View in EDS HTML Full Text PDF Full Text

Profiling Text Cohesion in the Development of L2 Chinese Reading Materials: Variation by Text Level and Genre

Saved in:
Bibliographic Details
Title: Profiling Text Cohesion in the Development of L2 Chinese Reading Materials: Variation by Text Level and Genre
Language: English
Authors: Yun Lin, Yanfang Su, Yiping Peng, Hua Liu
Source: Reading and Writing: An Interdisciplinary Journal. 2025 38(7):1949-1984.
Availability: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
Peer Reviewed: Y
Page Count: 36
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Descriptors: Connected Discourse, Reading Comprehension, Reading Materials, Material Development, Literary Genres, Second Language Learning, Chinese, Narration, Persuasive Discourse, Expository Writing, Effect Size
DOI: 10.1007/s11145-024-10585-w
ISSN: 0922-4777
1573-0905
Abstract: Developing reading materials tailored to the specific needs and proficiency levels of students is important yet challenging for educators and teachers. Current data-driven indices assisting this task primarily focus on lexical and syntactic text features, often overlooking text cohesion, a key discourse feature impacting reading comprehension. To fill this gap, this study explored the variation of text cohesion in second language (L2) Chinese reading materials across different text levels and genres. We used an NLP-based tool to examine the lexical, grammatical, and topical cohesion features of 450 representative L2 Chinese reading texts. A two-way MANOVA was then conducted to analyze the variation of text cohesion across different text levels and genres, as well as the interaction between text levels and genres. Our findings have revealed significant disparities in text cohesion across varying text levels, particularly in lexical cohesion features, and among different genres, with narratives standing out from argumentative and expository texts. Interaction effects between text level and genre were noted, albeit with small effect sizes. This study highlights the importance of text cohesion when developing reading materials for L2 Chinese learners and offers insights into developing text cohesion benchmarks for preparing reading materials of different levels and genres.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1482417
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHY_EcCYmW57TB4w_0XZcqyAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDHdTVqyvWyobqgVrbwIBEICBm1Dr4FQT46D093uUgJuhHQlBL-hTMPytVFCoV4bkEdVxccqBrOl55C58HOkwtkjRA2Q9gdCMvgvknhipvXz3bmuXo0uMhDYSXRgyzwnWazRpc7qeO931NJN4smD8mGhy4FVgnHTC-5WRG5TSiu8Z_ZC66ZZ22r08G9M638Ad64pPVWFbYDmKzkhZTgqGEqUUupOVwOE6vdqfX0l6
Text:
  Availability: 1
  Value: <anid>AN0187667272;2ap01sep.25;2025Sep03.06:43;v2.2.500</anid> <title id="AN0187667272-1">Profiling text cohesion in the development of L2 Chinese reading materials: variation by text level and genre </title> <p>Developing reading materials tailored to the specific needs and proficiency levels of students is important yet challenging for educators and teachers. Current data-driven indices assisting this task primarily focus on lexical and syntactic text features, often overlooking text cohesion, a key discourse feature impacting reading comprehension. To fill this gap, this study explored the variation of text cohesion in second language (L2) Chinese reading materials across different text levels and genres. We used an NLP-based tool to examine the lexical, grammatical, and topical cohesion features of 450 representative L2 Chinese reading texts. A two-way MANOVA was then conducted to analyze the variation of text cohesion across different text levels and genres, as well as the interaction between text levels and genres. Our findings have revealed significant disparities in text cohesion across varying text levels, particularly in lexical cohesion features, and among different genres, with narratives standing out from argumentative and expository texts. Interaction effects between text level and genre were noted, albeit with small effect sizes. This study highlights the importance of text cohesion when developing reading materials for L2 Chinese learners and offers insights into developing text cohesion benchmarks for preparing reading materials of different levels and genres.</p> <p>Keywords: Text cohesion; L2 Chinese reading materials; Text level; Text genre; Reading material development; Language; Communication and Culture Linguistics</p> <p>Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.</p> <hd id="AN0187667272-2">Introduction</hd> <p>Developing reading materials with appropriate proficiency levels and contents for target students is crucial for effective reading instruction (Sung et al., [<reflink idref="bib63" id="ref1">63</reflink>]). Typically, experts and teachers select or adapt suitable reading texts for learners across various text levels and genres (Mesmer et al., [<reflink idref="bib49" id="ref2">49</reflink>]). The introduction of corpus techniques has led to the development of data-driven indices that effectively characterize text complexity across diverse proficiency levels and genres (Lu, [<reflink idref="bib45" id="ref3">45</reflink>]). These indices, which include syntactic complexity, lexical complexity, and vocabulary profiles of texts (Chujo, [<reflink idref="bib11" id="ref4">11</reflink>]; Jin & Lu, [<reflink idref="bib34" id="ref5">34</reflink>]; Jin et al., [<reflink idref="bib35" id="ref6">35</reflink>]; Mesmer et al., [<reflink idref="bib49" id="ref7">49</reflink>]), have proved to be valuable in the preparation of English reading materials. While the importance of lexical and syntactic complexity features is widely acknowledged, excessively focusing on these aspects during adaptation could lead to disjointed ideas, resulting in texts that are less comprehensible to readers (Crossley et al., [<reflink idref="bib14" id="ref8">14</reflink>]; Hiebert & Mesmer, [<reflink idref="bib29" id="ref9">29</reflink>]). Hence, discourse features of text, such as cohesion, which refers to the explicit connections between different text segments achieved by various objective linguistic features (Crossley et al., [<reflink idref="bib14" id="ref10">14</reflink>]), should be given due consideration. In addition, while a multitude of data-driven indices are available for developing English reading materials, there is a notable scarcity of quantitative guidance for other language learning contexts, such as Chinese, a language that is currently witnessing a significant surge in learners (Chan et al., [<reflink idref="bib8" id="ref11">8</reflink>]).</p> <p>To address these research gaps, this study aims to examine the variation of text cohesion in L2 Chinese reading materials across different text levels and genres. It is hoped that this study can contribute to our understanding of the requirements of text cohesion in L2 Chinese reading materials across different genres and different levels of study, as well as informing both teachers and material developers in the selection and adaptation of more effective reading materials for L2 Chinese learners.</p> <hd id="AN0187667272-3">Literature review</hd> <p></p> <hd id="AN0187667272-4">Text level and genre in reading material development</hd> <p>The prerequisite of effective reading instruction is to develop reading materials that cater to the specific needs of the intended students. To facilitate successful learning, the reading materials should neither be too challenging which may impede comprehension nor too simplistic which might undermine motivation (Hiebert & Mesmer, [<reflink idref="bib29" id="ref12">29</reflink>]). Experts and teachers commonly take consideration of text levels and genres when selecting or adapting reading texts (Hiebert & Mesmer, [<reflink idref="bib29" id="ref13">29</reflink>]; Mesmer et al., [<reflink idref="bib49" id="ref14">49</reflink>]).</p> <p>The majority of research has been concentrated on profiling variations of linguistic features, with a special emphasis on lexical and syntactic features, across different text levels (Jin & Lu, [<reflink idref="bib34" id="ref15">34</reflink>]; Jin et al., [<reflink idref="bib35" id="ref16">35</reflink>]; Su et al., [<reflink idref="bib61" id="ref17">61</reflink>]). These studies highlighted the importance of linguistic characteristics in leveling and preparing reading materials. However, these studies have not considered the influence of genre, which can exert a substantial influence on the language proficiency required to comprehend the text (Tortorelli, [<reflink idref="bib65" id="ref18">65</reflink>]). Therefore, it is crucial to take into account the text genre during the process of the reading materials selection and adaptation (Jin et al., [<reflink idref="bib35" id="ref19">35</reflink>]; Mesmer et al., [<reflink idref="bib49" id="ref20">49</reflink>]; Sheehan et al., [<reflink idref="bib58" id="ref21">58</reflink>]). Some scholars identified variations of linguistic features, such as core vocabularies (Lee, [<reflink idref="bib38" id="ref22">38</reflink>]), average word length and word frequency, average sentence length, and referential cohesion (Sheehan et al., [<reflink idref="bib58" id="ref23">58</reflink>]), across genres. In a more recent study, Mesmer et al. ([<reflink idref="bib49" id="ref24">49</reflink>]) observed that reading texts from different content areas exhibited variations in word frequency, word length, and sentence length. They noted that the measurements alone were not a reliable discriminator for texts across different levels. A notable interaction effect between level and genre was also identified on the relationship between sentence length and reading comprehension. As a result, they advocated for a thoughtful consideration of the interaction between text level and genre when selecting reading texts.</p> <p>In summary, prior studies underscore the importance of considering both text level and text genre when preparing reading materials for instruction. The findings of these studies have contributed to validating the instructional effectiveness of reading materials and setting benchmarks for material selection and adaptation. Yet, while a significant amount of research has been devoted to investigating lexical and syntactic features, text cohesion features, which is critical for the clarity and continuity of the discourse (Hoey, [<reflink idref="bib30" id="ref25">30</reflink>]), remain relatively underexplored. Thus, there is a need for more concerted efforts towards profiling text cohesion in reading texts across varying text levels and genres.</p> <hd id="AN0187667272-5">Conceptualization and measurements of text cohesion</hd> <p>Text cohesion refers to the connections among different text segments explicitly achieved by various textual features to contribute to the overall unity of a text (Crossley et al., [<reflink idref="bib14" id="ref26">14</reflink>]; Hoey, [<reflink idref="bib30" id="ref27">30</reflink>]). As one of the text readability features (Sung et al., [<reflink idref="bib63" id="ref28">63</reflink>]), text cohesion is critical for reading comprehension (Tortorelli, [<reflink idref="bib65" id="ref29">65</reflink>]). It aids readers in forming a coherent mental representation of the text's content and organization, thereby making the coherence of the text explicit for the reader (Graesser et al., [<reflink idref="bib23" id="ref30">23</reflink>]; Halliday & Hasan, [<reflink idref="bib26" id="ref31">26</reflink>]; Hoey, [<reflink idref="bib30" id="ref32">30</reflink>]).</p> <p>Many researchers have investigated text cohesion with multiple dimensions of features in both English and Chinese. Halliday and Hasan ([<reflink idref="bib26" id="ref33">26</reflink>], [<reflink idref="bib27" id="ref34">27</reflink>]) were among the pioneers who systematically examined cohesion in English texts. They identified various cohesion devices that contribute to the structural connections of different elements in a text, including grammatical and lexical cohesive devices (Halliday & Hasan, [<reflink idref="bib26" id="ref35">26</reflink>], [<reflink idref="bib27" id="ref36">27</reflink>]). Specifically, grammatical cohesive devices are the specific grammatical elements that contribute to the structural connection of different segments in a text. Grammatical cohesion can be typically achieved by: (<reflink idref="bib1" id="ref37">1</reflink>) reference, such as pronouns, demonstrative determiners, definite articles, and comparatives, to refer back to something previously mentioned in the text or to point forward to something that will be mentioned (Halliday & Hasan, [<reflink idref="bib27" id="ref38">27</reflink>], p. 82; Li, [<reflink idref="bib40" id="ref39">40</reflink>], p. 12); (<reflink idref="bib2" id="ref40">2</reflink>) substitution, which involves replacing a word or phrase to avoid repetition and improve text flow; (<reflink idref="bib3" id="ref41">3</reflink>) ellipsis, which entails the omission of words or phrases that are easily understood in the context (Halliday & Hasan, [<reflink idref="bib27" id="ref42">27</reflink>], pp. 74–75); and (<reflink idref="bib4" id="ref43">4</reflink>) conjunction, which refers to using conjunctions (and, but, however, therefore, etc.) that link clauses or sentences to show additive, adversative, causal and temporal relationships (Halliday & Hasan, [<reflink idref="bib26" id="ref44">26</reflink>], p. 242–243). The second category, lexical cohesion, is more related to the semantic relations of lexical items in a text, usually realized through lexical reiteration or collocation (Halliday & Hasan, [<reflink idref="bib26" id="ref45">26</reflink>]). Specifically, reiteration involves using the same lexical item in different forms or related terms with varied sense relationships, such as synonyms and antonyms, to reinforce themes (Halliday & Hasan, [<reflink idref="bib26" id="ref46">26</reflink>]). Collocation refers to the frequent pairing of words such as <emph>boys</emph> and <emph>girls</emph>, <emph>red</emph> and <emph>green</emph>, which naturally co-occur in language (Halliday & Hasan, [<reflink idref="bib26" id="ref47">26</reflink>], pp. 284–285). In their later work, Halliday and Hasan ([<reflink idref="bib27" id="ref48">27</reflink>]) further refined the concept of lexical cohesion into two categories: general relations and instantial relations. General relations include meronymy, adding a layer of complexity by incorporating part-whole relationships into reiteration. Instantial relations detail how specific lexical choices operationalize cohesion within the text. This includes equivalence (using different terms to refer to the same concept or entity), naming (using a consistent name or names to refer to a specific entity or related entities throughout the text), and semblance (using semantically related but not necessarily synonymous words). The detailed categorization of text cohesion examples is shown in Figs. 1 and 2. The detailed categorization of text cohesion by Halliday and Hasan ([<reflink idref="bib26" id="ref49">26</reflink>], [<reflink idref="bib27" id="ref50">27</reflink>]) lays the foundation for future empirical analysis and theory development (Li, [<reflink idref="bib40" id="ref51">40</reflink>]).</p> <p>Graph: Fig. 1 Categories of grammatical cohesion (Halliday & Hasan, [<reflink idref="bib26" id="ref52">26</reflink>], [<reflink idref="bib27" id="ref53">27</reflink>])</p> <p>Graph: Fig. 2 Categories of lexical cohesion (Halliday & Hasan, [<reflink idref="bib26" id="ref54">26</reflink>], [<reflink idref="bib27" id="ref55">27</reflink>])</p> <p>Notably, the theories of cohesion developed by Halliday and Hasan ([<reflink idref="bib26" id="ref56">26</reflink>], [<reflink idref="bib27" id="ref57">27</reflink>]) was meaningfully expanded by Hoey ([<reflink idref="bib30" id="ref58">30</reflink>]), who emphasized analyzing text cohesion from a discoursal perspective. He introduced the concept of "links", which extends beyond the sentence level to describe the connections between different parts of a text. These links are established through various lexical and grammatical cohesive devices. For example, there is a link of complex lexical repetition between "writing" and "writer" in a text. A "bond", which is a significant relationship between two sentences or segments in a text, is formed when the number of links between these two sentences surpass a threshold. While a higher frequency of such links can indicate a stronger bond, the threshold number of links for what constitutes a bond may vary depending on the nature of the text and the impact of these links on the text's coherence (Hoey, [<reflink idref="bib30" id="ref59">30</reflink>]). By assessing the number of bonds a sentence has with others, central and marginal sentences can be distinguished within a text. Central sentences share bonds with many other sentences and are therefore crucial for conveying the main meaning of the text. In contrast, marginal sentences, which have fewer bonds, and do not significantly contributing to the main theme. Hoey's link-bond methods and the ideas of central and marginal sentences shifted the focus of text cohesion features from isolated grammatical or lexical items to larger text structures and their functions within discourse.