Writing Behavior as an Estimator of Self-Confidence on English Spelling Questions

Abstract

Reviewing plays an important role in the learning process of humans. A user's self-confidence is as important as the correctness of the answer to distinguish what to review. Giving feedback based on self-confidence enables learners to improve the quality of the review. Since it is still a familiar way for many students to write by hand, we propose a self-confidence estimation method based on handwriting behavior. To evaluate the performance of the proposed method, we employed the data collected while completing English spelling questions. The results revealed that the proposed method was able to predict self-confidence with 76% accuracy.