Gaze-Aware Quiz Interfaces for Better Retention

ISEF Category: Systems Software

Subcategory: Human/Machine Interface  ·  Difficulty: Intermediate  ·  Setup: School Lab  ·  Time: 1 to 2 Months

The Hook

Students often click an answer without really looking at the question. A gaze-aware quiz tries to catch that by watching where the eyes go. If your attention drifts, the system can ask again. That makes this project a real test of how much eye tracking can improve learning.

What Is It?

This project studies a quiz interface that uses webcam-based gaze tracking to estimate where you are looking. WebGazer.js is one example of this kind of tool. Instead of waiting for a wrong answer, the system can notice that your gaze leaves the answer area and re-prompt the question. Think of it like a teacher who says, “Look back here,” when you start zoning out.

The core idea mixes human-computer interaction with learning science. You are not just asking whether students get answers right. You are testing whether attention-aware software changes what students remember later. That makes the project about both interface design and memory, which is a strong combination for a science fair study.

Why This Is a Good Topic

This is a good science fair topic because you can test a clear design change and measure a clear outcome. The independent variable is the interface behavior, and the dependent variable is retention or answer accuracy. You can run it with simple web pages, a laptop webcam, and a small group of users. The topic connects to online learning, attention, and adaptive software, which makes the results easy to explain and easy to care about.

Research Questions

• How does a gaze-aware re-prompt affect immediate quiz accuracy compared with a standard quiz interface?
• What is the effect of gaze-triggered reminders on delayed retention after one day or one week?
• Does a gaze-aware interface reduce skipped questions or fast guessing on online quizzes?
• To what extent does answer-region fixation time predict later recall on Khan-Academy-style content?
• Which re-prompt threshold gives the best balance between catching inattention and avoiding false alerts?
• How does feedback style, such as a reminder versus a full question repeat, affect quiz performance?

Basic Materials

• Laptop or desktop computer with a webcam.
• Web browser that supports JavaScript.
• WebGazer.js or a similar open-source gaze tracking library.
• Simple web editor such as VS Code or an online HTML editor.
• Spreadsheet software for score tracking and data cleaning.
• Consent form for student participants.
• Timer or survey form for delayed recall testing.
• Quiet room with stable lighting.

Advanced Materials

• High-resolution external webcam.
• Second monitor for presenting quiz items separately from the tracking display.
• Local server setup for running the interface and logging events.
• JavaScript framework or plain HTML, CSS, and JavaScript project files.
• Python for statistical analysis and plotting.
• ImageJ or similar tool for checking screen capture regions, if needed.
• Ethics and consent documents approved by a school supervisor.

Software & Tools

• WebGazer.js: Tracks approximate gaze location from a webcam for an attention-aware quiz prototype.
• VS Code: Lets you build and edit the web interface in JavaScript and HTML.
• Google Sheets: Organizes participant scores, response times, and delayed recall data.
• Python: Helps you run statistical tests and make plots from the quiz results.
• JASP: Provides free statistical analysis if you want a point-and-click option.

Experiment Steps

1. Define the learning task and decide what counts as attention loss, such as gaze leaving the answer region for a set pattern of time or distance.
2. Build two versions of the quiz, one standard version and one gaze-aware version, so you can compare them fairly.
3. Plan a calibration method and a logging scheme that records gaze events, answer changes, and response timing.
4. Choose outcome measures before you recruit participants, including immediate accuracy, delayed recall, and user frustration ratings.
5. Design controls that keep the content, screen layout, and instructions the same across both versions.
6. Select a statistical test that matches your sample size and outcome type, then decide how you will visualize the results.

Common Pitfalls

• Treating raw webcam gaze as exact eye position, which makes the interface react to noise instead of real attention shifts.
• Changing the wording or difficulty of quiz questions between versions, which confounds the effect of the gaze-aware design.
• Using bright or uneven room lighting, which reduces gaze-tracking stability and adds false re-prompts.
• Logging only final scores and not gaze events, which leaves you unable to explain why the interface changed performance.
• Making the re-prompt too aggressive, which interrupts normal reading and makes the system feel annoying instead of helpful.

What Makes This Competitive

A stronger project would compare more than one gaze threshold, feedback style, or quiz format. You could also test whether the effect changes by content type, question difficulty, or learner age group. Good projects here do careful A/B testing, track false positives, and use statistics that match the data. The best versions connect software design choices to learning outcomes, not just to raw quiz scores.

Project Variations

• Test whether gaze-aware prompting works better for math questions, vocabulary questions, or diagram-based questions.
• Compare webcam-based gaze tracking with mouse movement or scroll behavior as a cheaper proxy for attention.
• Study whether a softer reminder message works better than a full repeat of the question when attention drifts.

Learn More

• WebGazer.js GitHub repository: Read the documentation, examples, and code for webcam-based gaze estimation on GitHub.
• MIT OpenCourseWare, Human-Computer Interaction: Find course materials on interface design, usability testing, and user studies.
• NIH PubMed: Search for review articles on attention, feedback, and learning retention in educational software.
• IEEE Xplore: Search for peer-reviewed papers on gaze tracking, adaptive interfaces, and online learning systems.
• JASP: Use this free statistics program for t-tests, ANOVA, and effect size analysis.
• Nielsen Norman Group: Read free articles on usability testing and user interface design principles.