Research Methods in Applied Linguistics最新文献

Language contact in peer interactions during study abroad has drawn considerable attention from applied linguists and educators. Scholars have developed various tools to collect language contact and social interaction information, such as the Language Contact Profile, the Language Log, and the Social Interaction Questionnaire. This paper employs a whole network design with peer evaluation to collect data on social interaction, use of languages, perceived language proficiency, and total time spent using different languages among 23 students with diverse backgrounds. These students were enrolled in an English-medium program in China that caters exclusively to international students. Findings show that the whole network approach can effectively cross-validate students’ self-reported language use and proficiency alongside their social interactions. The paper also offers several strategies, rooted in the whole network approach, to detect potential inconsistencies in these self-reported measures that warrant further investigations.

Frequency of (aural) language exposure and productive (oral) language use, context, and interlocutors are important factors for language learning. Applied linguists and study abroad (SA) researchers have studied these variables mostly retrospectively through questionnaires and learners’ self-reports which have unknown correspondence with learner's actual language use and social networks. Surprisingly few applied linguists have attempted to capture quantitative aspects of day-to-day language use through systematic long-term observation even though methods to do so have been used in psychological science for some time. In this article, we present arguments for the application of acoustic behavioral observation in applied linguistics and SA research. We introduce the Electronically Activated Recorder (EAR; Mehl, 2017) and illustrate its application in a case study about the use of first, second, and additional languages by two German high school students during their year abroad in the USA. The EAR is a digital audio recorder, currently available as an app for Android devices. It can be installed on smartphones worn by their users as they go about their daily activities. The app intermittently (e.g., 5 times per hour) records brief (e.g., 30 s) snippets of ambient sounds from the wearer's moment-to-moment sonic environment. EAR users are unaware of when exactly the device is recording. We describe the method, data collection, transcription, coding and analyses, address ethical concerns about its application, and report findings from the case study about the SA students’ language use to illustrate the potential and limitations of the method for applied linguistics research.

Despite the growing use of corpus linguistics across an ever-growing range of disciplines such as sociology, sports studies, journalism, media discourse or education, there is a dearth of research that examines the epistemological foundations of corpus methods in these disciplines. This paper builds on well-established conceptualisations about research methodology and the role of methods in the wider literature. Drawing on existing discussions about the use of research methods in objectivist and subjectivist conceptualisations of social reality, we seek to bring to the fore the underlying methodological tensions found in the use of corpus linguistics in the application of corpus methods in research that lies outside the interest of major linguistic disciplines. Through this process, we explore how the notions of natural language use and data elicitation are interpreted by current research in order to advance our understanding of how experts from different research camps engage with and epistemologically localise corpus linguistics.

Authentic materials are an important resource for teaching and assessment, especially when it comes to English for Specific Purposes. To this end, corpus linguistics has developed a supporting role as it provides a wide range of possibilities for extracting and exploring authentic data. With regard to Aeronautical English (AE), which refers to oral communications between pilots and air traffic controllers (ATCOs), corpora offer substantial benefits in the development of teaching and testing materials. However, putting together corpora in the context of AE faces challenges in availability, selection, compilation, and copyright issues as some countries limit access to radiotelephony communications due to their sensitive content. This paper reports the experience of compiling a collaborative corpus called “Aerocorpus” proposed during two webinars held for AE teachers and testers from all over the world. The compilation methodology aims to address the aforementioned challenges and offer alternative solutions. Despite existing limitations, we see considerable gains in using authentic materials and are looking into ways of protecting the “black box” containing sensitive information while continuing to compile the corpus, which currently features 38,135 words from 45 situations, and to make it available to a wider community of AE professionals.

This article proposes an innovative approach to examining garden-path (GP) effects in sentence processing. It applies GP effects as a method to a new domain, specifically to syntactic engagement of recently learned verbs. We tested twenty-seven English native speakers and twenty Spanish learners of English to verify method validity. Three main components characterise the method, namely training of new word meaning through definitions and example sentences, eye-tracking while reading plausible and implausible GP sentences after sleep consolidation, and a meaning recall test. We also examined if participants’ phonological working memory and vocabulary size play a role in how they syntactically engage new words. Results showed that recently learned verbs can elicit syntactic engagement in both native and nonnative readers. Both vocabulary size and phonological working memory capacity could predict ambiguity reprocessing, irrespective of language group. These results indicate that garden pathing can reliably signal effort to detect and resolve subject-object ambiguities in both first language (Frazier & Rayner, 1982; Pickering & Traxler, 1998) and second language readers (Chen et al., 2021, Jegerski, 2012). This feasibility study is a pioneering attempt to map new vocabulary knowledge as a window into emergent structural representations. The significance of this method lies in its potential to track syntactic engagement of new lexis, while accounting for individual differences, and following the principle that to know a word entails knowing its form, meaning, as well as its grammatical use (Nation, 2001).

