Xiaowan Zhang

Xiaowan Zhang

Auto-generated profile, based on publicly available information

Xiaowan Zhang

Duolingo, Pittsburgh, PA, USA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 August 2021 in Chemical Communications
Verified
Journal Article
Closed access
PDF file
WZWenli Zhu
Wenli Zhu
PQPeige Qin
Peige Qin
LHLizhen Han
Lizhen Han
XZXiaowan Zhang
Xiaowan Zhang
DLDan Li
Dan Li
ZCZongwei Cai
Zongwei Cai
Herein, a new gas-cycle-assisted (GCA) headspace solid-phase microextraction (HS-SPME) device was designed to rapidly extract organic pollutants with high Kow and boiling points, which have difficulty in volatilization from matrix to headspace. Organic pollutants, including three polycyclic aromatic hydrocarbons (PAHs), four polychlorinated biphenyls (PCBs), and five phthalate esters (PAEs), were selected to evaluate the performance of GCA HS-SPME. Compared with conventional HS-SPME, the equilibrium times of GCA HS-SPME for extraction of PAHs, PCBs, and PAEs were greatly shortened from 70–90 to 5–11 min. Moreover, the limits of detection for analysis of PAHs were achieved at pg mL−1 level by GCA HS-SPME coupled with gas chromatography-flame ionization detection.

28 April 2022 in Studies in Second Language Acquisition
Verified
Journal Article
Open access
PDF file
PWPaula Winke
Paula Winke
XZXiaowan Zhang
Xiaowan Zhang
SPSteven J. Pierce
Steven J. Pierce
Second language (L2) teachers may shy away from self-assessments because of warnings that students are not accurate self-assessors. This information stems from meta-analyses in which self-assessment scores on average did not correlate highly with proficiency test results. However, researchers mostly used Pearson correlations, when polyserial could be used. Furthermore, self-assessments today can be computer adaptive. With them, nonlinear statistics are needed to investigate their relationship with other measurements. We wondered, if we explored the relationship between self-assessment and proficiency test scores using more robust measurements (polyserial correlation, continuation-ratio modeling), would we find different results? We had 807 L2-Spanish learners take a computer-adaptive, L2-speaking self-assessment and the ACTFL Oral Proficiency Interview – computer (OPIc). The scores correlated at .61 (polyserial). Using continuation-ratio modeling, we found each unit of increase on the OPIc scale was associated with a 131% increase in the odds of passing the self-assessment thresholds. In other words, a student was more likely to move on to higher self-assessment subsections if they had a higher OPIc rating. We found computer-adaptive self-assessments appropriate for low-stakes L2-proficiency measurements, especially because they are cost-effective, make intuitive sense to learners, and promote learner agency.

20 April 2020 in Language Learning
Verified
Journal Article
Open access
PDF file
XZXiaowan Zhang
Xiaowan Zhang
PWPaula Winke
Paula Winke
SCShaunna Clark
Shaunna Clark
Answering calls to map college‐level proficiency development (Modern Language Association, 2007) and longitudinally chart language learning (Barkaoui, 2014; Ortega & Byrnes, 2008), we mapped the oral proficiency growth of 1,922 lower‐division college students of Chinese, French, Russian, and Spanish (in the second, third, or fourth semester of their programs), using the Oral Proficiency Interview–computer (OPIc). We recorded students’ gender, heritage‐learner status, high school language experience, interest in learning the target language, perceived importance of speaking, and outside‐of‐class second language (L2) contact, in order to differentiate growth. Latent growth curve analyses showed that learners’ oral proficiency progressed at an average rate of one sublevel per year on the scale of the American Council on the Teaching of Foreign Languages (2012), following a nonlinear developmental path that was mildly accelerating over time. Students’ proficiency development had a significant relationship with the language being learned, heritage‐learner status, high school language experience, interest in learning, and perceived importance of speaking. No unique effects were found for gender or out‐of‐class L2 contact.

24 June 2024 in TESOL Quarterly
Verified
Journal Article
Open access
XZXiaowan Zhang
Xiaowan Zhang
PWPaula Winke
Paula Winke
We investigated the time it takes 54,146 English learners (ELs) to attain English proficiency as they progressed from age 5 to 11 on average (Kindergarten through fifth grade in the United States). We also examine to what extent the time‐to‐proficiency estimate is affected by child‐internal and child‐external factors, including primary disability status, primary home language, home English use, instructional programming, and retention status. We used discrete‐time survival analysis to analyze the children's English growth. Survival analysis often concerns time from treatment until death: Here, “survival” is the time from EL‐program entry until “proficiency,” indicated via standardized testing. Half of the students attained proficiency in 5 years. Literacy skills (reading and writing) in general developed at slower rates than oral language skills (speaking and listening), with writing being ELs' biggest barrier to proficiency. While time to proficiency was significantly related to primary disability status, primary home language, and retention status, exposing ELs to their home language at home or at school does not have a substantial effect on their rates of English acquisition. The results are discussed for their research and practical implications.

27 July 2026 in Language Testing
Verified
Journal Article
Open access
YPYena Park
Yena Park
YAYigal Attali
Yigal Attali
XZXiaowan Zhang
Xiaowan Zhang
JCJacqueline Church
Jacqueline Church
KLKai-Ling Lo
Kai-Ling Lo
GLGeoffrey T. LaFlair
Geoffrey T. LaFlair
Despite task interactivity eliciting different aspects of language use, more interactive, adaptive speaking tasks have been challenging to implement in computer-delivered large-scale high-stakes contexts. Recent advances in large language models (LLMs) offer a potential means to address this tension through spoken dialogue systems (SDSs) but fall short of full adaptivity. To enhance adaptivity in an LLM-enhanced SDS-delivered interview task, this article proposes two solutions: (1) LLM-driven response evaluation using question-specific task completion rubrics for adaptive follow-up question selection and (2) item response theory (IRT) scoring accounting for prompt variability. We examined the extent to which these solutions functioned to simulate adaptivity, from 5,909 test takers’ responses to an adaptive speaking task with an avatar and a non-adaptive monologic task. Automated task completion evaluations were comparable to expert ratings, though follow-up question evaluation showed low agreement both among human raters and between raters and the system. IRT scoring yielded higher test–retest reliability, and the adaptive task elicited more reciprocal language use than the monologic task. Unlike earlier SDSs that prioritized consistency over adaptivity, the real-time meaning-level evaluation with IRT modeling strikes a balance between standardization and adaptivity. These findings support the viability of adaptive speaking tasks in computer-based standardized assessments.

1 September 2021 in Studies in Educational Evaluation
Verified
Journal Article
Closed access
XZXiaowan Zhang
Xiaowan Zhang