</p> <hd id="AN0187667272-6">Measurements of Chinese text cohesion</hd> <p>The Chinese text cohesion theories are primarily based upon the English ones. Building on the English discourse cohesion theories, Liao ([<reflink idref="bib41" id="ref60">41</reflink>], [<reflink idref="bib42" id="ref61">42</reflink>]) was among the first to introduce the concept of text cohesion in Chinese. He categorized Chinese connectives into temporal connectives and logical connectives (Liao, [<reflink idref="bib41" id="ref62">41</reflink>]). Temporal connectives, such as "qǐ chū" [at the beginning], "zuì hòu" [finally], and "xiān qián" [previously], express the relative timing of two or more events, states, or processes, or different stages of a single event. Logical connectives are used to express various relationships between events, including sequential (e.g., "qí yī...qí èr..." [for one thing...for another...], "zài zhě" [in addition]), adversative (e.g., "rán ér" [however], "qǐ liào" [unexpectedly]), and transitional relations (e.g., "shùn biàn yī tí" [by the way], "zhì yú" [in terms of]). Moreover, he explored referential identity and ellipses in Chinese (Liao, [<reflink idref="bib42" id="ref63">42</reflink>]). Subsequent to Liao ([<reflink idref="bib41" id="ref64">41</reflink>], [<reflink idref="bib42" id="ref65">42</reflink>]), Hu ([<reflink idref="bib33" id="ref66">33</reflink>]) proposed detailed categories of Chinese text cohesion features, which are comparable to the English cohesive features proposed by Halliday and Hasan ([<reflink idref="bib26" id="ref67">26</reflink>]). The categories of reference, structural cohesion, and logical connection proposed by Hu ([<reflink idref="bib33" id="ref68">33</reflink>]) are similar to what Halliday and Hasan ([<reflink idref="bib26" id="ref69">26</reflink>], [<reflink idref="bib27" id="ref70">27</reflink>]) referred to as English grammatical cohesion, whereas the lexical cohesion category corresponds to English lexical cohesion. He also proposed the analysis of topical cohesion in Chinese by analyzing theme and rheme under the framework of Systemic Functional Linguistics. Theme refers to the element in a clause that serves as the departure point of the message, which is typically represented by the topic in an unmarked Chinese sentence. Rheme is the part of the clause that constitutes the rest of the message about the theme. He then further analyzed how the thematic progression between two clauses could contribute to text cohesion by: (<reflink idref="bib1" id="ref71">1</reflink>) repeating the theme of the previous clause; (<reflink idref="bib2" id="ref72">2</reflink>) developing a new theme from the rheme of the previous clause; (<reflink idref="bib3" id="ref73">3</reflink>) creating a new theme from both the theme and rheme of the previous clause. Hu's introduction of theme and rheme structure is important for us to understand the concept of topic chains in Chinese, which is a notable distinction between English and Chinese in maintaining text cohesion (Wang et al., [<reflink idref="bib67" id="ref74">67</reflink>]).</p> <p>According to Sun ([<reflink idref="bib62" id="ref75">62</reflink>]), a topic chain typically comprises multiple clauses that are unified by a shared theme or subject. It is possible for a single paragraph to contain multiple topic chains, and conversely, a topic chain can extend across more than one paragraph. Within the same topic chain, even in the absence of explicit cohesive devices, textual cohesion can still be maintained. In this pattern, a topic is introduced at the beginning of the topic chain and subsequently implied through zero anaphora, i.e., a type of reference where the subject or object of a sentence can be omitted in subsequent sentences when it is clearly understandable from the context (Hu, [<reflink idref="bib33" id="ref76">33</reflink>]). As the chain progresses, the subsequent clause might even introduce a different, yet related, and explicitly stated topic, without interrupting the semantic continuity of the chain (Lu & Wu, [<reflink idref="bib46" id="ref77">46</reflink>]). In Example 1 below, the main topic "Like yellow DE people" is introduced at the beginning and then referred to through various means (e.g., zero anaphora, "This type people," and "They") throughout the sentence. This structure maintains semantic continuity while allowing for rich expressions of the topic, and therefore enhances the cohesion of the text. This approach avoids redundancy, improves the diversity of the semantic categories of topics, and maintains Chinese text cohesion without repeatedly mentioning the original topic. These contributions of previous scholars offer valuable insights into how cohesion is achieved and maintained in both English and Chinese texts.</p> <hd id="AN0187667272-7">Example 1</hd> <p> <bold>喜欢 黄色 的 人</bold> 一般 都 很 积极、乐观, ∅ 觉得 生活 很 容易, 不 会 有 大 问题。 <bold>这 种 人</bold> 从来 都 不 为 生活 和 工作 担心。<bold>他们</bold> 不 会 太 懒, ∅ 工作 可能 非常 努力, 但 ∅ 常常 不能 坚持 很 长 时间.</p> <p> <bold>Like yellow DE people</bold> generally all very positive, optimistic, ∅ feel life very easy, won't have big problem. <bold>This type people</bold> always all not for life and work worry. <bold>They</bold> won't too lazy, ∅ work possible very hard, but often not can persist very long time.</p> <p>(People who like yellow are generally very positive and optimistic. They find life easy and do not encounter major problems. This type of person never worries about life or work. They are not lazy and may work very hard, but they often cannot sustain their efforts for long periods.)</p> <p>The evolution of technology has led to the creation of numerous computational tools for automatic text cohesion analysis. One of the most popular tools for English texts is Coh-Metrix, which features three categories of cohesion measurements: co-referential cohesion, latent semantic analysis (LSA), and various types of connectives. A more recently developed tool is the Tool for the Automatic Analysis of Cohesion (TAACO), which, after two updates, now includes 194 indices covering local and global cohesion features, as well as semantic similarity features (Crossley et al., [<reflink idref="bib14" id="ref78">14</reflink>], [<reflink idref="bib15" id="ref79">15</reflink>], [<reflink idref="bib16" id="ref80">16</reflink>]). Some tools for automatic analysis of Chinese text cohesion have also been developed, such as Chinese Coh-Metrix (Chang, [<reflink idref="bib9" id="ref81">9</reflink>]), the Chinese Readability Index Explorer (CRIE) (Sung et al., [<reflink idref="bib63" id="ref82">63</reflink>]), and the Common Text Analysis Platform (CTAP) (Cui et al., [<reflink idref="bib19" id="ref83">19</reflink>]). However, these tools were developed from the original English tools and therefore neglect the unique characteristics of Chinese, where sentences without explicit cohesive devices can remain cohesive through relevant topics. Moreover, the validity of most of these tools is yet to be confirmed. More recently, Peng, Hu, and Wu ([<reflink idref="bib53" id="ref84">53</reflink>]) developed the NLP-based L2C-Cohesion tool for automatic analysis of Chinese text cohesion. This tool incorporates a variety of text cohesion measures. The selection of lexical and grammatical cohesion features is grounded in established English text cohesion theories (Halliday & Hasan, [<reflink idref="bib26" id="ref85">26</reflink>], [<reflink idref="bib27" id="ref86">27</reflink>]; Hoey, [<reflink idref="bib30" id="ref87">30</reflink>]). Distinguishing itself from previous tools, it also considers the unique characteristics of Chinese to add another dimension of topical cohesion as suggested by Hu ([<reflink idref="bib33" id="ref88">33</reflink>]). Consequently, L2C-Cohesion analyzes 27 cohesion features spanning lexical, grammatical, and topical dimensions, offering a robust and comprehensive solution for assessing text cohesion in Chinese. The effectiveness of this tool has been further validated by empirical studies (Peng et al., [<reflink idref="bib53" id="ref89">53</reflink>]).</p> <hd id="AN0187667272-8">Variation of text cohesion</hd> <p>Previous studies emphasize the importance of adapting lexical and syntactic features of reading materials to accommodate various text levels and text genres (Jin et al., [<reflink idref="bib35" id="ref90">35</reflink>]; Su et al., [<reflink idref="bib61" id="ref91">61</reflink>]; Mesmer et al., [<reflink idref="bib49" id="ref92">49</reflink>]). However, an excessive focus on these features during adaptation can result in disconnected ideas, thereby disrupting text cohesion (Crossley et al., [<reflink idref="bib15" id="ref93">15</reflink>]). In actual teaching scenarios, educators often intuitively incorporate modifications to simplify the text during the adaptation process, with a keen consideration of text cohesion (Rets et al., [<reflink idref="bib57" id="ref94">57</reflink>]). This practice aligns with the established research indicating text cohesion as a vital factor for reading comprehension (Kleijn et al., [<reflink idref="bib37" id="ref95">37</reflink>]). High cohesion texts employ cohesive devices to signal various relationships between text segments and thereby can reduce the inferential burden of readers in comprehending the text (Kleijn et al., [<reflink idref="bib37" id="ref96">37</reflink>]; O'Reilly & McNamara, [<reflink idref="bib51" id="ref97">51</reflink>]). Conversely, low cohesion texts require readers to engage more deeply, invoking prior knowledge and making effortful inferences, thereby presenting greater challenges (O'Reilly & McNamara, [<reflink idref="bib51" id="ref98">51</reflink>]). Hence, text cohesion stands as a key feature that educators consider in reading material development (Bilki & Plakans, [<reflink idref="bib4" id="ref99">4</reflink>]; Rets et al., [<reflink idref="bib57" id="ref100">57</reflink>]), i.e., the validity of text cohesion in reading material development has been confirmed. However, questions remain about how cohesion features vary across different text levels and genres, particularly in Chinese reading materials. To address these questions and build upon the existing research, this study aims to investigate the relationship among the three dimensions of cohesion features (lexical, grammatical, and topical), text level, and text genre in Chinese reading materials. In the following sections, we will review previous studies on the variation of text cohesion and, based on these theoretical foundations and research findings, formulate several hypotheses to guide our investigation.</p> <p>Multiple studies have already confirmed the variation of text cohesion features across text levels and text genres. A significant body of research has focused on investigating the developmental trajectories of text cohesion among various learning levels. Within this research strand, particular attention has been paid to the variation of lexical cohesion features. For instance, Graesser et al. ([<reflink idref="bib21" id="ref101">21</reflink>]) utilized Coh-Metrix and identified variations of deep cohesion features and referential cohesion features, across different text levels in English reading materials. Similarly, Plakins & Bilki ([<reflink idref="bib54" id="ref102">54</reflink>]) found that the overlap of content words between adjacent sentences, a key feature of lexical cohesion, increases with text level in ESL reading materials. Furthermore, Peng et al. ([<reflink idref="bib53" id="ref103">53</reflink>]) observed a positive correlation between lexical cohesion and Chinese L2 proficiency. Crossley et al. ([<reflink idref="bib17" id="ref104">17</reflink>]) observed increased semantic similarity along the L2 learning trajectories. Based on these findings, the following hypothesis has been formulated for this study.</p> <hd id="AN0187667272-9">H1</hd> <p>Lexical cohesion features might increase as text level rises.</p> <p>In addition to lexical cohesion features, studies have also explored the variation of grammatical cohesion features among different text levels. For example, Berendes et al. ([<reflink idref="bib1" id="ref105">1</reflink>]) found that adversative and concessive connectors and third-person personal pronoun cohesion features played a role in differentiating geography reading texts of different grade levels. Sung et al. ([<reflink idref="bib63" id="ref106">63</reflink>]) explored the combined effect of 30 linguistic features, including eight grammatical cohesion features, on levelling the readability of Chinese L2 reading materials. Unfortunately, they did not specify how reading materials at different levels varied in text cohesion. Some other studies have suggested that more proficient learners may rely less on explicit cohesive features and focus more on integrating abstract concepts and background information in their text processing. Plakins & Bilki ([<reflink idref="bib54" id="ref107">54</reflink>]) noted a decrease in the use of causal conjunctions as the grade level of ESL reading textbooks increased. Similar trends have been observed in writing. Yang and Sun ([<reflink idref="bib69" id="ref108">69</reflink>]) found that senior EFL learners used fewer personal and demonstrative pronouns than their sophomore counterparts. Likewise, research by Crowhurst ([<reflink idref="bib18" id="ref109">18</reflink>]) & Crossley, et al. ([<reflink idref="bib13" id="ref110">13</reflink>]) documented a reduction in conjunction usage as students progressed academically. Despite this decline in explicit cohesive devices, there was evidence of a more sophisticated employment of a variety of grammatical cohesive features as students' language proficiency developed. Based on these observations, the following hypotheses are proposed:</p> <hd id="AN0187667272-10">H2a</hd> <p>Density and ratio of grammatical cohesive devices might decrease as text level arises.</p> <hd id="AN0187667272-11">H2b</hd> <p>Diversity of grammatical cohesive devices might increase as text level arises.</p> <p>Moreover, few studies have explored the variation of topical cohesion in the written output of L2 Chinese learners. These studies have confirmed the relationship between subject density, diversity, and semantic categories with topical cohesion of texts. For example, more advanced learners can skillfully employ null subjects, i.e., subjects that is implied rather than overtly stated, to ensure continuity and minimize redundancy within topic chains, therefore produce texts with lower subject density (Lu & Wu, [<reflink idref="bib46" id="ref111">46</reflink>]; Peng et al., [<reflink idref="bib53" id="ref112">53</reflink>]). In addition, it has been observed that students of higher proficiency levels use subjects from diverse semantic classes and avoid repeating subjects in adjacent sentences (Peng et al., [<reflink idref="bib53" id="ref113">53</reflink>]; Zhou et al., [<reflink idref="bib72" id="ref114">72</reflink>]). Conversely, less proficient learners frequently rely on overt subjects, often pronouns, to sustain topic continuity (Peng et al., [<reflink idref="bib53" id="ref115">53</reflink>]). Based on the previous research, the following hypotheses are formed:</p> <hd id="AN0187667272-12">H3a</hd> <p>Subject density, ratio of pronoun subjects, and local and global subject cohesion might decrease as text level arises.</p> <hd id="AN0187667272-13">H3b</hd> <p>Subject diversity, diversity and ratio of noun subjects, and number and density of subject semantic classes might increase as text level arises.</p> <p>In addition to text level, the reliability of text cohesion in differentiating reading materials of different genres has also been confirmed. For example, in a study by McNamara et al., ([<reflink idref="bib47" id="ref116">47</reflink>]), narrative texts showed a lower level of referential cohesion and contained more additive and temporal connectives, whereas expository texts contained more causal and clarification connectives. Such a trend was also observed by Graesser et al. ([<reflink idref="bib21" id="ref117">21</reflink>]). They speculated that the higher referential cohesion level in expository texts might be related to authors' attempts to balance the inherent difficulty of scientific topics and ensure the comprehensibility of texts for readers. Such genre-specific variations in text cohesion features have also been noted in the writings of L2 Chinese learners. For instance, it is found that Chinese learners produced a larger number of topic chains and higher frequency of null subjects in personal narratives compared with argumentative texts (Wu, [<reflink idref="bib68" id="ref118">68</reflink>]). Similarly, Shi and Chen ([<reflink idref="bib59" id="ref119">59</reflink>]) observed that L2 Chinese learners tend to focus on the same topic when writing argumentative texts, whereas narrative compositions involve a broader array of topics. This suggests that the topic density in argumentative texts may be lower than in narratives. These observations lead us to propose Hypothesis 4:</p> <hd id="AN0187667272-14">H4</hd> <p>Lexical, grammatical, and topical cohesion features might vary by genre.</p> <p>In summary, prior research has affirmed the validity and reliability of text cohesion features in differentiating texts of for various levels and genres. Yet, the scope of these studies has often been constrained to a selected set of cohesion features, primarily in English reading materials. This limitation may be partially attributed to the lack of reliable automated tools for analyzing cohesion features in Chinese texts. With the advent of automatic analysis tools (e.g., Peng et al., [<reflink idref="bib53" id="ref120">53</reflink>]), a comprehensive investigation of text cohesion in Chinese reading materials across different levels and genres becomes feasible. Such investigation can improve our understanding of text cohesion variation in Chinese reading materials, in relation to both text level and genre, and offer crucial insights for Chinese reading material development.</p> <hd id="AN0187667272-15">The present study</hd> <p>Previous research has highlighted the importance of text cohesion in L2 reading materials of different levels as well as the role different genres plays in text cohesion. To guide teachers in developing reading materials appropriately for students, a systematic exploration of text cohesion features in L2 Chinese reading materials across various levels and genres is essential. Therefore, the present study aims to address the following research questions:</p> <p></p> <ulist> <item> Do text levels (Beginning, Intermediate, Advanced) influence cohesion features of L2 Chinese reading texts?</item> <p></p> <item> Do text genres ((Narrative, Argumentative, Expository) influence cohesion features of L2 Chinese reading texts?</item> <p></p> <item> In what ways, if any, does the interaction between text level and genre influence the cohesion features of texts?