nie, Jacques Dafflon, Margarita Grinvald, Max von Bülow, Georgios Kokosioulis, Luigi Frunzio, Marco Ravazzini