Shannon McCrocklin

We examined the limitations of observed course DFWI rates (% of D and F grades, withdrawals, and incompletes) as evaluation metrics, which obscure student characteristics, course design, instruction, structure, and latent factors, posing challenges in identifying courses that need improvement. An artificial neural network (ANN) was trained using student data to model risk, accounting for variations in student characteristics. The model’s predictions on test data were averaged at the course level, producing expected DFWI rates based on student composition. Courses with high observed DFWI rates and large deviations between observed and predicted DFWI rates (the GAP) were ranked and prioritized for review, as they may reflect aspects of course design, structure, or instructional practices warranting further qualitative evaluation. Our predictions are non-causal, and modeling calibration varies across subgroups; therefore, the original GAP rankings, robust to a post-hoc calibration check, are presented as risk-adjusted indicators for prioritizing courses for further review rather than as definitive causal measures of course quality. Rankings based on observed DFWI rates differ substantially from risk-adjusted GAP rankings, indicating that relying on observed DFWI rates alone may misidentify high-risk courses. Our methodology can assist educators and administrators in making fair resource allocation decisions and improving student outcomes.

Over the last 20 years, automatic speech recognition (ASR) has improved substantially and spread into people’s everyday lives through hands‐free texting and intelligent personal assistants. ASR has also gained recognition within the field of second language learning as the transcript provided by ASR programs can support noticing of pronunciation errors. Given the growing interest and research in ASR, it is important to examine the accuracy of the technology for nonnative speech, especially given that studies have documented that second language speakers have been frustrated by low levels of ASR recognition. Using a three‐way ANOVA, this study examines the transcription accuracy of two ASR‐dictation programs, Google and Windows Speech Recognition ( WSR ), to understand accuracy as compared to human listeners for learners of three different first language backgrounds (Arabic, Chinese, and Spanish) on two different types of tasks (free and read speech). The results show that while Google had a higher level of transcription accuracy than WSR , WSR actually showed greater correlation with human listener intelligibility. While the speakers’ language background was not shown to have a significant effect, the task did, with WSR losing substantial accuracy in the free speech task. The results show a need for continued development and advancement of ASR technologies for second language speech.

This exploratory analysis investigates the potential of Gliglish, an AI chatbot designed for second language learning, to support pronunciation learning. The study focused on Gliglish's ability to respond to requests for pronunciation information and activities across five major pronunciation learning activity types: explanation, perception, prediction, controlled production, and guided production, as well as feedback provided during communicative role-plays when pronunciation errors were introduced. Results indicated that Gliglish struggled at times to understand prompts, provide sufficient explanations or examples, and to provide accurate information, but performed best when prompted to provide activities for controlled production. Furthermore, the analysis of feedback revealed Gliglish transcribed segmental errors less than half of the time (36.36% for consonants and 45.45% for vowels) and even less frequently for suprasegmental features (7.69%). Further, Gliglish rarely indicated pronunciation issues in its responses in subsequent turns. The findings suggest essential areas for improvement, emphasizing the need for enhanced feedback mechanisms and the incorporation of various activity types.

Although the field of second language (L2) pronunciation teaching has moved toward intelligibility models and goals, several studies have shown that many L2 learners still profess a desire to sound native-like in their L2. This study explores the perspectives of bilinguals who have navigated their accent and identity in two different languages to explore the impact of increasing recognition of links between accent and identity on language-learning goals. Findings show that almost all bilingual participants (97%) recognized links between accent and identity, yet 72% still reported preferring a native accent if they learned a new language today. Intelligibility emerged as a key issue, with those wanting a native accent linking nativeness to increased intelligibility while those rejecting native accents separating the two dimensions (accent and intelligibility) while prioritizing intelligibility.

Despite growth in second language pronunciation research, there is limited research on Corrective Feedback (CF) frequency and pronunciation. This study examined the effects of CF frequency on learner uptake, repair, and CF preferences. Participants (n = 9) joined an English pronunciation tutoring session and received high (100%) or low (50% feedback on productive activities) feedback. Surveys (pretest and posttest) established biographical data and explored preferences for CF. Sessions were analyzed for pronunciation errors, uptake, and repair success. Participants reported mostly positive emotions when receiving feedback. Participants wanted frequent CF. Results showed CF frequency did not affect uptake. Frequency may have impacted repair success (i.e., marginal significance [p=.08] was found when comparing high and low feedback groups); the low feedback group had a higher repair success rate. Results suggest that while learners prefer high feedback, we need to explore CF frequency and repair.

Although studies have explored teaching practices when teachers were thrust online during COVID-19, more work is needed to understand how COVID responses affected teachers' beliefs. This study, part of a larger investigation, explored teacher beliefs about teaching various language skills, including pronunciation, as they transitioned online. Participants (n=109) representing a range of backgrounds completed an online survey. Using a three-way ANOVA, this study identified a surprising finding involving the impact of CALL training, experience teaching online prior to COVID-19, and experience teaching online during COVID-19 on a ranking task of comparative difficulty of teaching different skills online. Most skills (reading, writing, listening, speaking, grammar) were perceived to be easier to teach by those with CALL training as they gained experience teaching online. Pronunciation was distinct in that teachers with CALL training perceived pronunciation to be more difficult as they began teaching online, suggesting areas for improvement in both CALL and pronunciation training.

The social meaning of language is embedded in a cultural system of ideas that circulates throughout the historicity of peoples in contact. Border areas are useful for studying the permeability of category constructions, including language boundaries and fluidity of identities. The present study reports survey data from 168 Spanish-English bilinguals in the border region of the Rio Grande Valley of South Texas in the ways they define ‘Spanglish’ vis-à-vis ‘Tex-Mex’ as two distinct ways of speaking. In explaining the contact forms that each label represents and ideologies that have given rise to this distinction, results are discussed using iconization and fractal recursivity (Irvine and Gal 2000) and enregisterment (Agha 2005). Overall, results indicate that for those who make a distinction between Spanglish and Tex-Mex, Spanglish is valorized as more legitimate, whereas Tex-Mex is seen as a localized style of Spanglish also considered more informal.

Second language (L2) accents are defined by pronunciation features differing from a particular standard, but listeners perceive more than sound differences when evaluating accents. L2 accents are indexical to a speaker’s perceived place in an L2 community and convey pragmatic information about a speaker’s identity. Accent thus impacts the ways learners create identities and manage social roles. Such identity creation may be especially important in being heard as a valid speaker of the L2, a prerequisite to identity construction and further language learning. This chapter explores the pragmatic force of L2 accent for learners in educational settings, by using data collected from language learners and bilinguals and examining participants’ complex reactions towards accents, and studying the multifaceted goals for their pronunciation.