</item> </ulist> <hd id="AN0187667272-16">Methodology</hd> <p></p> <hd id="AN0187667272-17">Corpus</hd> <p>In this study, a corpus of L2 Chinese reading materials was compiled for analysis. To compile a balanced corpus, each proficiency level and each genre was represented by 150 texts respectively, amounting to a total of 450 texts (see Table 1). All the texts were randomly selected from the Corpus of Chinese Teaching Textbook, which comprises an extensive collection of 874 L2 Chinese textbooks that house 16,643 reading texts (Zhang & Jin, [<reflink idref="bib70" id="ref121">70</reflink>]). These texts are widely recognized in the international L2 Chinese community and offer a comprehensive representation of commonly used L2 Chinese textbooks (Zhang & Jin, [<reflink idref="bib70" id="ref122">70</reflink>]; Zhou et al., [<reflink idref="bib73" id="ref123">73</reflink>]). Two experienced Chinese teachers, each boasting over 10 years of L2 Chinese teaching experience, classified the texts according to their level and genre. Text proficiency levels were classified into the beginning level, the intermediate level, and the advanced level, as defined by the Chinese Proficiency Standards for International Chinese language Education (Ministry of Education, [<reflink idref="bib50" id="ref124">50</reflink>]). Genres were categorized into narrative, argumentative, and expository based on the operational definitions given in a classic Chinese genre reference book, i.e., The Dictionary of Text Genre (Jin, [<reflink idref="bib36" id="ref125">36</reflink>]). A consistency level of 94.5% was achieved in classifying the texts, and any discrepancies were discussed and resolved.</p> <p>Table 1 Detailed information about the corpus of L2 Chinese textbooks</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" /><th align="left"><p>Narrative texts</p></th><th align="left"><p>Argumentative texts</p></th><th align="left"><p>Expository texts</p></th><th align="left"><p>Total</p></th></tr></thead><tbody><tr><td align="left"><p>Beginning</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>150</p></td></tr><tr><td align="left"><p>Intermediate</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>150</p></td></tr><tr><td align="left"><p>Advanced</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>50</p></td><td char="." align="char"><p>150</p></td></tr><tr><td align="left"><p>Total</p></td><td char="." align="char"><p>150</p></td><td char="." align="char"><p>150</p></td><td char="." align="char"><p>150</p></td><td char="." align="char"><p>450</p></td></tr></tbody></table> </ephtml> </p> <hd id="AN0187667272-18">Analysis of cohesion features</hd> <p>Research on both L1 and L2 learning contexts has found that students are able to produce more cohesive texts as their grade level rises (Crowhurst, [<reflink idref="bib18" id="ref126">18</reflink>]; Yang & Sun, [<reflink idref="bib69" id="ref127">69</reflink>]; Peng et al., [<reflink idref="bib53" id="ref128">53</reflink>]). However, the use of different cohesive devices might vary with grade levels (Crossley et al., [<reflink idref="bib16" id="ref129">16</reflink>]). In this study, to examine the variation of text cohesion features by text level and genre in L2 Chinese reading materials, we employed L2C-Cohesion to automatically analyze the cohesion features of the text (Peng et al., [<reflink idref="bib53" id="ref130">53</reflink>]). This tool provides a comprehensive examination of twenty-seven cohesion features that include three dimensions: the lexical, grammatical, and topical cohesion (see Table 2).</p> <p>Table 2 Text cohesion dimensions and features</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Dimension</p></th><th align="left"><p>Feature</p></th><th align="left"><p>Label</p></th><th align="left"><p>Hypothesized relationship with text level</p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical cohesion</p></td><td align="left"><p>Local lexical cohesion</p></td><td align="left"><p>local_lexical_cohesion</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Global lexical cohesion</p></td><td align="left"><p>global_lexical_coheison</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Local noun cohesion</p></td><td align="left"><p>local_noun_cohesion</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Global noun cohesion</p></td><td align="left"><p>global_noun_cohesion</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Number of central sentences</p></td><td align="left"><p>central_sent_num</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Ratio of central sentences</p></td><td align="left"><p>central_sent_ratio</p></td><td align="left"><p> + </p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical cohesion</p></td><td align="left"><p>Pronoun density</p></td><td align="left"><p>pron_density</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Pronoun-noun ratio</p></td><td align="left"><p>pron_noun_ratio</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Personal pronoun-noun ratio</p></td><td align="left"><p>personal-pron_noun_ratio</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Ratio between personal pronouns and pronouns</p></td><td align="left"><p>personal-pron_pron_ratio</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Diversity of personal pronouns</p></td><td align="left"><p>personal-pron_TTR</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Diversity of pronouns</p></td><td align="left"><p>pron_TTR</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Number of pronouns per sentence</p></td><td align="left"><p>pron_num_per_sent</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Number of personal pronouns per centence</p></td><td align="left"><p>personal-pron_num_per_sent</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Conjunction density</p></td><td align="left"><p>conj_density</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Conjunction-noun ratio</p></td><td align="left"><p>conj_noun_ratio</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Number of conjunctions per sentence</p></td><td align="left"><p>conj_num_per_sent</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>conjunction diversity</p></td><td align="left"><p>conj_TTR</p></td><td align="left"><p> + </p></td></tr><tr><td align="left" rowspan="9"><p>Topical cohesion</p></td><td align="left"><p>Subject density</p></td><td align="left"><p>subj_density</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Subject diversity</p></td><td align="left"><p>subj_TTR</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Noun subject diversity</p></td><td align="left"><p>subj_noun_TTR</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Ratio of pronoun subject</p></td><td align="left"><p>subj_pron_ratio</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Ratio of noun subjects</p></td><td align="left"><p>subj_noun_ratio</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Local subject cohesion</p></td><td align="left"><p>local_subj_cohesion</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Global subject cohesion</p></td><td align="left"><p>global_subj_cohesion</p></td><td align="left"><p>−</p></td></tr><tr><td align="left"><p>Number of subject semantic class</p></td><td align="left"><p>subj_semantic_num</p></td><td align="left"><p> + </p></td></tr><tr><td align="left"><p>Density of subject semantic class</p></td><td align="left"><p>subj_semantic_density</p></td><td align="left"><p> + </p></td></tr></tbody></table> </ephtml> </p> <p> <sups>*</sups> " + " signifies a hypothesized positive relationship, indicating that the feature increases as the text level rises. Conversely, a "−"indicates a hypothesized negative relationship, where the feature decreases with the text level</p> <p>The first category, lexical cohesion, refers to the cohesive effect through the selection of specific words, which is mainly achieved by the word reiteration and collocation (Halliday & Hasan, [<reflink idref="bib26" id="ref131">26</reflink>]). Hoey ([<reflink idref="bib30" id="ref132">30</reflink>]) analyzes cohesion beyond individual sentences by focusing on central sentences within a text. To operationalize these theories for automatic analysis, we employed the Synonyms Cilin (extended version) tool (available at https://<ulink href="http://www.ltp-cloud.com/download#down%5fcilin),">www.ltp-cloud.com/download#down%5fcilin),</ulink> which categorizes and classifies a vast majority of Chinese synonyms and hyponyms into a structured five-layer semantic tree, to extract lexical cohesion relationships. The reliability and validity of Synonyms Cilin in analyzing semantic relationships have been well-documented in previous studies. As a first step of analyzing lexical cohesion, we first calculate the cohesion coefficient, which quantifies the strength of semantic connections between words based on the semantic tree (see Peng et al., [<reflink idref="bib53" id="ref133">53</reflink>] for the calculation formula). The higher the coefficient, the tighter the connection. For instance, the coefficient between "lǚ yóu" [travel] and "chū yóu" [tour] is 5, compared to 2 between "lǚ yóu" [travel] and "chī fàn" [eat].</p> <p>Then we calculate the average inter-sentence lexical cohesion coefficient from both local and global perspectives, as well as central sentences. Local cohesion is calculated only considering adjacent sentence pairs, while global cohesion considers any two sentences in the text. In addition, following the studies by Hoey ([<reflink idref="bib30" id="ref134">30</reflink>]) and Liu ([<reflink idref="bib43" id="ref135">43</reflink>]), after calculating the cohesion coefficients for all sentences in the discourse, those sentences whose average cohesion coefficient with other sentences exceeds 80 are designated as central sentences. The number and proportion of central sentences are examined as a third dimension of lexical cohesion features.</p> <hd id="AN0187667272-19">Grammatical cohesion</hd> <p>The second category, grammatical cohesion, primarily involves reference, substitution, ellipsis, and conjunction, as defined by Halliday and Hasan ([<reflink idref="bib26" id="ref136">26</reflink>]). Hu ([<reflink idref="bib33" id="ref137">33</reflink>]) emphasized the crucial role of pronouns and conjunctions in achieving this type of cohesion. Accordingly, this study focuses on analyzing pronouns and conjunctions as key features of grammatical cohesion in Chinese. In our analysis, we examine several aspects of pronouns and conjunctions, including the density and diversity of pronouns and conjunctions in a text, their ratio, and their proportion in a sentence (Peng et al., [<reflink idref="bib53" id="ref138">53</reflink>]). Substitution and ellipsis are excluded from the analysis given the absence of established formal markers and mature analytical tools for these two types of grammatical cohesive devices, as well as their infrequent use in both English and Chinese (Cameron et al., [<reflink idref="bib6" id="ref139">6</reflink>]; Peng et al., [<reflink idref="bib53" id="ref140">53</reflink>]).</p> <hd id="AN0187667272-20">Topical cohesion</hd> <p>The third category, i.e., topical cohesion, which is often achieved through the effective use of topic chains in Chinese, plays a crucial role in maintaining a coherent flow of information within a text (Li, [<reflink idref="bib39" id="ref141">39</reflink>]). Even though adjacent clauses lack explicit cohesive devices, the cohesion of the discourse remains unaffected due to their association with the same topic, rather than the presence of explicit cohesive devices (Peng et al., [<reflink idref="bib53" id="ref142">53</reflink>]). In Chinese grammar, the subject of a sentence is typically regarded as its topic (Zhu, [<reflink idref="bib74" id="ref143">74</reflink>], p.17; Zhao, [<reflink idref="bib71" id="ref144">71</reflink>], p.45). In analyzing topical cohesion in this study, nine features related to the subjects were calculated, including the density and diversity of subjects, the ratio of different subjects, global and local subject cohesion, and the number and density of subject semantic classes (Peng et al., [<reflink idref="bib53" id="ref145">53</reflink>]). Like lexical cohesion, the lexical-semantic information of subjects is retrieved from Synonyms Cilin. We group subjects into general semantic categories based on their top-level labels in the Cilin. The top-level labels represent broad conceptual domains. For example, "lǚ yóu" [travel], with the label H-j-48-B-01 in the Cilin, belongs to the H category. Another word, "xiōng zhǎng" [elder brother], with the label A-a-03-A-02, falls under the A category.</p> <hd id="AN0187667272-21">Data analysis</hd> <p>To answer the research questions, we first presented descriptive statistics of the cohesion features of L2 Chinese texts across different proficiency levels and genres. Before the test, the assumptions of MANOVA were examined. Some of the cohesion features did not meet the assumption of normal distribution or homogeneous covariance matrices. However, past research confirmed that MANOVA remains robust even when the assumption of normality is violated (Finch, [<reflink idref="bib20" id="ref146">20</reflink>]). It also demonstrates resilience against heterogeneous covariance matrices when sample sizes are equal (Hakstian et al., [<reflink idref="bib25" id="ref147">25</reflink>]; Oslon, [<reflink idref="bib52" id="ref148">52</reflink>]). Therefore, a two-way MANOVA was conducted to examine the influence of text level and genre on the text cohesion of L2 Chinese reading materials, with text level and text genre serving as the independent variables, and the 27 cohesion features being the dependent variables. The data analyses were performed using SPSS 26 and visualizations were created using Matplotlib in Python.</p> <hd id="AN0187667272-22">Results</hd> <p>A two-way MANOVA was conducted to examine the effects of text level and genre on text cohesion of L2 Chinese reading materials. Since the assumption of homogeneous covariance was violated, Pillai's Trace was used as the test statistics for the multivariate tests. Results for multivariate tests show that the main effect of both text level (Pillai's Trace = 0.80, <emph>F</emph> (<reflink idref="bib2" id="ref149">2</reflink>, 441) = 10.29, <emph>p</emph> < 0.001) and text genre (Pillai's Trace = 0.80, <emph>F</emph> (<reflink idref="bib2" id="ref150">2</reflink>, 441) = 10.23, <emph>p</emph> < 0.001) was significant. Moreover, the interaction between text level and genre also demonstrates a significant effect (Pillai's Trace = 0.58,<emph> F</emph> (<reflink idref="bib2" id="ref151">2</reflink>, 441) = 2.60, <emph>p</emph> < 0.001). To pinpoint the specific significant disparities among the three text levels and three text genres on cohesion features, Tukey's HSD post hoc tests were performed. Detailed results for the between-level and between-genre effects are presented in the following sections.</p> <hd id="AN0187667272-23">Text cohesion of L2 Chinese reading materials among different text levels</hd> <p>Table 3 summarizes descriptive statistics for text cohesion features in L2 Chinese reading texts spanning three text levels.</p> <p>Table 3 Descriptive statistics for cohesion features by text level<sups>*</sups></p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="2"><p>Dimension</p></th><th align="left" rowspan="2"><p>Measurements</p></th><th align="left" colspan="2"><p>Beginning</p></th><th align="left" colspan="2"><p>Intermediate</p></th><th align="left" colspan="2"><p>Advanced</p></th></tr><tr><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical cohesion</p></td><td align="left"><p>Local_lexical_cohesion</p></td><td char="." align="char"><p>162.03</p></td><td char="." align="char"><p>95.28</p></td><td char="." align="char"><p>251.08</p></td><td char="." align="char"><p>111.65</p></td><td char="." align="char"><p><bold>294.01</bold></p></td><td char="." align="char"><p>115.56</p></td></tr><tr><td align="left"><p>Global_lexical_coheison</p></td><td char="." align="char"><p>191.45</p></td><td char="." align="char"><p>109.01</p></td><td char="." align="char"><p>278.80</p></td><td char="." align="char"><p>134.77</p></td><td char="." align="char"><p><bold>309.24</bold></p></td><td char="." align="char"><p>127.29</p></td></tr><tr><td align="left"><p>Local_noun_cohesion</p></td><td char="." align="char"><p>10.02</p></td><td char="." align="char"><p>7.67</p></td><td char="." align="char"><p>13.63</p></td><td char="." align="char"><p>9.08</p></td><td char="." align="char"><p><bold>18.88</bold></p></td><td char="." align="char"><p>13.72</p></td></tr><tr><td align="left"><p>Global_noun_cohesion</p></td><td char="." align="char"><p>11.17</p></td><td char="." align="char"><p>9.50</p></td><td char="." align="char"><p>13.58</p></td><td char="." align="char"><p>11.49</p></td><td char="." align="char"><p><bold>17.89</bold></p></td><td char="." align="char"><p>14.15</p></td></tr><tr><td align="left"><p>Central_sent_num</p></td><td char="." align="char"><p>7.35</p></td><td char="." align="char"><p>5.25</p></td><td char="." align="char"><p>18.66</p></td><td char="." align="char"><p>8.23</p></td><td char="." align="char"><p><bold>22.13</bold></p></td><td char="." align="char"><p>5.81</p></td></tr><tr><td align="left"><p>Central_sent_ratio</p></td><td char="." align="char"><p>0.70</p></td><td char="." align="char"><p>0.26</p></td><td char="." align="char"><p>0.86</p></td><td char="." align="char"><p>0.10</p></td><td char="." align="char"><p><bold>0.88</bold></p></td><td char="." align="char"><p>0.10</p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical cohesion</p></td><td align="left"><p>Pron_density</p></td><td char="." align="char"><p><bold>0.08</bold></p></td><td char="." align="char"><p>0.05</p></td><td char="." align="char"><p>0.07</p></td><td char="." align="char"><p>0.04</p></td><td char="." align="char"><p>0.06</p></td><td char="." align="char"><p>0.03</p></td></tr><tr><td align="left"><p>Pron_noun_ratio</p></td><td char="." align="char"><p><bold>0.42</bold></p></td><td char="." align="char"><p>0.31</p></td><td char="." align="char"><p>0.42</p></td><td char="." align="char"><p>0.32</p></td><td char="." align="char"><p>0.28</p></td><td char="." align="char"><p>0.19</p></td></tr><tr><td align="left"><p>Person-pron_noun_ratio</p></td><td char="." align="char"><p><bold>0.22</bold></p></td><td char="." align="char"><p>0.25</p></td><td char="." align="char"><p>0.21</p></td><td char="." align="char"><p>0.25</p></td><td char="." align="char"><p>0.13</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Person-pron_pron_ratio</p></td><td char="." align="char"><p>0.39</p></td><td char="." align="char"><p>0.24</p></td><td char="." align="char"><p><bold>0.40</bold></p></td><td char="." align="char"><p>0.21</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.21</p></td></tr><tr><td align="left"><p>Person-pron_TTR</p></td><td char="." align="char"><p><bold>0.32</bold></p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.31</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.29</p></td><td char="." align="char"><p>0.20</p></td></tr><tr><td