Search

The purpose of this study is to analyze eye movements of second language learners of Mandarin Chinese during online reading to determine patterns of attention. I intend to replicate and expand part of Stickler and Shi's (2014) original eye tracking experiments on Chinese second language learners completing an online reading activity with Chinese characters and Pinyin transcription scaffolding. Since Chinese uses characters, reading can prove a daunting task for learners, especially those at the beginning levels (Wang, 2014). I have personally experienced this frustration myself and observed similar difficulties for other students. Although, many characters do contain some phonetic information, it is usually easier for students to learn to read with the help of a phonetic transcription system such as Pinyin, which uses the Roman alphabet “to approximate the pronunciation of Chinese characters” (ibid, p. 53). Often textbooks and reading materials will annotate some or all of the included Chinese characters in such a transcription system, at least for beginner levels, for the benefit of first or second language learners (See Figures 1 and 2).

According to Barzegar (2013), the way we arrange words in different positions of the clause is essential to communicate in any language. In other words, word order is highly important when creating meaning. Second language learners usually aim to have a clear understanding of how and where to place words, so they can achieve successful communication in the target language. Kessler (2004) found that native speakers and advance language learners of English have a similar level of syntactical knowledge, especially in terms of word order. However, this word-order knowledge does not seem equally applied to different structures and languages. In this regard, Jin (2008) discovered that English speakers have much difficulty in learning marked structures in Chinese. Similarly, Solís Hernández (2006) demonstrated that L2 English learners usually struggle with structures that do not follow the same pattern at all times, especially in relation to adverb placement. She described that most students did not clearly understand the reasons why adverbs in English are placed in different positions of the clause. In my experience, I have noticed that small differences in clause constructions (e.g., “I saw him yesterday” and “Yesterday I saw him”), can be confusing to L2 learners. This is not only because of the syntactic order, but because sometimes it is difficult for learners to understand the reasons that have motivated a speaker or a writer to select a less common word order construction to communicate a specific message.

The purpose of this paper is to present an overview of ESL reading test development and analysis in the context of a small-scale ESL classroom. We created 12 multiple-choice items for an ESL reading mid-term exam, administered the test in the Community English Program at Teachers College in fall 2008, and analyzed the test results to evaluate the reliability and validity of the test. We first describe the nature of reading ability by reviewing the literature on second language reading and reading assessment. Based on a widely shared definition of reading ability, we suggest a theoretical construct of reading ability and relevant observable variables. Following Bachman and Palmer (1996), this paper provides practical guidance for language teachers with regard to how to create reading test items and assess the test quality from describing the target language use (TLU) domain and task types, developing a test design statement, generating the blueprint for test operationalization, coding multiple choice items, to conducting item and distractor analyses. Issues revolving around L2 reading test development are further discussed.

A prominent hypothesis holds that by speaking to infants in infant-directed speech (IDS) as opposed to adult-directed speech (ADS), parents help them learn phonetic categories. Specifically, two characteristics of IDS have been claimed to facilitate learning: hyperarticulation, which makes the categories more separable, and variability, which makes the generalization more robust. Here, we test the separability and robustness of vowel category learning on acoustic representations of speech uttered by Japanese adults in ADS, IDS (addressed to 18- to 24-month olds), or read speech (RS). Separability is determined by means of a distance measure computed between the five short vowel categories of Japanese, while robustness is assessed by testing the ability of six different machine learning algorithms trained to classify vowels to generalize on stimuli spoken by a novel speaker in ADS. Using two different speech representations, we find that hyperarticulated speech, in the case of RS, can yield better separability, and that increased between-speaker variability in ADS can yield, for some algorithms, more robust categories. However, these conclusions do not apply to IDS, which turned out to yield neither more separable nor more robust categories compared to ADS inputs. We discuss the usefulness of machine learning algorithms run on real data to test hypotheses about the functional role of IDS. © 2021 The Authors. Cognitive Science published by Wiley Periodicals LLC on behalf of Cognitive Science Society (CSS).