Automated Classification of Classroom Climate by Audio Analysis

Abstract

While in training, teachers are often given feedback about their teaching style by experts who observe the classroom. Trained observer coding of classroom such as the Classroom Assessment Scoring System (CLASS) provides valuable feedback to teachers, but the turnover time for observing and coding makes it hard to generate instant feedback. We aim to design technological platforms that analyze real-life data in learning environments, and generate automatic objective assessments in real-time.To this end, we adopted state-of-the-art speech processing technologies and conducted trials in real-life teaching environments. Although much attention has been devoted to speech processing for numerous applications, few researchers have attempted to apply speech processing for analyzing activities in classrooms. To address this shortcoming, we developed speech processing algorithms that detect speakers and social behavior from audio recordings in classrooms. Specifically, we aim to infer the climate in the classroom from non-verbal speech cues. We extract non-verbal speech cues and lowlevel audio features from speech segments and we train classifiers based on those cues. We were able to distinguish between positive and negative CLASS climate scores with 70-80% accuracy (estimated by leave-one-out crossvalidation). The results indicate the potential of predicting classroom climate automatically from audio recordings.

DIAMOND: A tool for monitoring the participation of students in online lectures

“…However, the proposed system performs poorly as compared to modern diarization techniques, since the authors reported a Diarization Error Rate (DER) equal to 25%, which is much less than current systems that produce DERs between 5 and 10%. James et al ( 2019 ) propose a system that uses speech processing algorithms in order to detect speakers and social behavior from audio recordings in classrooms. The proposed system extracts non-verbal speech cues and low-level audio features from speech segments in order to infer the general climate in the classroom.…”

What Classroom Audio Tells About Teaching: A Cost-effective Approach for Detection of Teaching Practices Using Spectral Audio Features

“…James et al [14] collected data from 92 preschool classrooms and classified the overall climate of the lessons as positive or negative classroom climate according to the CLASS protocol. Afterward, they removed the silence from the recordings via thresholding and performed speaker diarization using the LIUM [25] toolkit.…”

“…Owens et al [27] developed an automated model for decibel analysis using decision trees and identified different teaching patterns related to active learning situations by analyzing the mean volume and its standard deviation along the lessons. James et al [14] used low-level audio features and conversational features to classify the overall climate of lessons according to the CLASS protocol and revealed that spectral features are better predictors for classroom climate as they carry emotional information because of their dependency on vocal folds.…”

Toward Automated Classroom Observation: Predicting Positive and Negative Climate

“…However, their system does not actually predict the CLASS scores themselves. James et al [34], [40] pursued an architecture similar to our prior work [15] for automatic recognition of CLASS climate scores. However, in contrast to the CLASS definition, which defines Positive Climate and Negative Climate as independent dimensions, their work treats these as two sides of a spectrum.…”