Explainable Automated Essay Scoring: Deep Learning Really Has Pedagogical Value

Abstract

Automated essay scoring (AES) is a compelling topic in Learning Analytics for the primary reason that recent advances in AI find it as a good testbed to explore artificial supplementation of human creativity. However, a vast swath of research tackles AES only holistically; few have even developed AES models at the rubric level, the very first layer of explanation underlying the prediction of holistic scores. Consequently, the AES black box has remained impenetrable. Although several algorithms from Explainable Artificial Intelligence have recently been published, no research has yet investigated the role that these explanation models can play in: (a) discovering the decision-making process that drives AES, (b) fine-tuning predictive models to improve generalizability and interpretability, and (c) providing personalized, formative, and fine-grained feedback to students during the writing process. Building on previous studies where models were trained to predict both the holistic and rubric scores of essays, using the Automated Student Assessment Prize's essay datasets, this study focuses on predicting the quality of the writing style of Grade-7 essays and exposes the decision processes that lead to these predictions. In doing so, it evaluates the impact of deep learning (multilayer perceptron neural networks) on the performance of AES. It has been found that the effect of deep learning can be best viewed when assessing the trustworthiness of explanation models. As more hidden layers were added to the neural network, the descriptive accuracy increased by about 10%. This study shows that faster (up to three orders of magnitude) SHAP implementations are as accurate as the slower model-agnostic one. It leverages the state-of-the-art in natural language processing, applying feature selection on a pool of 1592 linguistic indices that measure aspects of text cohesion, lexical diversity, lexical sophistication, and syntactic sophistication and complexity. In addition to the list of most globally important features, this study reports (a) a list of features that are important for a specific essay (locally), (b) a range of values for each feature that contribute to higher or lower rubric scores, and (c) a model that allows to quantify the impact of the implementation of formative feedback.

Predicting Student Academic Performance by Means of Associative Classification

“…They highlighted the need for using XAI in the educational field. In the context of automated essay scoring, the authors in [34] have studied the impact and trustworthiness of neural networks by means of the SHAP explanation framework [35]). Similar attempts have been made in the domains of computation thinking [36] and knowledge tracing [29].…”

Learning Automated Essay Scoring Models Using Item-Response-Theory-Based Scores to Decrease Effects of Rater Biases

“…Implementing a function to check the quality of predicted scores is another future research direction because we are sometimes interested in knowing the reliability or confidence level of scores predicted by AES. Furthermore, taking advantage of the unique property of the proposed method, namely, its high interpretability in terms of rater biases, another future direction is to analyze how rater biases affect the behavior of AES models, for example, by using an explanation model, as in [83].…”

Towards Trustworthy AutoGrading of Short, Multi-lingual, Multi-type Answers

“…Explainability techniques have also been assessed in the context of autograders (Kumar & Boulanger 2020). Explanations can increase understanding of automatic grading decisions and provide justification of those decisions.…”