Investigating relationships of sentiments, emotions, and performance in professional development K-12 CS teachers

ABSTRACT

Background and Context: Professional development (PD) programs for K-12 computer science teachers use surveys to measure teachers’ knowledge and attitudes while recognizing daily sentiment and emotion changes can be crucial for providing timely teacher support.

Objective: We investigate approaches to compute sentiment and emotion scores automatically and identify associations between the scores and teachers’ performance.

Method: We compute the scores from teachers’ assignments using a machine-assisted tool and measure score changes with standard deviation and linear regression slopes. Further, we compare the scores to teachers’ performance and post-PD qualitative survey results.

Findings: We find significant associations between teachers’ sentiment and emotion scores and their performance across demographics. Additionally, we find significant associations that are not captured by post-PD qualitative surveys.

Implications: The sentiment and emotion scores can viably reflect teachers’ performance and enrich our understanding of teachers’ learning behaviors. Further, the sentiment and emotion scores can complement conventional surveys with additional insights related to teachers’ learning performance.

KEYWORDS:

• Professional development
• K-12 teachers
• computer science education
• sentiment analysis
• emotion analysis
• machine-based text analysis

Acknowledgments

This work was supported by a National Science Foundation grant no. 1837476. All findings and opinions are those of the authors and not necessarily of the funding agency. The authors would like to acknowledge the instructors and teaching assistants and teachers of the four summer institutes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Science Foundation [1837476].