We present and evaluate a machine learning based system that automatically grades audios of students speaking a foreign language. The use of automated systems to aid the assessment of student performance holds great promise in augmenting the teacher’s ability to provide meaningful feedback and instruction to students. Teachers spend a significant amount of time grading student work and the use of these tools can save teachers a significant amount of time on their grading. This additional time could be used to give personalized attention to each student. Significant prior research has focused on the grading of closed-form problems, open-ended essays and textual content. However, little research has focused on audio content that is much more prevalent in language study education. In this paper, we explore the development of automated assessment tools for audio responses in a college-level Chinese language-learning course. We analyze several challenges faced while working with data of this type as well as the generation and extraction of features for the purpose of building machine learning models to aid in the assessment of student language learning.
more »
« less
Distinguishing Anxiety Subtypes of English Language Learners Towards Augmented Emotional Clarity
Public Speaking Anxiety (PSA) and Foreign Language Anxiety (FLA) afflict most English Language Learners (ELLs) during a presentation. However, few tools are available to help multicultural learners clearly identify which type of anxiety they are feeling. In this paper, we present a field study conducted in real language classrooms. We developed machine learning models based on features of electrodermal activity (EDA) to predict non-verbal behaviors manifested as PSA and FLA. The students were labeled with the anxiety categories both PSA and FLA, PSA more, FLA more, or no anxiety. To classify the ELLs into their respective anxiety categories, prominent EDA features were employed that supported the predictions of anxiety sources. These results may encourage both ELLs and instructors to be aware of the origins of anxiety subtypes and develop a customized practice for public speaking in a foreign language.
more »
« less
- Award ID(s):
- 1730033
- NSF-PAR ID:
- 10198278
- Date Published:
- Journal Name:
- International Conference on Artificial Intelligence in Education
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We present and evaluate a machine learning based system that automatically grades audios of students speaking a foreign language. The use of automated systems to aid the assessment of student performance holds great promise in augmenting the teacher’s ability to provide meaningful feedback and instruction to students. Teachers spend a significant amount of time grading student work and the use of these tools can save teachers a significant amount of time on their grading. This additional time could be used to give personalized attention to each student. Significant prior research has focused on the grading of closed-form problems, open-ended essays and textual content. However, little research has focused on audio content that is much more prevalent in the language-study education. In this paper, we explore the development of automated assessment tools for audio responses in a college-level Chinese language-learning course. We analyze several challenges faced while working with data of this type as well as the generation and extraction of features for the purpose of building machine learning models to aid in the assessment of student language learning.more » « less
-
more » « less
Abstract Although language has long been regarded as a primarily arbitrary system,
sound symbolism , or non‐arbitrary correspondences between the sound of a word and its meaning, also exists in natural language. Previous research suggests that listeners are sensitive to sound symbolism. However, little is known about the specificity of these mappings. This study investigated whether sound symbolic properties correspond to specific meanings, or whether these properties generalize across semantic dimensions. In three experiments, native English‐speaking adults heard sound symbolic foreign words for dimensional adjective pairs (big/small, round/pointy, fast/slow, moving/still) and for each foreign word, selected a translation among English antonyms that either matched or mismatched with the correct meaning dimension. Listeners agreed more reliably on the English translation for matched relative to mismatched dimensions, though reliable cross‐dimensional mappings did occur. These findings suggest that although sound symbolic properties generalize to meanings that may share overlapping semantic features, sound symbolic mappings offer semantic specificity. -
This work explores the experiences of administrators supporting teachers in the [redacted] program which emphasizes broadening participation in engineering at the high school level. In order to understand how efforts to broaden participation can leverage multiple levels of the school system, this work sought to identify the experiences of administrators. Two rounds of convenience samples of administrators in public and independent schools occurred in spring and fall of 2022. This recruitment led to two focus groups (with both public and independent school administrators) and a single administrator interview (independent school). During these conversations, administrators were asked to reflect on the implementation of the [redacted] program at their school, their personal experiences with this process, and barriers or suggestions in expanding this program both locally and more broadly. The transcripts of these interviews and focus groups were analyzed using descriptive coding (Saldaña, 2014) by two researchers. The researchers then met and compared codes to ensure inter-rater reliability with a percent agreement above 90%. During this process the codes were categorized and then emergent themes were identified. The findings indicate that administrators have a range of personal experience with implementing this engineering program, and that often these experiences were reported as a benefit to the entire school. For instance, administrators often referred to connections made to local universities by or as a result of the program, which then served as a positive outcome for the school at large. This suggests that a multifaceted approach to implementing engineering courses at the high school level which includes curriculum as well as human connections is seen as a benefit by administrators. Administrators also refer to the coursework and teacher actions as supporting the goal of broadening participation by speaking about specific engineering projects that engage a variety of learners, mostly through the type of project employed within the course. Finally, the administrators identified potential barriers to implementing engineering programs on their campus, with most barriers relating to teacher staffing issues, such as the need to engage in professional development to train these teachers externally. The themes that emerged from these various administrators perspectives provide an understanding of how to approach broadening participation through leveraging the role of administrators and how to encourage communication within the school system to create more access for students both to enroll in the engineering courses and to see a future self within engineering.more » « less
-
Does the lexicon of a language have consequences for cognition? Here, we provide evidence that the ease with which category features can be named can influence category learning. Across two experiments, participants learned to distinguish images composed of colors (Experiment 1) and shapes (Experiment 2) that were either easy or more difficult to name in English. Holding the category structure constant, when the underlying features of the category were easy to name, participants were faster and more accurate in learning the novel category. We argue that these findings suggest that labels allow learners to form more compact hypotheses, which in turn can be confirmed or disconfirmed in the course of learning. These results have consequences for considering how cross-linguistic differences in lexical inventory affect how readily novel categories are learned.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1007/978-3-030-52240-7_29