The present study aims to validate a new scale, i.e., self-efficacy beliefs in peer feedback (SEBPF), using an advanced method of bifactor ESEM. Informed by sociocognitive theory and self-regulated learning, the SEBPF was designed to measure five distinct factors: self-incentives for peer feedback, perceived use of peer feedback, confidence in peer feedback, managing stress in peer feedback, and setting proximal goals for peer feedback. The items for the SEBPF scale were developed by extracting relevant themes from published studies and conducting interviews with a sample of English as a Foreign Language (EFL) student writers in China. This paper provides a comprehensive account of the developing and validating process of the SEBPF scale. Comparisons of confirmatory factor analysis and exploratory factor analysis support the criterion-related validity of a coherent 5-factor structure comprising 22 items, indicating the reliability and multifaceted nature of the SEBPF scale. The findings demonstrate the robustness and practical applicability of the SEBPF scale in assessing EFL writers’ self-efficacy beliefs associated with peer feedback, as well as its predictive effects on different sub-scores of writing performance. This paper is an empirical application of validating a new scale through bifactor ESEM. Implications were proposed based on the findings.

While general first language corpora are composed of samples from various naturalistic sources (e.g., websites, books), language samples in most written learner corpora (LC) are texts produced in response to prompts. In this context, LC users need to develop a clear awareness of the affordances and limitations of specific prompts and how responses to said prompts may affect the investigation of their intended object(s) of study. Through an analysis of the presence/absence of specific Spanish verb tenses in texts written in response to two supposedly narrative prompts in a Spanish LC (COWS-L2H; Yamada et al., 2020), this article illustrates the impact of inter- and intra-prompt response variation on LC data interpretation. Based on this evidence, we caution against rapid assumptions about text content based solely on the superficial phrasing of LC writing prompts. Instead, we recommend that LC users perform in-depth quantitative and qualitative analyses of learners’ samples written in response to each prompt they aim to include in their study prior to running statistical models on those data.

The power, convenience, and efficiency made possible by online communication tools have become increasingly important in qualitative research. The tradition's most central data collection method, interviewing, is in turn rapidly changing shape as these tools are integrated into data collection processes. Drawing from the perspective of interviewing as a social process (Talmy, 2010), as well as the Online Interviewing Ecology framework (Meskill et al., 2024), which includes three main elements—modalities, time, and relationship, this study examines the emerging strategies and characteristics of qualitative online interviewing in applied linguistics. Applying specific selection criteria, we interviewed the interviewer/authors of 6 recent empirical studies in applied linguistics that employed online qualitative interviewing for data collection. Thematic analysis of resulting transcripts in tandem with the authors' published empirical papers provides the basis for rich description of how researchers conceived of and deepened interviews through a dialogic meaning making approach. Findings suggest that online interviewing can be a very productive means for co-constructing meaning between interviewers and interviewees. Participants' responses also helped us to reconceive our initial framework: the interconnected categories of time and modalities serve to support the development of the interviewer-interviewee relationships that more directly shapes the flow and quality of the co-constructed interview discourse.

This paper presents the development of two novel discourse-gating tasks to investigate the processing of pragmatic information, namely, the timing of the recognition of genuine (sincere) and ostensible (transparently insincere) refusals in Chinese and provides preliminary validity evidence for the tasks. Gating tasks were introduced to investigate spoken word recognition and have been successfully extended to spoken language processing, most notably sentences. Following Grosjean's (1996) observation that gating tasks could be used to investigate a variety of linguistic features, we extended the gating tasks to spoken discourse using turns as gates. The open-prediction gating task allows participants to make a single prediction about the outcome of each of 12 recorded conversations as soon as they can. The fixed-prediction gating task asks participants to make predictions at regular intervals while listening to a second set of 12 conversations. One hundred and seven participants (60 L1 speakers and 47 third- and fourth-year learners of Chinese) were recruited to test the tasks. The tasks reveal a lag in speech-act identification not found when retrospective speech-act identification tasks are used. The fixed-prediction task additionally reveals alternatives that are considered during processing. The paper discusses the benefits of the discourse gating tasks and the merits of each, the quantitative and qualitative evidence for the tasks, and future directions for discourse gating tasks.

This tutorial demonstrates how to automate the scoring of two oft-used English morphosyntactic forms, be-passive and third person singular -s, using commercial Natural Language Processing services. It focuses specifically on the context of elicited imitation (EI) tests drawing on previously web-collected EI data (Kim & Godfroid, 2023; Kim et al., 2024). We provide step-by-step instructions and example codes covering three key stages of data processing: (1) speech-to-text transcription, (2) identification of morphosyntactic structures, and (3) the scoring algorithm. This method can be applied to various form-based EI scoring schemes or other form-based automatic scoring tasks, enhancing the broader adoption and practical application of automated scoring in both research and educational settings.