align="left"><p>Pron_TTR</p></td><td char="." align="char"><p><bold>0.52</bold></p></td><td char="." align="char"><p>0.18</p></td><td char="." align="char"><p>0.50</p></td><td char="." align="char"><p>0.18</p></td><td char="." align="char"><p>0.51</p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left"><p>Pron_num_per_sent</p></td><td char="." align="char"><p>1.07</p></td><td char="." align="char"><p>0.64</p></td><td char="." align="char"><p><bold>1.34</bold></p></td><td char="." align="char"><p>0.62</p></td><td char="." align="char"><p>1.13</p></td><td char="." align="char"><p>0.57</p></td></tr><tr><td align="left"><p>Person-pron_num_per_sent</p></td><td char="." align="char"><p>0.53</p></td><td char="." align="char"><p>0.46</p></td><td char="." align="char"><p><bold>0.63</bold></p></td><td char="." align="char"><p>0.49</p></td><td char="." align="char"><p>0.50</p></td><td char="." align="char"><p>0.43</p></td></tr><tr><td align="left"><p>Conj_density</p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Conj_noun_ratio</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p><bold>0.17</bold></p></td><td char="." align="char"><p>0.09</p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.07</p></td></tr><tr><td align="left"><p>Conj_num_per_sent</p></td><td char="." align="char"><p>0.43</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.64</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.73</p></td><td char="." align="char"><p>0.40</p></td></tr><tr><td align="left"><p>Conj_TTR</p></td><td char="." align="char"><p>0.53</p></td><td char="." align="char"><p>0.27</p></td><td char="." align="char"><p><bold>0.64</bold></p></td><td char="." align="char"><p>0.13</p></td><td char="." align="char"><p>0.59</p></td><td char="." align="char"><p>0.15</p></td></tr><tr><td align="left" rowspan="9"><p>Topical cohesion</p></td><td align="left"><p>Subj_density</p></td><td char="." align="char"><p><bold>0.10</bold></p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Subj_TTR</p></td><td char="." align="char"><p>0.63</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.62</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p><bold>0.65</bold></p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Subj_noun_TTR</p></td><td char="." align="char"><p>0.61</p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.67</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p><bold>0.68</bold></p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Subj_pron_ratio</p></td><td char="." align="char"><p>0.27</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.27</p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.21</p></td><td char="." align="char"><p>0.14</p></td></tr><tr><td align="left"><p>Subj_noun_ratio</p></td><td char="." align="char"><p>0.46</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.50</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p><bold>0.56</bold></p></td><td char="." align="char"><p>0.15</p></td></tr><tr><td align="left"><p>Local_subj_cohesion</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.13</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Global_subj_cohesion</p></td><td char="." align="char"><p><bold>0.20</bold></p></td><td char="." align="char"><p>0.20</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.11</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Subj_semantic_num</p></td><td char="." align="char"><p>3.93</p></td><td char="." align="char"><p>1.45</p></td><td char="." align="char"><p>5.89</p></td><td char="." align="char"><p>1.56</p></td><td char="." align="char"><p><bold>6.82</bold></p></td><td char="." align="char"><p>1.35</p></td></tr><tr><td align="left"><p>Subj_semantic_density</p></td><td char="." align="char"><p>2.54</p></td><td char="." align="char"><p>1.03</p></td><td char="." align="char"><p>3.82</p></td><td char="." align="char"><p>1.33</p></td><td char="." align="char"><p><bold>4.26</bold></p></td><td char="." align="char"><p>1.18</p></td></tr></tbody></table> </ephtml> </p> <p> <sups>*</sups>The highest values for each measurement across the three genres are bolded</p> <p>The main effect of text level was significant for all the lexical cohesion measurements. Eight out of 12 grammatical cohesion features and seven out of nine topical cohesion features were significantly varied across text levels (see Table 4). Specifically, four features exhibit large effect size belongs to the lexical cohesion features, including the central sentence number (central_sent_num): <emph>F</emph> (<reflink idref="bib2" id="ref152">2</reflink>, 441) = 234.15, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.52; the central sentence ratio (central_sent_ratio): <emph>F</emph> (<reflink idref="bib2" id="ref153">2</reflink>, 441) = 56.85, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.21, the local lexical cohesion (local_lexical_cohesion): <emph>F</emph> (<reflink idref="bib2" id="ref154">2</reflink>, 441) = 59.66, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.21; and the global lexical cohesion (global_lexical_cohesion): <emph>F</emph> (<reflink idref="bib2" id="ref155">2</reflink>, 441) = 38.12, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.15. All four measurements exhibit a gradual increase from lower to higher level of texts. In addition, two topical features showing large effect size also gradually grow from lower to higher levels, including the subject semantic class number (subj_semantic_num): <emph>F</emph> (<reflink idref="bib2" id="ref156">2</reflink>, 441) = 157.79, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.42; and the subject semantic class density (subj_semantic_density): <emph>F</emph> (<reflink idref="bib2" id="ref157">2</reflink>, 441) = 86.66, <emph>p</emph> < 0.001, partial η<sups>2</sups> = 0.28. In comparison, eight grammatical cohesion features only exhibit small to medium effect sizes, indicating that text level might have a weaker influence on grammatical cohesion.</p> <p>Table 4 Main effects of text level on cohesion features</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Dimension</p></th><th align="left"><p>Measurements</p></th><th align="left"><p><italic>F</italic> (2, 441)</p></th><th align="left"><p><italic>p</italic> value<sup>a</sup></p></th><th align="left"><p>Partial Eta Squared (η<sup>2</sup>)<sup>b</sup></p></th><th align="left"><p>Post-hoc</p><p>tests<sup>c</sup></p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical cohesion</p></td><td align="left"><p>Local_lexical_cohesion</p></td><td char="." align="char"><p>59.66</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.21<sup>***</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr><tr><td align="left"><p>Global_lexical_coheison</p></td><td char="." align="char"><p>38.12</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.15<sup>***</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left"><p>Local_noun_cohesion</p></td><td char="." align="char"><p>31.70</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.13<sup>**</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr><tr><td align="left"><p>Global_noun_cohesion</p></td><td char="." align="char"><p>13.99</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.06<sup>**</sup></p></td><td align="left"><p>2 < 3, 1 < 3</p></td></tr><tr><td align="left"><p>Central_sent_num</p></td><td char="." align="char"><p>234.15</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.52<sup>***</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr><tr><td align="left"><p>Central_sent_ratio</p></td><td char="." align="char"><p>56.85</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.21<sup>***</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical cohesion</p></td><td align="left"><p>Pron_density</p></td><td char="." align="char"><p>19.22</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.08<sup>**</sup></p></td><td align="left"><p>2 > 3, 1 > 3</p></td></tr><tr><td align="left"><p>Pron_noun_ratio</p></td><td char="." align="char"><p>16.19</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>2 > 3, 1 > 3</p></td></tr><tr><td align="left"><p>Person-pron_noun_ratio</p></td><td char="." align="char"><p>12.40</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.05<sup>*</sup></p></td><td align="left"><p>2 > 3, 1 > 3</p></td></tr><tr><td align="left"><p>Person-pron_pron_ratio</p></td><td char="." align="char"><p>2.62</p></td><td char="." align="char"><p>0.07</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Person-pron_TTR</p></td><td char="." align="char"><p>1.03</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Pron_TTR</p></td><td char="." align="char"><p>1.18</p></td><td char="." align="char"><p>0.31</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Pron_num_per_sent</p></td><td char="." align="char"><p>11.14</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.05<sup>*</sup></p></td><td align="left"><p>1 < 2, 2 > 3</p></td></tr><tr><td align="left"><p>Person-pron_num_per_sent</p></td><td char="." align="char"><p>5.98</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td><td align="left"><p>1 < 2, 1 > 3</p></td></tr><tr><td align="left"><p>Conj_density</p></td><td char="." align="char"><p>7.93</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left"><p>Conj_noun_ratio</p></td><td char="." align="char"><p>1.03</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Conj_num_per_sent</p></td><td char="." align="char"><p>29.08</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.12<sup>**</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left"><p>Conj_TTR</p></td><td char="." align="char"><p>14.81</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.06<sup>**</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left" rowspan="9"><p>Topical cohesion</p></td><td align="left"><p>Subj_density</p></td><td char="." align="char"><p>24.54</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.10<sup>**</sup></p></td><td align="left"><p>1 > 2 > 3</p></td></tr><tr><td align="left"><p>Subj_TTR</p></td><td char="." align="char"><p>1.17</p></td><td char="." align="char"><p>0.31</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Subj_n_TTR</p></td><td char="." align="char"><p>12.16</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.05<sup>*</sup></p></td><td align="left"><p>1 < 2, 1 < 3</p></td></tr><tr><td align="left"><p>Subj_p_ratio</p></td><td char="." align="char"><p>9.60</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td><td align="left"><p>2 > 3, 1 > 3</p></td></tr><tr><td align="left"><p>Subj_n_ratio</p></td><td char="." align="char"><p>16.73</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr><tr><td align="left"><p>Local_subj_cohesion</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p>0.92</p></td><td char="." align="char"><p>0.00</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Global_subj_cohesion</p></td><td char="." align="char"><p>8.72</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td><td align="left"><p>1 > 2, 1 > 3</p></td></tr><tr><td align="left"><p>Subj_semantic_num</p></td><td char="." align="char"><p>157.79</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.42<sup>***</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr><tr><td align="left"><p>Subj_semantic_density</p></td><td char="." align="char"><p>86.66</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.28<sup>***</sup></p></td><td align="left"><p>1 < 2 < 3</p></td></tr></tbody></table> </ephtml> </p> <p> <sups>a</sups> <emph>p</emph>-values indicating significant findings are highlighted in bold <sups>b</sups>Asterisks denote the effect size: <sups>*</sups>represents a small effect size; <sups>**</sups>signifies a medium effect size; <sups>***</sups>indicates a large effect size <sups>c</sups>The column shows significant results in post-hoc tests. 1 represents beginning level texts, 2 represents intermediate level texts, and 3 represents advanced level texts</p> <p>The descriptive statistics for text cohesion by text level revealed an upward trend across all lexical cohesion measurements from lower to higher text levels. Significant differences were observed in two local lexical cohesion features (local_lexical_cohesion & local_noun_cohesion) and one central sentence feature (central_sent_num) across three levels. The feature of global lexical cohesion (global_lexical_cohesion) and central sentence ratio (central_sent_ratio) showed significant differences between the beginning and the intermediate text levels, while the global noun cohesion (global_noun_cohesion) exhibit significant differences between the intermediate and advanced levels. Given these findings, it can be concluded that overall lexical cohesion features increase as text level rises, which confirms Hypothesis 1 (H1).</p> <p>In terms of grammatical cohesion, none of the features exhibited significant differences across all three text levels (see Table 4). As shown in Table 3, the density and ratio of different types of pronouns showed a downward trend from lower to higher levels. Three features related to pronouns, i.e., the pronoun density (pron_density), the pronoun-noun ratio (pron_noun_ratio), and the personal pronoun-noun ratio (person-pron_noun_ratio), demonstrated significant differences between the intermediate and advanced levels, but not between the beginning and the intermediate levels. The results indicate that advanced level texts utilized fewer pronouns to improve text cohesion than intermediate or beginning level texts. As for measurements related to the use of conjunctions, the beginning level texts lagged behind the intermediate and advanced levels of texts in the conjunction density (conj_density), the conjunction number per sentence (conj_num_per_sent), and the conjunction diversity (conj_TTR). This implies an increase in both the frequency and diversity of conjunction usage with text level advancement. Example 2 demonstrates the contrast between texts from the beginning and advanced levels in the use of pronouns and conjunctions. The beginning level text employed more personal pronouns as subjects and fewer conjunctions. In contrast, the advanced level text exhibited increased use of zero anaphora and a greater variety of conjunctions. Only one grammatical cohesion feature, i.e., the pronoun number per sentence (pron_num_per_sent) demonstrated no significant results between the advanced and beginning level of texts. Therefore, both Hypothesis 2a and Hypothesis 2b are partially confirmed.</p> <hd id="AN0187667272-24">Example 2</hd> <p></p> <ulist> <item> Beginning level</item> <p></p> <item> <bold> 我 家 </bold> 有 三 口 人: 爸爸、妈妈 和 我。<bold>我</bold> 没有 哥哥、姐姐, ∅ 也 没有 弟弟、妹妹。<bold>我 爸爸</bold> 是 老师, <bold>妈妈</bold> 是 大夫。<bold>我</bold> 是 学生。</item> <p></p> <item> <bold> My family </bold> has three people: father, mother and me. <bold>I</bold> have no older brother, older sister, ∅ also have no younger brother, younger sister. <bold>My father</bold> is teacher, <bold>mother</bold> is doctor. <bold>I</bold> am student.</item> <p></p> <item> (There are three people in my family: my father, my mother, and me. I don't have any brothers or sisters. My father is a teacher, and my mother is a doctor. I am a student.)</item> <p></p> <item> Advanced level</item> <p></p> <item> <bold> 中国 的 饮食 文化 </bold> 历经 数 千 年, ∅ 始终 具有 魅力, 是 <uline>因为</uline><bold>它</bold><uline>不仅</uline> 民族 特性 鲜明, <uline>而且</uline> ∅ 善于 吸收 不同 国家、不同 区域、不同 民族 的 优异 之 处, <uline>以至</uline> ∅ 辉煌 至 今。</item> <p></p> <item> <bold> China's dietary culture </bold> through several thousand years, ∅ always has charm, is <uline>because</uline><bold>it</bold><uline>not only</uline> ethnic characteristics distinct, <uline>but also</uline> ∅ good at absorbing different countries, different regions, different ethnicities DE excellence, <uline>hence</uline> ∅ flourishing till now.</item> <p></p> <item> (China's dietary culture has endured for thousands of years and remained captivating because it not only has distinct ethnic characteristics but is also adept at absorbing the best elements from different countries, regions, and ethnic groups, resulting in its continued brilliance to this day.)</item> </ulist> <p>As for the topical cohesion, although the descriptive statistics generally follow the tendency in Hypothesis 3a and Hypothesis 3b, not all the features presented statistically significant results across the three levels. Intermediate level texts showed significantly lower subject density (subj_density) and global subject cohesion (global_subj_cohesion) than beginning level texts but were comparable to those of advanced level texts. Example 3 presented a clear contrast of topical cohesion between the beginning and the advanced level texts. In the advanced text, subject density decreased as the use of null subjects became more prevalent. Simultaneously, there was an increase in both the ratio and diversity of noun subjects (e.g., yellow mud road, road slope, tourists). The variety of semantic categories for subjects also broadened accordingly. The number (subj_semantic_num) and density (subj_semantic_density) of subject semantic class escalated significantly from lower to higher levels across all three levels. Two measurements related to noun subjects increased from lower to higher levels. Specifically, the noun subject ratio (subj_noun_ratio) differed significantly across all three levels. The noun subject diversity (subj_noun_TTR) can discriminate beginning and intermediate level texts but is relatively stable in intermediate and advanced level texts. In comparison, the pronoun subject ratio (subj_pron_ratio) declined as the text level increased, with beginning and intermediate level texts displaying a significantly higher ratio of pronoun subjects than advanced level texts.</p> <hd id="AN0187667272-25">Example 3</hd> <p></p> <ulist> <item> Beginning level</item> <p></p> <item> <bold> 阿山 </bold> 喜欢 汉语, <bold>他</bold> 现在 学习 汉语。<bold>他</bold> 常常 念 课文、写 汉字、做 练习, ∅ 也 常常 说 汉语。<bold>他</bold> 没有 时间 看 朋友, ∅ 也 没有 时间 玩儿。</item> <p></p> <item> <bold> Ashan </bold> likes Chinese, <bold>he</bold> now studies Chinese. <bold>He</bold> often reads texts, writes characters, does exercises, ∅ also often speaks Chinese. <bold>He</bold> has no time to see friends, ∅ also has no time to play.</item> <p></p> <item> (Ashan likes Chinese and is currently studying it. He often reads texts, writes characters, does exercises, and frequently speaks Chinese. He has no time to see friends or to play.)