Kindergarten Science Classroom Dialogue Sentiment Analysis from Taiwan

98 kindergarten children and seven teachers in northern Taiwan participated in this study of sentiment during science activities. The research, authored by Ren-Cheng Zhang, used lexicon-based sentiment analysis on classroom conversations to correlate teacher strategies and student practices with expressed sentiments. Results indicate a significant correlation between teacher and child sentiments, with more negative sentiment during direct instruction and more positive sentiment during dialogic strategies like probing.

Use Cases

• Analyze sentiment alignment between teachers and students based on classroom dialogue transcripts.
• Investigate the impact of instructional strategies like 'Providing direct instruction' or 'Probing' on classroom affect.
• Study the relationship between specific scientific practices (e.g., 'Measuring', 'Planning') and student sentiment.
• Develop models for real-time sentiment feedback in educational settings based on conversational data.

Strengths

• Data is based on a study involving 98 children and 7 teachers, providing a concrete sample size.
• Findings include specific correlations, such as lower student sentiment during 'Employing equipment and tools'.
• The dataset is licensed under CC-BY-4.0, allowing for open sharing and reuse.

Limitations

• The dataset is very small at 30.5 KB, indicating limited scope and likely a small number of conversation samples.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data is from four kindergarten classes in northern Taiwan, which may limit generalizability to other regions or age groups.

Provenance

Source

figshare, authored by Ren-Cheng Zhang.

Collection Method

Lexicon-based sentiment analysis applied to transcribed classroom conversations in project-based science classrooms.

Freshness

Last updated 2026-05-15 05:39:51; freshness should be verified.

Geography

Four kindergarten classes in northern Taiwan.