English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Developing models to automatically score students' written responses toscience problems is critical for science education. However, collecting andlabeling sufficient student responses for training models is time andcost-consuming. Recent studies suggest that pre-trained language models can beadapted to downstream tasks without fine-tuning with prompts. However, noresearch has employed such a prompt approach in science education. As studentresponses are presented with natural language, aligning the scoring procedureas the next sentence prediction task using prompts can skip the costlyfine-tuning stage. In this study, we developed a zero-shot approach toautomatically score student responses via Matching Exemplars as Next SentencePrediction (MeNSP). This approach employs no training samples. We first applyMeNSP in scoring three assessment tasks of scientific argumentation and foundmachine-human scoring agreements, Cohen's Kappa ranges from 0.30 to 0.57, andF1 score ranges from 0.54 to 0.81. To improve the performance, we extend ourresearch to the few-shots setting, either randomly selecting labeled studentresponses or manually constructing responses to fine-tune the models. We findthat one task's performance is improved with more samples, Cohen's Kappa from0.30 to 0.38, and F1 score from 0.54 to 0.59; for the two others, scoringperformance is not improved. We also find that randomly selected few-shotsperform better than the human expert-crafted approach. This study suggests thatMeNSP can yield referable automatic scoring for student responses whilesignificantly reducing the cost of model training. This method can benefitlow-stakes classroom assessment practices in science education. Future researchshould further explore the applicability of the MeNSP in different types ofassessment tasks in science education and improve the model performance.

Matching Exemplar as Next Sentence Prediction (MeNSP): Zero-shot Prompt Learning for Automatic Scoring in Science Education