</item> <p></p> <item> Advanced level</item> <p></p> <item> <bold> ∅ 听 司机 讲, _B_我们</bold> 是 一 群 幸运儿, 要是 ∅ 碰上 下雨, <bold>黄泥 路</bold> 就 变得 又 烂 又 滑, 再 加上 <bold>公路 坡度</bold> 很 大, 即使 空车 ∅ 也 不敢 上山, <bold>游客</bold> 自然 要 失望 而 归 了。</item> <p></p> <item> <bold> ∅ Listen driver say, _B_we</bold> are a group of lucky ones, if ∅ encounter rain, <bold>yellow mud road</bold> then becomes both muddy and slippery, plus <bold>road slope</bold> very steep, even empty vehicle ∅ also not dare go up mountain, <bold>tourists</bold> naturally will disappointedly return.</item> <p></p> <item> (According to the driver, we are a lucky group. If it had rained, the yellow dirt road would have become both muddy and slippery. Coupled with the steep slope of the road, even an empty vehicle wouldn't dare to go up the mountain, and tourists would naturally have to return disappointedly.)</item> </ulist> <hd id="AN0187667272-26">Text cohesion of L2 Chinese textbooks across different genres</hd> <p>Table 5 presents the descriptive statistics for cohesion features in the L2 Chinese reading materials across three genres, narrative, argumentative, and expository texts.</p> <p>Table 5 Descriptive statistics for cohesion features by text genre<sups>*</sups></p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="2"><p>Dimension</p></th><th align="left" rowspan="2"><p>Measurements</p></th><th align="left" colspan="2"><p>Narrative</p></th><th align="left" colspan="2"><p>Argumentative</p></th><th align="left" colspan="2"><p>Expository</p></th></tr><tr><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th><th align="left"><p>Mean</p></th><th align="left"><p>SD</p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical cohesion</p></td><td align="left"><p>Local_lexical_cohesion</p></td><td char="." align="char"><p>208.75</p></td><td char="." align="char"><p>130.83</p></td><td char="." align="char"><p><bold>263.16</bold></p></td><td char="." align="char"><p>109.05</p></td><td char="." align="char"><p>234.34</p></td><td char="." align="char"><p>115.74</p></td></tr><tr><td align="left"><p>Global_lexical_coheison</p></td><td char="." align="char"><p>220.47</p></td><td char="." align="char"><p>133.55</p></td><td char="." align="char"><p><bold>299.19</bold></p></td><td char="." align="char"><p>120.70</p></td><td char="." align="char"><p>258.58</p></td><td char="." align="char"><p>134.80</p></td></tr><tr><td align="left"><p>Local_noun_cohesion</p></td><td char="." align="char"><p>8.18</p></td><td char="." align="char"><p>5.81</p></td><td char="." align="char"><p>16.67</p></td><td char="." align="char"><p>12.60</p></td><td char="." align="char"><p><bold>17.55</bold></p></td><td char="." align="char"><p>11.08</p></td></tr><tr><td align="left"><p>Global_noun_cohesion</p></td><td char="." align="char"><p>7.66</p></td><td char="." align="char"><p>5.74</p></td><td char="." align="char"><p><bold>17.62</bold></p></td><td char="." align="char"><p>14.51</p></td><td char="." align="char"><p>17.21</p></td><td char="." align="char"><p>11.76</p></td></tr><tr><td align="left"><p>Central_sent_num</p></td><td char="." align="char"><p><bold>18.45</bold></p></td><td char="." align="char"><p>10.42</p></td><td char="." align="char"><p>14.85</p></td><td char="." align="char"><p>8.10</p></td><td char="." align="char"><p>14.86</p></td><td char="." align="char"><p>8.10</p></td></tr><tr><td align="left"><p>Central_sent_ratio</p></td><td char="." align="char"><p>0.75</p></td><td char="." align="char"><p>0.25</p></td><td char="." align="char"><p><bold>0.86</bold></p></td><td char="." align="char"><p>0.10</p></td><td char="." align="char"><p>0.82</p></td><td char="." align="char"><p>0.16</p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical cohesion</p></td><td align="left"><p>Pron_density</p></td><td char="." align="char"><p><bold>0.10</bold></p></td><td char="." align="char"><p>0.04</p></td><td char="." align="char"><p>0.07</p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.04</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Pron_noun_ratio</p></td><td char="." align="char"><p><bold>0.56</bold></p></td><td char="." align="char"><p>0.28</p></td><td char="." align="char"><p>0.36</p></td><td char="." align="char"><p>0.30</p></td><td char="." align="char"><p>0.20</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Person-pron_noun_ratio</p></td><td char="." align="char"><p><bold>0.36</bold></p></td><td char="." align="char"><p>0.24</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.19</p></td><td char="." align="char"><p>0.06</p></td><td char="." align="char"><p>0.06</p></td></tr><tr><td align="left"><p>Person-pron_pron_ratio</p></td><td char="." align="char"><p><bold>0.58</bold></p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.33</p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.25</p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left"><p>Person-pron_TTR</p></td><td char="." align="char"><p>0.22</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.35</p></td><td char="." align="char"><p>0.18</p></td><td char="." align="char"><p><bold>0.35</bold></p></td><td char="." align="char"><p>0.22</p></td></tr><tr><td align="left"><p>Pron_TTR</p></td><td char="." align="char"><p>0.38</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.56</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p><bold>0.60</bold></p></td><td char="." align="char"><p>0.17</p></td></tr><tr><td align="left"><p>Pron_num_per_sent</p></td><td char="." align="char"><p><bold>1.53</bold></p></td><td char="." align="char"><p>0.58</p></td><td char="." align="char"><p>1.23</p></td><td char="." align="char"><p>0.56</p></td><td char="." align="char"><p>0.79</p></td><td char="." align="char"><p>0.47</p></td></tr><tr><td align="left"><p>Person-pron_num_per_sent</p></td><td char="." align="char"><p><bold>0.94</bold></p></td><td char="." align="char"><p>0.44</p></td><td char="." align="char"><p>0.48</p></td><td char="." align="char"><p>0.37</p></td><td char="." align="char"><p>0.24</p></td><td char="." align="char"><p>0.23</p></td></tr><tr><td align="left"><p>Conj_density</p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.01</p></td><td char="." align="char"><p><bold>0.04</bold></p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.03</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Conj_noun_ratio</p></td><td char="." align="char"><p>0.12</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p><bold>0.21</bold></p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.09</p></td></tr><tr><td align="left"><p>Conj_num_per_sent</p></td><td char="." align="char"><p>0.38</p></td><td char="." align="char"><p>0.28</p></td><td char="." align="char"><p><bold>0.78</bold></p></td><td char="." align="char"><p>0.40</p></td><td char="." align="char"><p>0.63</p></td><td char="." align="char"><p>0.39</p></td></tr><tr><td align="left"><p>Conj_TTR</p></td><td char="." align="char"><p>0.53</p></td><td char="." align="char"><p>0.24</p></td><td char="." align="char"><p><bold>0.63</bold></p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.59</p></td><td char="." align="char"><p>0.18</p></td></tr><tr><td align="left" rowspan="9"><p>Topical cohesion</p></td><td align="left"><p>Subj_density</p></td><td char="." align="char"><p><bold>0.09</bold></p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p>0.02</p></td><td char="." align="char"><p>0.08</p></td><td char="." align="char"><p>0.02</p></td></tr><tr><td align="left"><p>Subj_TTR</p></td><td char="." align="char"><p>0.57</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.66</p></td><td char="." align="char"><p>0.13</p></td><td char="." align="char"><p><bold>0.67</bold></p></td><td char="." align="char"><p>0.11</p></td></tr><tr><td align="left"><p>Subj_noun_TTR</p></td><td char="." align="char"><p><bold>0.68</bold></p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.65</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.64</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Subj_pron_ratio</p></td><td char="." align="char"><p><bold>0.38</bold></p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.22</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.10</p></td></tr><tr><td align="left"><p>Subj_noun_ratio</p></td><td char="." align="char"><p>0.41</p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.50</p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p><bold>0.61</bold></p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Local_subj_cohesion</p></td><td char="." align="char"><p><bold>0.25</bold></p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.15</p></td><td char="." align="char"><p>0.16</p></td><td char="." align="char"><p>0.13</p></td></tr><tr><td align="left"><p>Global_subj_cohesion</p></td><td char="." align="char"><p><bold>0.22</bold></p></td><td char="." align="char"><p>0.17</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.13</p></td><td char="." align="char"><p>0.14</p></td><td char="." align="char"><p>0.12</p></td></tr><tr><td align="left"><p>Subj_semantic_num</p></td><td char="." align="char"><p>5.25</p></td><td char="." align="char"><p>2.00</p></td><td char="." align="char"><p><bold>5.88</bold></p></td><td char="." align="char"><p>1.90</p></td><td char="." align="char"><p>5.49</p></td><td char="." align="char"><p>1.69</p></td></tr><tr><td align="left"><p>Subj_semantic_density</p></td><td char="." align="char"><p><bold>3.84</bold></p></td><td char="." align="char"><p>1.43</p></td><td char="." align="char"><p>3.29</p></td><td char="." align="char"><p>1.28</p></td><td char="." align="char"><p>3.48</p></td><td char="." align="char"><p>1.40</p></td></tr></tbody></table> </ephtml> </p> <p> <sups>*</sups>The highest values for each measurement across the three genres are bolded</p> <p>The main effect of text genre was significant on almost all cohesion features. Of the twenty-six cohesion features showing significant results, thirteen exhibited a large effect size, nine showed a medium effect size, and four displayed a small effect size (see Table 6). Only one topical cohesion measurement, i.e., the noun subject diversity (subj_noun_TTR): <emph>F</emph> (<reflink idref="bib2" id="ref158">2</reflink>, 441) = 2.18, <emph>p</emph> = 0.11), did not yield a significant main effect. Based on these findings, it can be concluded that overall text cohesion features vary across different genres, thereby confirming Hypothesis 4 (H4).</p> <p>Table 6 Main effects of text genre on cohesion features</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Dimension</p></th><th align="left"><p>Measurements</p></th><th align="left"><p><italic>F</italic> (2, 441)</p></th><th align="left"><p><italic>p</italic> value</p></th><th align="left"><p>Partial Eta Squared (η<sup>2</sup>)</p></th><th align="left"><p>Post-hoc tests<sup>c</sup></p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical Cohesion</p></td><td align="left"><p>Local_lexical_cohesion</p></td><td char="." align="char"><p>9.86</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td><td align="left"><p>N < A</p></td></tr><tr><td align="left"><p>Global_lexical_coheison</p></td><td char="." align="char"><p>16.04</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>N < E < A</p></td></tr><tr><td align="left"><p>Local_noun_cohesion</p></td><td char="." align="char"><p>43.79</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.17<sup>***</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left"><p>Global_noun_cohesion</p></td><td char="." align="char"><p>39.45</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.15<sup>***</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left"><p>Central_sent_num</p></td><td char="." align="char"><p>16.88</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>N > A, N > E</p></td></tr><tr><td align="left"><p>Central_sent_ratio</p></td><td char="." align="char"><p>15.41</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical Cohesion</p></td><td align="left"><p>Pron_density</p></td><td char="." align="char"><p>128.00</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.37<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Pron_noun_ratio</p></td><td char="." align="char"><p>84.40</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.28<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Person-pron_noun_ratio</p></td><td char="." align="char"><p>114.12</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.34<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Person-pron_pron_ratio</p></td><td char="." align="char"><p>161.78</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.42<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Person-pron_TTR</p></td><td char="." align="char"><p>23.82</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.10<sup>**</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left"><p>Pron_TTR</p></td><td char="." align="char"><p>87.94</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.29<sup>***</sup></p></td><td align="left"><p>N < A < E</p></td></tr><tr><td align="left"><p>Pron_num_per_sent</p></td><td char="." align="char"><p>72.41</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.25<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Person-pron_num_per_sent</p></td><td char="." align="char"><p>147.31</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.40<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Conj_density</p></td><td char="." align="char"><p>50.76</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.19<sup>***</sup></p></td><td align="left"><p>N < E < A</p></td></tr><tr><td align="left"><p>Conj_noun_ratio</p></td><td char="." align="char"><p>28.46</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.11<sup>**</sup></p></td><td align="left"><p>N < A,E < A</p></td></tr><tr><td align="left"><p>Conj_num_per_sent</p></td><td char="." align="char"><p>52.60</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.19<sup>***</sup></p></td><td align="left"><p>N < E < A</p></td></tr><tr><td align="left"><p>Conj_TTR</p></td><td char="." align="char"><p>12.02</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.05<sup>*</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left" rowspan="9"><p>Topical Cohesion</p></td><td align="left"><p>Subj_density</p></td><td char="." align="char"><p>13.28</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.06<sup>**</sup></p></td><td align="left"><p>N > A, N > E</p></td></tr><tr><td align="left"><p>Subj_TTR</p></td><td char="." align="char"><p>31.52</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.13<sup>**</sup></p></td><td align="left"><p>N < A, N < E</p></td></tr><tr><td align="left"><p>Subj_noun_TTR</p></td><td char="." align="char"><p>2.18</p></td><td char="." align="char"><p>0.11</p></td><td char="." align="char"><p>0.01</p></td><td align="left"><p>Null</p></td></tr><tr><td align="left"><p>Subj_pron_ratio</p></td><td char="." align="char"><p>127.21</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.37<sup>***</sup></p></td><td align="left"><p>N > A > E</p></td></tr><tr><td align="left"><p>Subj_noun_ratio</p></td><td char="." align="char"><p>73.00</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.25<sup>***</sup></p></td><td align="left"><p>N < A < E</p></td></tr><tr><td align="left"><p>Local_subj_cohesion</p></td><td char="." align="char"><p>15.79</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.07<sup>**</sup></p></td><td align="left"><p>N > A, N > E</p></td></tr><tr><td align="left"><p>Global_subj_cohesion</p></td><td char="." align="char"><p>20.04</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.08<sup>**</sup></p></td><td align="left"><p>N > A, N > E</p></td></tr><tr><td align="left"><p>Subj_semantic_num</p></td><td char="." align="char"><p>7.33</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td><td align="left"><p>N < A</p></td></tr><tr><td align="left"><p>Subj_semantic_density</p></td><td char="." align="char"><p>8.66</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td><td align="left"><p>N > A, N > E</p></td></tr></tbody></table> </ephtml> </p> <p> <sups>a</sups> <emph>p</emph>-values indicating significant findings are highlighted in bold <sups>b</sups>Asterisks denote the effect size: <sups>*</sups>represents a small effect size; <sups>**</sups>signifies a medium effect size; <sups>***</sups>indicates a large effect size <sups>c</sups>The column shows significant results in post-hoc tests. N represents narrative texts, A represents argumentative texts, E represents expository texts</p> <p>Tukey's HSD post hoc tests show that twelve features were identified significant between-genre differences among all three genres. For the remaining measurements that show significant results in 2-way MANOVA, thirteen features revealed no significant variations between argumentative and expository texts. This indicates a high degree of similarity between these two genres in cohesion, particularly in lexical and topical aspects. In contrast, narrative and argumentative texts displayed significant differences across all cohesion measurements that yielded significant results in the 2-way MANOVA analysis.</p> <p>In terms of the different categories of cohesion measurements, argumentative texts exhibit highest lexical cohesion on four out of six lexical features, followed by expository texts and narrative texts. The four features include the local (local_lexical_cohesion) lexical cohesion, global noun cohesion (global_noun_cohesion), and the central sentence ratio (central_sent_ratio). Compared with argumentative and expository texts, narrative texts were significantly lower in the local noun cohesion (local_noun_cochesion), but higher in the central sentence number (central_cent_num).</p> <p>As for grammatical cohesion, nine out of twelve grammatical cohesion features could discriminate three genres. Specifically, narrative texts ranked highest on four measurements related to the density and ratio of pronouns, and the number of pronouns and personal pronouns per sentence. However, narrative texts exhibited significantly lower diversity in both personal pronouns and pronouns. Regarding measurements related to the use of conjunctions, argumentative texts scored the highest. In addition, expository texts and narrative texts showed no significant differences in the conjunction-noun ratio (conj_noun_ratio). Argumentative texts and expository texts were not significantly different in the conjunction diversity (conj_TTR).