MyVillage Project

MyVillage Project Confidence Analyzer

A research initiative testing how math narrative's messaging impacts student confidence in mathematics. Using custom-made AI models to analyze text, images, and video to measure confidence in educational settings.

Test the Study
Research Methods

How Confidence is Measured

Student-developed custom AI models analyze multiple indicators to assess confidence levels

Text Analysis

Sample Input
"Im doing better at understanding algebra still a little confused with the geometry"
AI Analysis Result
Confident 0%
Kinda Confident 100%
Not Confident 0%

Video Analysis

Sample Video Frame
Video analysis example - Not Confident
What the AI Analyzes: Posture, facial expressions, hand gestures, head position
Sample Video Result
Confident 0%
Kinda Confident 0%
Not Confident 100%

Image Analysis

Sample Image
Image analysis example - Kinda Confident
What the AI Detects: Posture, facial expressions, shoulder alignment, presence
Sample Image Result
Confident 100%
Kinda Confident 0%
Not Confident 0%
Study Methodology

Research Approach

This research uses AI-powered analysis to objectively measure confidence indicators across multiple modalities.

Multi-Modal Data Collection

Gather confidence data through written responses, video recordings, and images for comprehensive analysis.

Objective Measurement

AI models provide consistent, unbiased analysis of confidence indicators across all participants.

Research Insights

Generate detailed confidence metrics to help understand the impact of organizational messaging.

Non-Verbal Analysis

Body language and posture analysis provides additional data points beyond written responses.

Longitudinal Tracking

Monitor changes in student confidence over the study period to measure intervention effectiveness.

Participant Privacy

All data is handled with strict confidentiality in accordance with research ethics guidelines.

Research Technology

The AI Models

Custom-trained models developed specifically for this research study

How the AI Models Work

Two student-developed custom AI models are used to understand confidence. These models learned from thousands of real examples to recognize patterns that indicate how confident someone feels:

  • Text Model: Analyzes the way something is said and what is being said to identify confidence. View on Hugging Face
  • Image/Video Model: Looks at posture and facial expressions to determine confidence levels. View on Hugging Face
  • Each submission receives a clear result: Confident, Kinda Confident, or Not Confident
  • Results help researchers understand how messages affect student confidence in math

8,605

Text Examples Used for Training

897

Images Used for Training

3

Ways to Measure Confidence

< 5s

Time to Get Results

Assessment Methods

Data Collection Modalities

Multiple methods for measuring student math confidence

Text Analysis

Share your thoughts in writing, and the AI model analyzes the way something is said and what is being said.

  • Learned from 8,605 real examples
  • Analyzes word choice and phrasing
  • Clear confidence rating
  • Results in seconds
Learn More

Video Analysis

Record a short video, and the AI model analyzes your posture and facial expressions to determine confidence.

  • Analyzes posture and facial expressions
  • Evaluates each moment
  • Works in different lighting
  • Shows confidence over time
Learn More

Image Analysis

Take or upload a photo, and the AI model analyzes your posture and facial expressions to determine confidence.

  • Learned from 897 real images
  • Analyzes posture and facial expressions
  • Works with any background
  • Take a photo or upload one
Learn More

Ready to Participate?

Join this research study on student math confidence. Your participation helps researchers understand how organizational messaging impacts learning.