Search

Background: Readability metrics provide us with an objective and efficient way to assess the quality of educational texts. We can use the readability measures for finding assessment items that are difficult to read for a given grade level. Hard‐to‐read math word problems can put some students at a disadvantage if they are behind in their literacy learning. Despite their math abilities, these students can perform poorly on difficult‐to‐read word problems because of their poor reading skills. Less readable math tests can create equity issues for students who are relatively new to the language of assessment. Less readable test items can also affect the assessment's construct validity by partially measuring reading comprehension. Objectives: This study shows how large language models help us improve the readability of math assessment items. Methods: We analysed 250 test items from grades 3 to 5 of EngageNY, an open‐source curriculum. We used the GPT‐3 AI system to simplify the text of these math word problems. We used text prompts and the few‐shot learning method for the simplification task. Results and Conclusions: On average, GPT‐3 AI produced output passages that showed improvements in readability metrics, but the outputs had a large amount of noise and were often unrelated to the input. We used thresholds over text similarity metrics and changes in readability measures to filter out the noise. We found meaningful simplifications that can be given to item authors as suggestions for improvement. Takeaways: GPT‐3 AI is capable of simplifying hard‐to‐read math word problems. The model generates noisy simplifications using text prompts or few‐shot learning methods. The noise can be filtered using text similarity and readability measures. The meaningful simplifications AI produces are sound but not ready to be used as a direct replacement for the original items. To improve test quality, simplifications can be suggested to item authors at the time of digital question authoring. Lay Description: What is Known About the Subject: Difficult to read math assessment items cause measurement and equity issuesReadability of math word problems negatively correlated with outcomes What Our Paper Adds: GPT‐3 AI system is capable to simplify math word problemsPrompt‐based and few‐shot learning‐based approaches are able to create meaningful simplifications, but with a very low accuracy rateText similarity and readability measures can be used to filter out noisy outputs and discover interesting simplifications What are the Implications for Practitioners: GPT‐3 AI is capable of generating relevant math world problem simplificationsSimplified text can be used as suggestions to question authors during the digital item authoring process [ABSTRACT FROM AUTHOR]

Multiple external representations (MERs) play an important role in the learning field of mathematics. Whereas the cognitive theory of multimedia learning and the integrative text and picture comprehension model assume that the heterogeneous combination of symbolic and analogous representations fosters learning; the design, functions, and tasks framework holds that learning benefits depend on the specific functions of MERs. The current paper describes a conceptual replication study of one of the few studies comparing single representations, heterogeneous, and homogeneous MERs in the context of mathematics learning. In a balanced incomplete block design, the participants were provided single representations (a graphic, text, or formula) or a heterogeneous (e.g., text + graphic) or homogeneous (text + formula) combination of these to solve linear system of equations problems. In accordance with previous research, performance was superior in conditions providing MERs compared to single‐representation conditions. Moreover, heterogeneous MERs led to time savings over homogeneous MERs which triggered an increase in cognitive load. Contrary to previous research, text was the least fixated representation whereas the graphical representation proved to be most beneficial. With regard to practical implications, experts should be fostered through more challenging homogeneous MERs whereas novices should be supported through the accessible graphic contained in heterogeneous MERs. Lay Description: What is already known about this topic: Providing multiple instead of single representations can foster mathematics learning.However, some combinations of representations hamper the performance of novice learners.The classical multimedia view explains the positive effect of combining text with picture.Recent research suggests the combination of multiple symbolic representations (text and formula). What this paper adds: The study replicated the effect of multiple symbolic representations for equation solving.In contrast to former research, each representation (text, formula, and graphic) was functionally equivalent.Tasks with differently coded representations were solved faster and with less mental effort.Gaze behavior indicated that students mostly used the graphical representation. Implications for practice and/or policy: Beginners should be supported with a graphic when solving equations.Experts should be challenged with a combination of symbolic representations to benefit the most. [ABSTRACT FROM AUTHOR]

This paper deals with multiobjective 0–1 programming problems having block angular structure. We propose an interactive fuzzy rule-satisfying method in order to obtain satisfactory solutions that take into account objective functions expressed by the decision maker in fuzzy form. In the proposed method, a membership function corresponding to each fuzzy object of the decision maker is specified by interaction with the decision maker. Subsequently, the Parato optimum solution, closest in the minimax sense to a reference point set by the decision maker in the membership function space, is derived by using a genetic algorithm. If satisfaction is not reached, the reference point is updated interactively so that the solution eventually derived is satisfactory to the solution maker. Since the minimax problem solved in this interactive process is a single-object 0–1 programming problem with block angular structure, a genetic algorithm is employed that contains a decomposition process for solving the problem by means of a special configuration. By means of simple numerical experiments, the effectiveness of the proposed method is demonstrated. © 1998 Scripta Technica, Electron Comm Jpn Pt 3, 81(8): 10–17, 1998 [ABSTRACT FROM AUTHOR]