</p> <p>Two topical cohesion features, including the pronoun subject ratio (subj_pron_ratio) and the noun subject ratio (subj_noun_ratio), effectively distinguish all three genres of texts. Narrative texts exhibited the highest ratio of pronoun subjects, whereas expository texts had the highest ratio of noun subjects. Argumentative and expository texts were quite similar on the remaining cohesion measurements. The subject density (subj_density) in narrative texts was significantly higher than in argumentative and expository texts, but the subject diversity (subj_TTR) was considerably lower in narrative texts. In addition, both the local and global subject cohesion of narrative texts significantly surpassed the other two text types. Meanwhile, narrative texts featured a significantly larger number of subject semantic classes (subj_semantic_num) compared to the other two genres, though the density of subject semantic classes (subj_semantic_density) was markedly lower.</p> <hd id="AN0187667272-27">Interaction between text level and genre</hd> <p>So far, we have separately examined the effect of text level and genre on text cohesion. A significant interaction effect was also identified between text level and genre in 2-way MANOVA (see Table 7). Specifically, this effect was evident in three lexical cohesion features (see Figs. 3, 4, 5), including the local noun cohesion (local_noun_cohesion): <emph>F</emph> (<reflink idref="bib2" id="ref159">2</reflink>, 441) = 2.92; <emph>p</emph> < 0.05, the central sentence number (central_sent_num): <emph>F</emph> (<reflink idref="bib2" id="ref160">2</reflink>, 441) = 7.32; <emph>p</emph> < 0.001, and the central sentence ratio (central_sent_ratio): <emph>F</emph> (<reflink idref="bib2" id="ref161">2</reflink>, 441) = 5.65; <emph>p</emph> < 0.001. Five grammatical cohesion features also demonstrated the interaction effect (see Figs. 6, 7, 8, 9, 10), including the pronoun density (pron_density): <emph>F</emph> (<reflink idref="bib2" id="ref162">2</reflink>, 441) = 3.58; <emph>p</emph> = 0.01, the personal pronoun-noun ratio (person-pron_noun_ratio): <emph>F</emph> (<reflink idref="bib2" id="ref163">2</reflink>, 441) = 2.92; <emph>p</emph> < 0.05, the personal pronoun diversity (TTR): <emph>F</emph> (<reflink idref="bib2" id="ref164">2</reflink>, 441) = 3.21; <emph>p</emph> = 0.01, the conjunction-noun ratio (conj_noun_ratio):<emph> F</emph> (<reflink idref="bib2" id="ref165">2</reflink>, 441) = 3.96; <emph>p</emph> < 0.001, and the conjunction diversity (conj_TTR):<emph> F</emph> (<reflink idref="bib2" id="ref166">2</reflink>, 441) = 21.37; <emph>p</emph> < 0.001. In addition, interaction effects were also observed on two topical cohesion features (see Figs. 11, 12), i.e., the subject density (subj_density):<emph> F</emph> (<reflink idref="bib2" id="ref167">2</reflink>, 441) = 4.07; <emph>p</emph> < 0.001 and the subject semantic class number (subj_semantic_num): <emph>F</emph> (<reflink idref="bib2" id="ref168">2</reflink>, 441) = 3.03; <emph>p</emph> < 0.05. Most of the ten features displayed a small effect size, with only two measures, namely the central sentence number and the conjunction diversity, demonstrating medium to large effect sizes. These results indicate that the influence of text level on these cohesion features was moderated by text genre.</p> <p>Table 7 Interaction effects of text level and genre</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Dimension</p></th><th align="left"><p>Measurements</p></th><th align="left"><p><italic>F</italic> (2, 441)</p></th><th align="left"><p><italic>p</italic> value</p></th><th align="left"><p>Partial Eta squared (η<sup>2</sup>)</p></th></tr></thead><tbody><tr><td align="left" rowspan="6"><p>Lexical cohesion</p></td><td align="left"><p>Local_lexical_cohesion</p></td><td char="." align="char"><p>1.05</p></td><td char="." align="char"><p>0.38</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Global_lexical_coheison</p></td><td char="." align="char"><p>1.45</p></td><td char="." align="char"><p>0.22</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Local_noun_cohesion</p></td><td char="." align="char"><p>2.92</p></td><td char="." align="char"><p><bold>0.02</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td></tr><tr><td align="left"><p>Global_noun_cohesion</p></td><td char="." align="char"><p>1.50</p></td><td char="." align="char"><p>0.20</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Central_sent_num</p></td><td char="." align="char"><p>7.32</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.06<sup>**</sup></p></td></tr><tr><td align="left"><p>Central_sent_ratio</p></td><td char="." align="char"><p>5.65</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.05<sup>*</sup></p></td></tr><tr><td align="left" rowspan="12"><p>Grammatical cohesion</p></td><td align="left"><p>Pron_density</p></td><td char="." align="char"><p>3.58</p></td><td char="." align="char"><p><bold>0.01</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td></tr><tr><td align="left"><p>Pron_noun_ratio</p></td><td char="." align="char"><p>1.12</p></td><td char="." align="char"><p>0.34</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Person-pron_noun_ratio</p></td><td char="." align="char"><p>2.92</p></td><td char="." align="char"><p><bold>0.02</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td></tr><tr><td align="left"><p>Person-pron_pron_ratio</p></td><td char="." align="char"><p>0.75</p></td><td char="." align="char"><p>0.56</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Person-pron_TTR</p></td><td char="." align="char"><p>3.21</p></td><td char="." align="char"><p><bold>0.01</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td></tr><tr><td align="left"><p>Pron_TTR</p></td><td char="." align="char"><p>1.43</p></td><td char="." align="char"><p>0.22</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Pron_num_per_sent</p></td><td char="." align="char"><p>0.58</p></td><td char="." align="char"><p>0.68</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Person-pron_num_per_sent</p></td><td char="." align="char"><p>0.34</p></td><td char="." align="char"><p>0.85</p></td><td char="." align="char"><p>0.00</p></td></tr><tr><td align="left"><p>Conj_density</p></td><td char="." align="char"><p>2.16</p></td><td char="." align="char"><p>0.07</p></td><td char="." align="char"><p>0.02<sup>*</sup></p></td></tr><tr><td align="left"><p>Conj_noun_ratio</p></td><td char="." align="char"><p>3.96</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td></tr><tr><td align="left"><p>Conj_num_per_sent</p></td><td char="." align="char"><p>0.71</p></td><td char="." align="char"><p>0.59</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Conj_TTR</p></td><td char="." align="char"><p>21.37</p></td><td char="." align="char"><p><bold>0.00</bold></p></td><td char="." align="char"><p>0.16<sup>***</sup></p></td></tr><tr><td align="left" rowspan="9"><p>Topical cohesion</p></td><td align="left"><p>Subj_density</p></td><td char="." align="char"><p>4.07</p></td><td char="." align="char"><p>0.00</p></td><td char="." align="char"><p>0.04<sup>*</sup></p></td></tr><tr><td align="left"><p>Subj_TTR</p></td><td char="." align="char"><p>2.36</p></td><td char="." align="char"><p>0.05</p></td><td char="." align="char"><p>0.02<sup>*</sup></p></td></tr><tr><td align="left"><p>Subj_n_TTR</p></td><td char="." align="char"><p>0.93</p></td><td char="." align="char"><p>0.44</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Subj_p_ratio</p></td><td char="." align="char"><p>0.96</p></td><td char="." align="char"><p>0.43</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Subj_n_ratio</p></td><td char="." align="char"><p>0.82</p></td><td char="." align="char"><p>0.51</p></td><td char="." align="char"><p>0.01</p></td></tr><tr><td align="left"><p>Local_subj_cohesion</p></td><td char="." align="char"><p>0.38</p></td><td char="." align="char"><p>0.82</p></td><td char="." align="char"><p>0.00</p></td></tr><tr><td align="left"><p>Global_subj_cohesion</p></td><td char="." align="char"><p>0.34</p></td><td char="." align="char"><p>0.85</p></td><td char="." align="char"><p>0.00</p></td></tr><tr><td align="left"><p>Subj_semantic_num</p></td><td char="." align="char"><p>3.03</p></td><td char="." align="char"><p><bold>0.02</bold></p></td><td char="." align="char"><p>0.03<sup>*</sup></p></td></tr><tr><td align="left"><p>Subj_semantic_density</p></td><td char="." align="char"><p>0.56</p></td><td char="." align="char"><p>0.70</p></td><td char="." align="char"><p>0.01</p></td></tr></tbody></table> </ephtml> </p> <p> <sups>a</sups>In the table, <emph>p</emph>-values indicating significant findings are highlighted in bold <sups>b</sups>Asterisks denote the effect size: <sups>*</sups>represents a small effect size; <sups>**</sups>signifies a medium effect size; <sups>***</sups> indicates a large effect size</p> <p>Graph: Fig. 3 Estimated marginal means of cohesion features with significant interaction effects (local_noun_cohesion)</p> <p>Graph: Fig. 4 Estimated marginal means of cohesion features with significant interaction effects (central_sent_number)</p> <p>Graph: Fig. 5 Estimated marginal means of cohesion features with significant interaction effects (central_sent_ratio)</p> <p>Graph: Fig. 6 Estimated marginal means of cohesion features with significant interaction effects (pron_density)</p> <p>Graph: Fig. 7 Estimated marginal means of cohesion features with significant interaction effects (person-pron_noun_ratio)</p> <p>Graph: Fig. 8 Estimated marginal means of cohesion features with significant interaction effects (person-pron_TTR)</p> <p>Graph: Fig. 9 Estimated marginal means of cohesion features with significant interaction effects (conj_noun_ratio)</p> <p>Graph: Fig. 10 Estimated marginal means of cohesion features with significant interaction effects (conj_TTR)</p> <p>Graph: Fig. 11 Estimated marginal means of cohesion features with significant interaction effects (subj_density)</p> <p>Graph: Fig. 12 Estimated marginal means of cohesion features with significant interaction effects (subj_semantic_num)</p> <hd id="AN0187667272-28">Summary</hd> <p>The two-way MANOVA and post-hoc tests revealed significant main effects of both text level and genre on text cohesion, and a significant interaction between them. Text cohesion exhibited notable variation across beginning, intermediate, and advanced levels in terms of lexical, topical, and grammatical aspects, with the greatest disparities observed in lexical cohesion. Text genre also had a considerable impact on text cohesion from lexical, topical, and grammatical perspectives. Argumentative and expository texts were quite similar in text cohesion, while narrative texts demonstrated significant differences from the other two genres on most of the cohesion features. Interaction effects between text level and genre were observed in several cohesion features, mainly with small effect sizes. The study highlights the importance of considering text cohesion in developing L2 Chinese reading materials.</p> <hd id="AN0187667272-29">Discussion</hd> <p></p> <hd id="AN0187667272-30">Influence of text level on text cohesion</hd> <p>The MANOVA results revealed significant differences in most cohesion features among different text levels. All six lexical cohesion features exhibited significant variations, showing a clear progression from lower to higher levels. Grammatical and topical cohesion features also distinguished text levels effectively, though a few features in these dimensions showed no significant differences.</p> <p>Specifically, all six lexical cohesion features significantly distinguish text levels, with a clear upward trend. Such a trend aligns with the development trajectories of language acquisition. Where learners become more adept at employing various semantic relations to create cohesive texts as their proficiency level progresses (Crossley et al., [<reflink idref="bib17" id="ref169">17</reflink>]; Peng et al., [<reflink idref="bib53" id="ref170">53</reflink>]). Moreover, the results corroborate research on the readability of Chinese L1 texts, which found as the length of the text increases with grade levels, the lexical richness of the texts tended to decrease because of increased proportion of repeated vocabulary (Cheng et al., [<reflink idref="bib10" id="ref171">10</reflink>]). The increased repeated vocabulary resulted to increased lexical cohesion. In addition to local and global lexical cohesion features, the central sentence number clearly discriminated all three text levels. This implies that as the complexity of contents increases, more central sentences are needed to convey them accurately. Consequently, the texts become increasingly complex and the semantic relations between sentences tighten with the rise of text level. All lexical features, except noun cohesion features, showed a more notable growth from the beginning to the intermediate level, but this growth rate decelerated from the intermediate to advanced level. This trend aligns with the findings of Plakans and Bilki ([<reflink idref="bib54" id="ref172">54</reflink>]), who observed more pronounced lexical cohesion distinctions between beginning and intermediate levels than between intermediate and advanced levels. These findings might be related to the relationship between text cohesion and cognitive processing of texts, as described by McNamara et al. ([<reflink idref="bib48" id="ref173">48</reflink>]), who found that the beneficial effects of cohesive devices for reading comprehension may be more crucial for learners at lower proficiency levels, as they help bridge the gap in their understanding of the text (McNamara et al., [<reflink idref="bib48" id="ref174">48</reflink>]).</p> <p>As for grammatical cohesion features, our results showed divergent trends for pronoun and conjunction use across text levels. The results can be interpreted from various perspectives. First, the results demonstrate students' anticipated mastery of more complex cohesive devices as they progress to higher levels of study. As mentioned earlier, zero anaphora are important yet challenging feature for Chinese L2 learners (Lu & Wu, [<reflink idref="bib46" id="ref175">46</reflink>]). The reduced proportion of pronouns in texts indicate learners' gradual master of zero anaphora in Chinese. In addition, L1 Chinese readability research has identified a positive relationship between the text difficulty and the proportion of conjunctions (Cheng et al., [<reflink idref="bib10" id="ref176">10</reflink>]). An increased use of conjunction use not only reflects developmental progress of L2 Chinese learners but also aids in comprehending more complex texts by clarifying logical relationships within the discourse (Gunnerud et al., [<reflink idref="bib24" id="ref177">24</reflink>]). Secondly, the observed trends align with the pedagogical goals outlined in the Chinese Proficiency Grading Standards for International Chinese Language Education (Grading Standards) (Hanban/Confucius Institute Headquarters, [<reflink idref="bib28" id="ref178">28</reflink>]; Ministry of Education, [<reflink idref="bib50" id="ref179">50</reflink>]). According to the Grading Standards, students at the beginning level need to understand and master the usages of basic Chinese pronouns. At the intermediate or advanced levels, students are introduced to a wider range of conjunctions to grapple with more intricate sentence structures and increasingly complex discourse relations (Yang & Sun, [<reflink idref="bib69" id="ref180">69</reflink>]). Consequently, as the text level rises, pronoun proportion decreases to accommodate a greater diversity of reference methods, such as varied conjunctions. Third, the results indicate the increasing information density and text difficulty as text level arises. Texts with higher difficulty levels often contain more information and complex concepts. To accurately convey this information and avoid ambiguity, authors may use more nouns and noun phrases while reducing the use of pronouns. In comparison, the primary function of conjunctions lies in connecting words, phrases, clauses, and sentences, and showing the relationship between them. Accordingly, the increased density and diversity of conjunctions indicate the greater length, complexity, and richness of information (Cheng et al., [<reflink idref="bib10" id="ref181">10</reflink>]).</p> <p>In terms of topical cohesion, the observation that subject density decreases with increasing text level suggests that higher-level texts utilize more implicit or understood subjects rather than explicitly stating them, which might be related to the use of "noun-pronoun-zero forms", i.e., zero anaphora, to reduce redundancy and improve text cohesion in Chinese (Sun, [<reflink idref="bib62" id="ref182">62</reflink>]). While previous studies have identified the omission of subjects or objects in sentences as a challenging yet compelling learning objective at the intermediate level (Cao, [<reflink idref="bib7" id="ref183">7</reflink>]; Ministry of Education, [<reflink idref="bib50" id="ref184">50</reflink>]), the employment of more zero anaphora also demonstrates the structural complexity of clauses in higher level texts (Hu et al., [<reflink idref="bib31" id="ref185">31</reflink>]). Secondly, it is found that higher-level texts demonstrate increased diversity in noun subjects, alongside an increased proportion of noun subjects and a decreased proportion of pronoun subjects. This indicates that higher-level texts use a more varied range of subjects and discuss a broader scope of topics. Thirdly, as text complexity increases, there is an increase in the number and density of subject semantic classes. This trend indicates that higher-level texts employ a richer variety of subjects within the same semantic field.</p> <hd id="AN0187667272-31">Influence of text genre on text cohesion</hd> <p>In addition to the text level, our study also underscores the significant influence of text genre on text cohesion. Genre refers to the specific form of language used in the community, aiming to fulfill particular social functions and uphold established cultural values (Wu, [<reflink idref="bib68" id="ref186">68</reflink>]). Texts of the same genre share similar lexico-grammatical choices, including the means to achieve text cohesion and coherence (Bublitz et al., [<reflink idref="bib5" id="ref187">5</reflink>]). All lexical and grammatical cohesion features exhibit significant differences among different text genres. Topical cohesion features also effectively differentiated text genres, with only one feature failing to show significant differences.</p> <p>Most lexical cohesion features are notably higher in argumentative and expository texts than in narratives. The higher lexical cohesion in argumentative and expository texts, as per Graesser et al. ([<reflink idref="bib21" id="ref188">21</reflink>]), might counterbalance their difficult topics. Another reason might be that these two genres often rely heavily on specific terminology, repeated concepts, and direct references to establish clear, logical connections and arguments. In comparison, narratives, focusing more on storylines and character development, may employ a wider variety of language and less repetition (Clinton et al., [<reflink idref="bib12" id="ref189">12</reflink>]; McNamara et al., [<reflink idref="bib47" id="ref190">47</reflink>]). These observations echo the study of Liu and Braine ([<reflink idref="bib44" id="ref191">44</reflink>]) on EFL writing, which found that lexical cohesion devices constituted over half of the cohesion devices used by students in argumentative writings.</p> <p>Regarding grammatical cohesion, narratives frequently use a wider array and higher number of pronouns, especially personal ones, compared to expository and argumentative texts, which align with findings in multiple languages, including, Chinese, English, Dutch, Hebrew, French, and Spanish (Berman & Nir-Sagiv, [<reflink idref="bib2" id="ref192">2</reflink>]; Clinton et al., [<reflink idref="bib12" id="ref193">12</reflink>]; Qi & Liao, [<reflink idref="bib55" id="ref194">55</reflink>]). This pattern is likely because of the detailed character interactions and continuous events, requiring frequent reference back to previously mentioned subjects for fluid storytelling in narratives (Jin, [<reflink idref="bib36" id="ref195">36</reflink>]). In contrast, argumentative texts favored the use of conjunctions, which might attribute to their high demands on the logical relationship between propositions and the smooth progression of argumentative structure (Su, Liu, Lai, & Jin, [<reflink idref="bib60" id="ref196">60</reflink>]). Conjunctions play a pivotal role in linking ideas and clarifying relationships between different points (Uccelli et al., [<reflink idref="bib66" id="ref197">66</reflink>]).</p> <p>The different natures of the three genres were also manifested in the topical cohesion features. First, narrative texts have more pronoun subjects and fewer noun subjects, while argumentative and expository texts show the opposite pattern. This indicates that the main content of narrative texts is related to characters, time, and events associated with the story plot, whereas argumentative and expository texts primarily focus on substantive subject matter (Graesser et al., [<reflink idref="bib21" id="ref198">21</reflink>]). Second, compared with expository and argumentative texts, narratives achieved a high topical cohesion level on the density of subjects and subject semantic classes, and the local and global subject cohesion, which might be explained by the fact that narratives typically have a clear theme and center around specific characters or events to maintains the flow of story (Clinton et al., [<reflink idref="bib12" id="ref199">12</reflink>]; Graesser et al., [<reflink idref="bib22" id="ref200">22</reflink>]). In contrast, argumentative and expository texts explore various aspects of a topic or presenting multiple arguments, resulting in a higher diversity of subjects (Graesser et al., [<reflink idref="bib22" id="ref201">22</reflink>]).</p> <hd id="AN0187667272-32">Interactions between text level and genre</hd> <p>The present study identified a significant interaction effect between text level and genre, indicating that the impact of text level on text cohesion is genre-dependent, and vice versa. Overall, interaction effects between text level and genre are most prevalent in grammatical cohesion features, followed by lexical cohesion features. In contrast, only two topical cohesion features exhibited such interaction effects. The findings suggest that text level and genre interact in complex ways to shape the cohesion of a text. As such, teachers should pay special attention to the cohesion features exhibiting significant interaction effect in selecting or adapting L2 Chinese reading materials. For instance, in narrative genre, pronoun usage typically decreases from lower to higher educational levels due to the initial high use of personal pronouns common in storytelling. As students progress, they increasingly adopt "zero anaphora," a technique especially prevalent in Chinese narratives. In contrast, in argumentative and expository texts, learning focuses initially on mastering personal pronouns. At more advanced levels, the need in argumentative texts to present various perspectives and in expository texts to clarify complex information by referring back to earlier mentioned concepts or objects, as well as the shift of learning focus towards effectively using demonstrative and interrogative pronouns (Peng et al., [<reflink idref="bib53" id="ref202">53</reflink>]), initially counteracts the decline in personal pronoun usage due to zero anaphora. As observed by Hu et al. ([<reflink idref="bib32" id="ref203">32</reflink>]), when texts progress to intermediate and advanced levels, the argumentative and expository texts may cover topics that are difficult or unfamiliar for students, including those related to social life or specialized areas. In this case, teachers might need to prepare texts of higher lexical cohesion levels to help students maintain the flow and a coherent mental representation of the texts (Plakans & Bilki, [<reflink idref="bib54" id="ref204">54</reflink>]). Similar patterns were also observed in Chinese L2 writing research where students' use of pronouns increased after two months learning (Qi & Liao, [<reflink idref="bib55" id="ref205">55</reflink>]). However, as students become proficient, the effect of zero anaphora becomes more pronounced, leading to a decrease in overall pronoun density.</p> <p>The interplay between level and genre is consistent with both L1 and L2 language development research. For example, Berman and Nir-Sagiv ([<reflink idref="bib3" id="ref206">3</reflink>]) found that L1 students' comprehension or construction of a coherent text is associated with both their proficiency level and text genre. Likewise, L2 writing researchers have identified unique patterns of the employment of cohesion features across genres and these patterns also exhibited a certain degree of development over time, although no clear development trajectory was identified (Abdi Tabari & Johnson, [<reflink idref="bib64" id="ref207">64</reflink>]). However, it is noted that a higher proficiency level did not necessarily ensure enhanced register awareness in the usage of discourse markers, including conjunctions, which emphasize the need to take into consideration both the proficiency level and text register or genre in L2 instruction (Qin & Uccelli, [<reflink idref="bib56" id="ref208">56</reflink>]).</p> <hd id="AN0187667272-33">Conclusion</hd> <p>This study revealed the variations of text cohesion in L2 Chinese reading materials influenced by text level and genre. The findings of the present study have enhanced our understanding of text cohesion in L2 Chinese reading resources and offered meaningful pedagogical implications for reading material preparation and reading instruction.</p> <p>The present study emphasizes the need for considering the intricate interplay of text cohesion, text level, and text genre in L2 Chinese reading material preparation and reading instruction. First, when preparing reading materials, teachers should account for text cohesion features to ensure they offer appropriate challenges and support for learners at various language proficiency stages (Bilki & Plakans, [<reflink idref="bib4" id="ref209">4</reflink>]; Clinton et al., [<reflink idref="bib12" id="ref210">12</reflink>]). Instructions should progressively guide students through the intricate repertoire of cohesion features, moving from simpler elements like pronouns to more complex ones like various types of conjunctions. This gradual approach helps students build a robust understanding of how text cohesion features contribute to the overall text coherence. Moreover, text genre might influence the formation of coherent mental representations. Therefore, genre-based tasks that expose students a variety of text genres and corresponding coherence relations might be helpful (Bilki & Plakans, [<reflink idref="bib4" id="ref211">4</reflink>]). Explicit instructions with concrete examples of how cohesion devices vary across genres can improve students' genre awareness and the ability to organize texts accordingly (Graesser et al., [<reflink idref="bib22" id="ref212">22</reflink>]; Qin & Uccelli, [<reflink idref="bib56" id="ref213">56</reflink>]). Teachers can implement exercises that enable students to practice identifying and using different cohesion devices in texts of various genres, such as text reconstruction, where students have to reassemble a disordered text using their understanding of cohesion, or contrastive analysis, where students critically analyze how cohesion is used in texts of different genres or levels (Qin & Uccelli, [<reflink idref="bib56" id="ref214">56</reflink>]).</p> <p>Despite these significant findings, our study has some limitations that warrant further exploration. First, our study primarily concentrated on a quantitative assessment of text cohesion features, the findings of which may help educators determine if a text satisfies certain numerical criteria for text cohesion. Still, future research could employ qualitative analysis to delve deeper into the specific uses of these devices across text levels or genres, thereby shedding light on how well reading materials align with the learning objectives outlined in the Grading Standards. In addition, the present study focused on analyzing text cohesion features of L2 Chinese reading texts without considering learner characteristics. Future studies could enrich our understanding of how text cohesion affects L2 Chinese reading comprehension and learning by incorporating data from reading comprehension tests.</p> <hd id="AN0187667272-34">Acknowledgements</hd> <p>This research was supported by the Outstanding Innovative Talents Cultivation Funded Programs for Doctoral Students of Jinan University (2023CXB007) to the first author of this article (Yun Lin).</p> <hd id="AN0187667272-35">Publisher's Note</hd> <p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p> <ref id="AN0187667272-36"> <title> References </title> <blist> <bibl id="bib1" idref="ref37" type="bt">1</bibl> <bibtext> Berendes K, Vajjala S, Meurers D, Bryant D, Wagner W, Chinkina M, Trautwein U. Reading demands in secondary school: Does the linguistic complexity of textbooks increase with grade level and the academic orientation of the school track?. Journal of Educational Psychology. 2018; 110; 4: 518-543. 10.1037/edu0000225</bibtext> </blist> <blist> <bibl id="bib2" idref="ref40" type="bt">2</bibl> <bibtext> Berman RA, Nir-Sagiv B. Linguistic indicators of inter-genre differentiation in later language development. Journal of Child Language. 2004; 31: 339-380. 10.1017/s0305000904006038</bibtext> </blist> <blist> <bibl id="bib3" idref="ref41" type="bt">3</bibl> <bibtext> Berman RA, Nir-Sagiv BNippold M, Scott C. The language of expository texts: Developmental perspectives. Expository discourse in children, adolescents, and adults: Development and disorders. 2009; Taylor & Francis: 101-123</bibtext> </blist> <blist> <bibl id="bib4" idref="ref43" type="bt">4</bibl> <bibtext> Bilki Z, Plakans L. Levels of textual cohesion and the meaning‐construction process of advanced second language readers. TESOL Journal. 2022. 10.1002/tesj.656</bibtext> </blist> <blist> <bibl id="bib5" idref="ref187" type="bt">5</bibl> <bibtext> Bublitz, W, Lenk, U, & Ventola, E. (1999). Coherence in Spoken and Written Discourse: How to create it and how to describe it. Selected papers from the International Workshop on Coherence, Augsburg, 24–27 April 1997 (Vol. 63). John Benjamins Publishing.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref139" type="bt">6</bibl> <bibtext> Cameron CA, Lee K, Webster S, Munro K, Hunt AK, Linton MJ. Text cohesion in children's narrative writing. Applied Psycholinguistics. 2008; 16; 3: 257-269. 10.1017/S0142716400007293</bibtext> </blist> <blist> <bibl id="bib7" idref="ref183" type="bt">7</bibl> <bibtext> Cao X. The use of discourse reference in the writings of Korean learners of Chinese. Chinese Teaching in the World. 2000; 54; 4: 77-83</bibtext> </blist> <blist> <bibl id="bib8" idref="ref11" type="bt">8</bibl> <bibtext> Chan J, Woore R, Molway L, Mutton T. Learning and teaching Chinese as a foreign language: A scoping review. Review of Education. 2022; 10; 3: 1-35. 10.1002/rev3.3370</bibtext> </blist> <blist> <bibl id="bib9" idref="ref81" type="bt">9</bibl> <bibtext> Chang, H. (2014). Analyses of text characteristics on children's Chinese text. [Master's Thesis: National Taichung University of Education]. National Taichung University of Education Institutional Repository. https://ntcuir.ntcu.edu.tw/bitstream/987654321/895/1/102NTCT0629044-001.pdf</bibtext> </blist> <blist> <bibtext> Cheng Y, Xu D, Dong J. On key factors of text reading difficulty grading and adaptability formula based on Chinese textbook corpus. Applied Linguistics [yuyan Wenzi Yingyong]. 2020; 1: 132-143</bibtext> </blist> <blist> <bibtext> Chujo K. Measuring vocabulary levels of English textbooks and tests using a BNC lemmatized high frequency word list. Language and Computers. 2004; 51: 231-249</bibtext> </blist> <blist> <bibtext> Clinton V, Taylor T, Bajpayee S, Davison ML, Carlson SE, Seipel B. Inferential comprehension differences between narrative and expository texts: A systematic review and meta-analysis. Reading and Writing. 2020; 33: 2223-2248. 10.1007/s11145-020-10044-2</bibtext> </blist> <blist> <bibtext> Crossley, S. A, Roscoe, R, & McNamara, D. S. (2011). Predicting human scores of essay quality using computational indices of linguistic and textual features. In G. Biswas, S. Bull, J. Kay, & A. Mitrovic (Eds.), Artificial Intelligence in Education: 15th International Conference (pp. 438–440). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-21869-9_62</bibtext> </blist> <blist> <bibtext> Crossley SA, Kyle K, Dascalu M. The tool for the automatic analysis of cohesion 2.0: Integrating semantic similarity and text overlap. Behavior Research Methods. 2019; 51: 14-27. 10.3758/s13428-018-1142-4</bibtext> </blist> <blist> <bibtext> Crossley SA, Kyle K, McNamara DS. The tool for the automatic analysis of text cohesion (TAACO): Automatic assessment of local, global, and text cohesion. Behavior Research Methods. 2016; 48: 1227-1237. 10.3758/s13428-015-0651-7</bibtext> </blist> <blist> <bibtext> Crossley SA, Kyle K, McNamara DS. The development and use of cohesive devices in L2 writing and their relations to judgments of essay quality. Journal of Second Language Writing. 2016; 32: 1-16. 10.1016/j.jslw.2016.01.003</bibtext> </blist> <blist> <bibtext> Crossley SA, Salsbury T, McNamara DS. The development of semantic relations in second language speakers: A case for Latent Semantic Analysis. Vigo International Journal of Applied Linguistics. 2010; 7: 55-74</bibtext> </blist> <blist> <bibtext> Crowhurst M. Cohesion in argument and narration at three grade levels. Research in the Teaching of English. 1987; 21: 185-201</bibtext> </blist> <blist> <bibtext> Cui, Y, Zhu, J, Yang, L, Fang, X, Chen, X, Wang, Y, & Yang, E. (2022). CTAP for Chinese: A linguistic complexity feature automatic calculation platform. In Proceedings of the Thirteenth Language Resources and Evaluation Conference (pp. 5525–5538). https://aclanthology.org/2022.lrec-1.592.pdf</bibtext> </blist> <blist> <bibtext> Finch H. Comparison of the performance of nonparametric and parametric MANOVA test statistics when assumptions are violated. Methodology. 2005; 1; 1: 27-38. 10.1027/1614-1881.1.1.27</bibtext> </blist> <blist> <bibtext> Graesser AC, McNamara DS, Cai Z, Conley M, Li H, Pennebaker J. Coh-Metrix measures text characteristics at multiple levels of language and discourse. The Elementary School Journal. 2014; 115; 2: 210-229. 10.1086/678293</bibtext> </blist> <blist> <bibtext> Graesser AC, McNamara DS, Louwerse MMSweet AP, Snow CE. What do readers need to learn in order to process coherence relations in narrative and expository text. Rethinking reading comprehension. 2003; The Guilford Press: 82-98</bibtext> </blist> <blist> <bibtext> Graesser AC, McNamara DS, Louwerse MM, Cai Z. Coh-Metrix: Analysis of text on cohesion and language. Behavior Research Methods, Instruments, & Computers. 2004; 36; 2: 193-202. 10.3758/BF03195564</bibtext> </blist> <blist> <bibtext> Gunnerud HL, Foldnes N, Melby-Lervåg M. Levels of skills and predictive patterns of reading comprehension in bilingual children with an early age of acquisition. Reading and Writing. 2022; 35; 10: 2365-2387. 10.1007/s11145-022-10286-2</bibtext> </blist> <blist> <bibtext> Hakstian AR, Roed JC, Lind JC. Two-sample T2 procedure and the assumption of homogeneous covariance matrices. Psychological Bulletin. 1979; 86: 1255-1263</bibtext> </blist> <blist> <bibtext> Halliday MAK, Hasan R. Cohesion in English. 1976; Longman</bibtext> </blist> <blist> <bibtext> Halliday MAK, Hasan R. Language, context, and text: Aspects of language in a social semiotic perspective. 19852; Oxford University Press</bibtext> </blist> <blist> <bibtext> Hanban/Confucius Institute Headquarters. Chinese proficiency grading standards for international Chinese language education. 2009; Foreign Language Teaching and Research Press</bibtext> </blist> <blist> <bibtext> Hiebert EH, Mesmer HAE. Upping the ante of text complexity in the Common Core State Standards: Examining its potential impact on young readers. Educational Researcher. 2013; 42; 1: 44-51. 10.3102/0013189X12459802</bibtext> </blist> <blist> <bibtext> Hoey M. Patterns of lexis in text. 1991; Oxford University Press</bibtext> </blist> <blist> <bibtext> Hu R, Wu J, Lu X. Word-combination-based measures of phraseological diversity, sophistication, and complexity and their relationship to second language Chinese proficiency and writing quality. Language Learning. 2022; 72; 4: 1128-1169. 10.1111/lang.12511</bibtext> </blist> <blist> <bibtext> Hu R, Zhu Q, Yang L. The research and construction of a topic corpus for teaching Chinese as a second language. Journal of Chinese Information Processing. 2015; 29; 6: 62-68</bibtext> </blist> <blist> <bibtext> Hu Z. Discourse cohesion and coherence. 1994; Shanghai Foreign Language Education Press</bibtext> </blist> <blist> <bibtext> Jin T, Lu X. A data-driven approach to text adaptation in teaching material preparation: Design, implementation, and teacher professional development. TESOL Quarterly. 2018; 52; 2: 457-467. 10.1002/tesq.434</bibtext> </blist> <blist> <bibtext> Jin T, Lu X, Ni J. Syntactic complexity in adapted teaching materials: Differences among grade levels and implications for benchmarking. The Modern Language Journal. 2020; 104; 1: 192-208. 10.1111/modl.12622</bibtext> </blist> <blist> <bibtext> Jin Z. Dictionary of genre. 1986; Northeast Normal University Press</bibtext> </blist> <blist> <bibtext> Kleijn S, Pander Maat HL, Sanders TJ. Comprehension effects of connectives across texts, readers, and coherence relations. Discourse Processes. 2019; 56; 5–6: 447-464. 10.1080/0163853X.2019.1605257</bibtext> </blist> <blist> <bibtext> Lee DY. Defining core vocabulary and tracking its distribution across spoken and written genres: Evidence of a gradience of variation from the British National Corpus. Journal of English Linguistics. 2001; 29; 3: 250-278. 10.1177/0075424012200536</bibtext> </blist> <blist> <bibtext> Li W. Topic chains in Chinese discourse. Discourse Processes. 2004; 37; 1: 25-45. 10.1207/s15326950dp3701_2</bibtext> </blist> <blist> <bibtext> Li Y. Cohesion in text and text aboutness. 2012; Sun Yat-Sen University Press</bibtext> </blist> <blist> <bibtext> Liao Q. Xiandai Hanyu Pianzhang Zhong de Lianjie Chengfen. [the discourse coherence elements in modem Chinese]. Zhongguo Yuwen. 1986; 6: 413-427</bibtext> </blist> <blist> <bibtext> Liao Q. Xiandai Hanyu Yupian zhong Zhitong de Biaoda [expressions of referential identity in Chinese discourse]. Zhongguo Yuwen. 1986; 2: 88-96</bibtext> </blist> <blist> <bibtext> Liu G. Jiyu Xianjie Tezheng de Yingyu Xuexizhe Shumianyu Yupian Lianguan Zidong Pingjia Yanjiu [Research on the automatic evaluation of written discourse coherence for English learners based on cohesive features]. 2018; Foreign Langauge Teaching and Research Press</bibtext> </blist> <blist> <bibtext> Liu M, Braine G. Cohesive features in argumentative writing produced by Chinese undergraduates. System. 2005; 33; 4: 623-636. 10.1016/j.system.2005.02.002</bibtext> </blist> <blist> <bibtext> Lu X. Computational methods for corpus annotation and analysis. 2014; Springer</bibtext> </blist> <blist> <bibtext> Lu X, Wu J. Noun-phrase complexity measures in Chinese and their relationship to L2 Chinese writing quality: A comparison with topic–comment-unit-based measures. The Modern Language Journal. 2022; 106; 1: 267-283. 10.1111/modl.12766</bibtext> </blist> <blist> <bibtext> McNamara DS, Graesser AC, Louwerse MMSabatini JP, Albro E. Sources of text difficulty: Across the ages and genres. Assessing reading in the 21st century: Aligning and applying advances in the reading and measurement sciences. 2012: Lanham, MD; R&L Education</bibtext> </blist> <blist> <bibtext> McNamara DS, Kintsch E, Songer NB, Kintsch W. Are good texts always better? Interactions of text coherence, background knowledge, and levels of understanding in learning from text. Cognition and Instruction. 1996; 14; 1: 1-43. 10.1207/s1532690xci14011</bibtext> </blist> <blist> <bibtext> Mesmer HAE, Hiebert EH, Cunningham JW, Kapania M. Does one size fit all? Exploring the contribution of text features, text content, and grade of use on comprehension. Reading Psychology. 2021; 42; 1: 42-72. 10.1080/02702711.2020.1861864</bibtext> </blist> <blist> <bibtext> Ministry of Education. (2021). Chinese proficiency grading standards for international Chinese education. <ulink href="http://www.moe.gov.cn/jyb%5fsjzl/ziliao/A19/202111/W020211118507389477190.pdf">http://www.moe.gov.cn/jyb%5fsjzl/ziliao/A19/202111/W020211118507389477190.pdf</ulink> (accessed 28 May 2023).</bibtext> </blist> <blist> <bibtext> O'Reilly T, McNamara DS. The impact of science knowledge, reading skill, and reading strategy knowledge on more traditional "high-stakes" measures of high school students' science achievement. American Educational Research Journal. 2007; 44; 1: 161-196. 10.3102/0002831206298171</bibtext> </blist> <blist> <bibtext> Olson CL. Comparative robustness of six tests in multivariate analysis of variance. Journal of the American Statistical Association. 1974; 69: 894-908</bibtext> </blist> <blist> <bibtext> Peng, Y, Hu, R, & Wu, J. (2023). Automatic analysis and application of cohesion features of Chinese texts. Yuyan Wenzi Yingyong [Applied Linguistics], 01, 114–129.</bibtext> </blist> <blist> <bibtext> Plakans L, Bilki Z. Cohesion features in ESL reading: Comparing beginning, intermediate and advanced textbooks. Reading in a Foreign Language. 2016; 26; 1: 79-100</bibtext> </blist> <blist> <bibtext> Qi H, Liao J. An investigation into Chinese linguistic development in L2 narrative and argumentative writing. Chinese Teaching in the World. 2019; 33; 4: 563-576</bibtext> </blist> <blist> <bibtext> Qin W, Uccelli P. Beyond linguistic complexity: Assessing register flexibility in EFL writing across contexts. Assessing Writing. 2020; 45: 100465. 10.1016/j.asw.2020.100465</bibtext> </blist> <blist> <bibtext> Rets I, Astruc L, Coughlan T, Stickler U. Approaches to simplifying academic texts in English: English teachers' views and practices. English for Specific Purposes. 2022; 68: 31-46. 10.1016/j.esp.2022.06.001</bibtext> </blist> <blist> <bibtext> Sheehan, K. M, Kostin, I, & Futagi, Y. (2008). When do standard approaches for measuring vocabulary difficulty, syntactic complexity and referential cohesion yield biased estimates of text difficulty. In Proceedings of the 30th Annual Conference of the Cognitive Science Society, Washington DC.</bibtext> </blist> <blist> <bibtext> Shi W, Chen J. The linguistic features of L2 Chinese writing performance in different genres. Chinese Language Learning. 2019; 6: 76-85</bibtext> </blist> <blist> <bibtext> Su, Y, Liu, K, Lai, C, & Jin, T. (2021). The progression of collaborative argumentation among English learners: A qualitative study. System, 98, 102471. https://doi.org/10.1016/j.system.2021.102471</bibtext> </blist> <blist> <bibtext> Su, Y, Liu, K, Liu, F, Lee, J. & Jin, T. (2024). Lexical complexity in exemplar EFL texts: Towards text adaptation for 12 grades of basic English curriculum in China. International Review of Applied Linguistics in Language Teaching, 62(1), 137–164. https://doi-org.bham-ezproxy.idm.oclc.org/10.1515/iral-2022-0236</bibtext> </blist> <blist> <bibtext> Sun K. The nature and characteristics of topic chains in Chinese written discourse—a contrastive study of topic chain and zero anaphora. Chinese Language Learning. 2014; 6: 52-63</bibtext> </blist> <blist> <bibtext> Sung YT, Lin WC, Dyson SB, Chang KE, Chen YC. Leveling L2 texts through readability: Combining multilevel linguistic features with the CEFR. The Modern Language Journal. 2015; 99; 2: 371-391. 10.1111/modl.12213</bibtext> </blist> <blist> <bibtext> Tabari MA, Johnson MD. Exploring new insights into the role of cohesive devices in written academic genres. Assessing Writing. 2023; 57: 100749. 10.1016/j.asw.2023.100749</bibtext> </blist> <blist> <bibtext> Tortorelli LS. Beyond first grade: Examining word, sentence, and discourse text factors associated with oral reading rate in informational text in second grade. Reading and Writing. 2020; 33; 1: 143-170. 10.1007/s11145-019-09956-5</bibtext> </blist> <blist> <bibtext> Uccelli P, Dobbs CL, Scott J. Mastering academic language: Organization and stance in the persuasive writing of high school students. Written Communication. 2013; 30; 1: 36-62. 10.1177/0741088312469013</bibtext> </blist> <blist> <bibtext> Wang Z, Bukhari N, Tang Q. The basic differences of textual cohesion between English and Chinese. SHS Web of Conferences. 2023; 159: 1-4. 10.1051/shsconf/202315901018</bibtext> </blist> <blist> <bibtext> Wu J. Comparative study of linguistic features in different genres in Chinese writing by South Korean students. Language Teaching and Linguistic Studies. 2019; 199; 5: 1-12</bibtext> </blist> <blist> <bibtext> Yang W, Sun Y. The use of cohesive devices in argumentative writing by Chinese EFL learners at different proficiency levels. Linguistics and Education. 2012; 23; 1: 31-48. 10.1016/j.linged.2011.09.004</bibtext> </blist> <blist> <bibtext> Zhang, N, & Jin, T. (2021). The Corpus of Chinese Teaching Textbook. Retrieved from https://languagedata.net/corpus/</bibtext> </blist> <blist> <bibtext> Zhao Y. A grammar of spoken Chinese. 1979; The Commercial Press</bibtext> </blist> <blist> <bibtext> Zhou J, Luo Y, Chen B. A study on the expressiveness of the theme from the perspective of the subject. Yuyan Wenzi Yingyong [applied Linguistics]. 2018; 01: 61-70</bibtext> </blist> <blist> <bibtext> Zhou X, Bo W, Wang L, Li Y. A study of the construction and application of the TCSOL material corpus. Yuyan Wenzi Yingyong [applied Linguistics]. 2017; 01: 125-135. 10.16499/j.cnki.1003-5397.2017.01.014</bibtext> </blist> <blist> <bibtext> Zhu D. Lecture notes on grammar. 1982; The Commercial Press</bibtext> </blist> </ref> <aug> <p>By Yun Lin; Yanfang Su; Yiping Peng and Hua Liu</p> <p>Reported by Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib63" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib49" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib45" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib11" firstref="ref4"></nolink> <nolink nlid="nl5" bibid="bib34" firstref="ref5"></nolink> <nolink nlid="nl6" bibid="bib35" firstref="ref6"></nolink> <nolink nlid="nl7" bibid="bib14" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib29" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib61" firstref="ref17"></nolink> <nolink nlid="nl10" bibid="bib65" firstref="ref18"></nolink> <nolink nlid="nl11" bibid="bib58" firstref="ref21"></nolink> <nolink nlid="nl12" bibid="bib38" firstref="ref22"></nolink> <nolink nlid="nl13" bibid="bib30" firstref="ref25"></nolink> <nolink nlid="nl14" bibid="bib23" firstref="ref30"></nolink> <nolink nlid="nl15" bibid="bib26" firstref="ref31"></nolink> <nolink nlid="nl16" bibid="bib27" firstref="ref34"></nolink> <nolink nlid="nl17" bibid="bib40" firstref="ref39"></nolink> <nolink nlid="nl18" bibid="bib41" firstref="ref60"></nolink> <nolink nlid="nl19" bibid="bib42" firstref="ref61"></nolink> <nolink nlid="nl20" bibid="bib33" firstref="ref66"></nolink> <nolink nlid="nl21" bibid="bib67" firstref="ref74"></nolink> <nolink nlid="nl22" bibid="bib62" firstref="ref75"></nolink> <nolink nlid="nl23" bibid="bib46" firstref="ref77"></nolink> <nolink nlid="nl24" bibid="bib15" firstref="ref79"></nolink> <nolink nlid="nl25" bibid="bib16" firstref="ref80"></nolink> <nolink nlid="nl26" bibid="bib19" firstref="ref83"></nolink> <nolink nlid="nl27" bibid="bib53" firstref="ref84"></nolink> <nolink nlid="nl28" bibid="bib57" firstref="ref94"></nolink> <nolink nlid="nl29" bibid="bib37" firstref="ref95"></nolink> <nolink nlid="nl30" bibid="bib51" firstref="ref97"></nolink> <nolink nlid="nl31" bibid="bib21" firstref="ref101"></nolink> <nolink nlid="nl32" bibid="bib54" firstref="ref102"></nolink> <nolink nlid="nl33" bibid="bib17" firstref="ref104"></nolink> <nolink nlid="nl34" bibid="bib69" firstref="ref108"></nolink> <nolink nlid="nl35" bibid="bib18" firstref="ref109"></nolink> <nolink nlid="nl36" bibid="bib13" firstref="ref110"></nolink> <nolink nlid="nl37" bibid="bib72" firstref="ref114"></nolink> <nolink nlid="nl38" bibid="bib47" firstref="ref116"></nolink> <nolink nlid="nl39" bibid="bib68" firstref="ref118"></nolink> <nolink nlid="nl40" bibid="bib59" firstref="ref119"></nolink> <nolink nlid="nl41" bibid="bib70" firstref="ref121"></nolink> <nolink nlid="nl42" bibid="bib73" firstref="ref123"></nolink> <nolink nlid="nl43" bibid="bib50" firstref="ref124"></nolink> <nolink nlid="nl44" bibid="bib36" firstref="ref125"></nolink> <nolink nlid="nl45" bibid="bib43" firstref="ref135"></nolink> <nolink nlid="nl46" bibid="bib39" firstref="ref141"></nolink> <nolink nlid="nl47" bibid="bib74" firstref="ref143"></nolink> <nolink nlid="nl48" bibid="bib71" firstref="ref144"></nolink> <nolink nlid="nl49" bibid="bib20" firstref="ref146"></nolink> <nolink nlid="nl50" bibid="bib25" firstref="ref147"></nolink> <nolink nlid="nl51" bibid="bib52" firstref="ref148"></nolink> <nolink nlid="nl52" bibid="bib10" firstref="ref171"></nolink> <nolink nlid="nl53" bibid="bib48" firstref="ref173"></nolink> <nolink nlid="nl54" bibid="bib24" firstref="ref177"></nolink> <nolink nlid="nl55" bibid="bib28" firstref="ref178"></nolink> <nolink nlid="nl56" bibid="bib31" firstref="ref185"></nolink> <nolink nlid="nl57" bibid="bib12" firstref="ref189"></nolink> <nolink nlid="nl58" bibid="bib44" firstref="ref191"></nolink> <nolink nlid="nl59" bibid="bib55" firstref="ref194"></nolink> <nolink nlid="nl60" bibid="bib60" firstref="ref196"></nolink> <nolink nlid="nl61" bibid="bib66" firstref="ref197"></nolink> <nolink nlid="nl62" bibid="bib22" firstref="ref200"></nolink> <nolink nlid="nl63" bibid="bib32" firstref="ref203"></nolink> <nolink nlid="nl64" bibid="bib64" firstref="ref207"></nolink> <nolink nlid="nl65" bibid="bib56" firstref="ref208"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1482417
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Profiling Text Cohesion in the Development of L2 Chinese Reading Materials: Variation by Text Level and Genre
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Yun+Lin%22">Yun Lin</searchLink><br /><searchLink fieldCode="AR" term="%22Yanfang+Su%22">Yanfang Su</searchLink><br /><searchLink fieldCode="AR" term="%22Yiping+Peng%22">Yiping Peng</searchLink><br /><searchLink fieldCode="AR" term="%22Hua+Liu%22">Hua Liu</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Reading+and+Writing%3A+An+Interdisciplinary+Journal%22"><i>Reading and Writing: An Interdisciplinary Journal</i></searchLink>. 2025 38(7):1949-1984.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 36
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Connected+Discourse%22">Connected Discourse</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Comprehension%22">Reading Comprehension</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Materials%22">Reading Materials</searchLink><br /><searchLink fieldCode="DE" term="%22Material+Development%22">Material Development</searchLink><br /><searchLink fieldCode="DE" term="%22Literary+Genres%22">Literary Genres</searchLink><br /><searchLink fieldCode="DE" term="%22Second+Language+Learning%22">Second Language Learning</searchLink><br /><searchLink fieldCode="DE" term="%22Chinese%22">Chinese</searchLink><br /><searchLink fieldCode="DE" term="%22Narration%22">Narration</searchLink><br /><searchLink fieldCode="DE" term="%22Persuasive+Discourse%22">Persuasive Discourse</searchLink><br /><searchLink fieldCode="DE" term="%22Expository+Writing%22">Expository Writing</searchLink><br /><searchLink fieldCode="DE" term="%22Effect+Size%22">Effect Size</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1007/s11145-024-10585-w
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0922-4777<br />1573-0905
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Developing reading materials tailored to the specific needs and proficiency levels of students is important yet challenging for educators and teachers. Current data-driven indices assisting this task primarily focus on lexical and syntactic text features, often overlooking text cohesion, a key discourse feature impacting reading comprehension. To fill this gap, this study explored the variation of text cohesion in second language (L2) Chinese reading materials across different text levels and genres. We used an NLP-based tool to examine the lexical, grammatical, and topical cohesion features of 450 representative L2 Chinese reading texts. A two-way MANOVA was then conducted to analyze the variation of text cohesion across different text levels and genres, as well as the interaction between text levels and genres. Our findings have revealed significant disparities in text cohesion across varying text levels, particularly in lexical cohesion features, and among different genres, with narratives standing out from argumentative and expository texts. Interaction effects between text level and genre were noted, albeit with small effect sizes. This study highlights the importance of text cohesion when developing reading materials for L2 Chinese learners and offers insights into developing text cohesion benchmarks for preparing reading materials of different levels and genres.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1482417
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1482417
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s11145-024-10585-w
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 36
        StartPage: 1949
    Subjects:
      – SubjectFull: Connected Discourse
        Type: general
      – SubjectFull: Reading Comprehension
        Type: general
      – SubjectFull: Reading Materials
        Type: general
      – SubjectFull: Material Development
        Type: general
      – SubjectFull: Literary Genres
        Type: general
      – SubjectFull: Second Language Learning
        Type: general
      – SubjectFull: Chinese
        Type: general
      – SubjectFull: Narration
        Type: general
      – SubjectFull: Persuasive Discourse
        Type: general
      – SubjectFull: Expository Writing
        Type: general
      – SubjectFull: Effect Size
        Type: general
    Titles:
      – TitleFull: Profiling Text Cohesion in the Development of L2 Chinese Reading Materials: Variation by Text Level and Genre
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Yun Lin
      – PersonEntity:
          Name:
            NameFull: Yanfang Su
      – PersonEntity:
          Name:
            NameFull: Yiping Peng
      – PersonEntity:
          Name:
            NameFull: Hua Liu
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 09
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0922-4777
            – Type: issn-electronic
              Value: 1573-0905
          Numbering:
            – Type: volume
              Value: 38
            – Type: issue
              Value: 7
          Titles:
            – TitleFull: Reading and Writing: An Interdisciplinary Journal
              Type: main
ResultId 1