View in EDS HTML Full Text PDF Full Text

The Effects of Schools Providing Compensatory Cultural Capital on Student Reading in the Case of Taiwanese Students Participating in PISA

Saved in:
Bibliographic Details
Title: The Effects of Schools Providing Compensatory Cultural Capital on Student Reading in the Case of Taiwanese Students Participating in PISA
Language: English
Authors: Tien-Hui Chiang (ORCID 0000-0001-8255-1049), Pei-Ming Chiang, Su-Wei Lin, Allen Thurston (ORCID 0000-0001-9518-3257)
Source: British Educational Research Journal. 2025 51(4):1820-1852.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 33
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Education Level: Secondary Education
Descriptors: Cultural Capital, Socioeconomic Status, Foreign Countries, Achievement Tests, International Assessment, Secondary School Students, Reading, Reading Habits, Cognitive Development, Learning Strategies
Geographic Terms: Taiwan
Assessment and Survey Identifiers: Program for International Student Assessment
DOI: 10.1002/berj.4157
ISSN: 0141-1926
1469-3518
Abstract: The structural approach of cultural capital theories neglects the idea that the impact of structural constraints on educational results can be reduced by agency. This predicament can be solved when the advantages of abundant educational resources available in schools are unpacked, since doing so can compensate for the paucity of such resources often seen in the low-socioeconomic status (SES) family social space. Although the gains available from such resources remain embedded in the school social space, it can be assumed that their compensatory function can be activated through reading activities that contribute significantly to students' cognitive development. This situation prompts two research questions related to the contributions of reading to academic disciplines such as mathematics and science, and the amelioration of the structural impact of SES and cultural capital on low-SES students' learning outcomes. To explore these questions, we used regression analysis and hierarchical linear modelling (HLM) to analyse data from the stratified random sample of Taiwanese students (n = 7342) contained in the Programme for International Student Assessment (PISA) 2018 dataset. The estimates of regression analysis showed that reading ability functioned as a reliable indicator of Taiwanese students' mathematics and science scores on PISA 2018. The results of HLM analysis further demonstrated that the predominant influence of economic, social and cultural status and cultural capital can be attenuated significantly when the independent variables of school support and student personal efforts/learning strategies are considered. Accordingly, reading resources can be regarded as a compensatory genre of cultural capital embedded within the school social space, at least in the case of Taiwan, as the benefits they create need to be achieved through reading plans/projects scientifically implemented by schools.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1480416
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFHZr57dOUyvSXae_uqQD5MAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDFEzngGUO6ABbt4ozwIBEICBm1ZmohDi0YVLsprcc6K7q-t6gybT8PiE1pg7HkgZGKgqGbOzegVZbVmIiPXO_aKZkGU4sd77GUiAjlCDB62kZXxEwQ0-VjHzTHCrp2Z4Jug0SXKXo-9C6VbKoeNpfRtwE10_RLHDAr3OG3sj7186ENRHDgdCVHM0UKsXTuAWqdNzFyKBfoJrk9mChTZhGkMVNCgGM42Rz8_BbODf
Text:
  Availability: 1
  Value: <anid>AN0187391082;bed01aug.25;2025Aug20.05:27;v2.2.500</anid> <title id="AN0187391082-1">The effects of schools providing compensatory cultural capital on student reading in the case of Taiwanese students participating in PISA </title> <p>The structural approach of cultural capital theories neglects the idea that the impact of structural constraints on educational results can be reduced by agency. This predicament can be solved when the advantages of abundant educational resources available in schools are unpacked, since doing so can compensate for the paucity of such resources often seen in the low‐socioeconomic status (SES) family social space. Although the gains available from such resources remain embedded in the school social space, it can be assumed that their compensatory function can be activated through reading activities that contribute significantly to students' cognitive development. This situation prompts two research questions related to the contributions of reading to academic disciplines such as mathematics and science, and the amelioration of the structural impact of SES and cultural capital on low‐SES students' learning outcomes. To explore these questions, we used regression analysis and hierarchical linear modelling (HLM) to analyse data from the stratified random sample of Taiwanese students (n = 7342) contained in the Programme for International Student Assessment (PISA) 2018 dataset. The estimates of regression analysis showed that reading ability functioned as a reliable indicator of Taiwanese students' mathematics and science scores on PISA 2018. The results of HLM analysis further demonstrated that the predominant influence of economic, social and cultural status and cultural capital can be attenuated significantly when the independent variables of school support and student personal efforts/learning strategies are considered. Accordingly, reading resources can be regarded as a compensatory genre of cultural capital embedded within the school social space, at least in the case of Taiwan, as the benefits they create need to be achieved through reading plans/projects scientifically implemented by schools.</p> <p>Keywords: cultural capital; PISA; reading; SES</p> <p>Key insights What is the main issue that the paper addresses?This study sets out to uncover the contribution of reading to ameliorating cultural reproduction. What are the main insights that the paper provides?The findings show that reading resources function as compensatory cultural capital embedded within the school social space, as the benefits they create need to be achieved through reading programmes scientifically implemented by schools.</p> <hd id="AN0187391082-2">INTRODUCTION</hd> <p>Inequity in educational results has been systematically proven as a social fact across different levels of education, such as preschool education (Linberg et al., [<reflink idref="bib66" id="ref1">66</reflink>]), primary schools, secondary schools (Møllegaard & Jæger, [<reflink idref="bib75" id="ref2">75</reflink>]) and higher education institutes (Matherly et al., [<reflink idref="bib72" id="ref3">72</reflink>]). Bourdieu conceptualises it as cultural reproduction, proposing the principle that cultural capital invisibly manipulates educational results (Bourdieu, [<reflink idref="bib17" id="ref4">17</reflink>], [<reflink idref="bib21" id="ref5">21</reflink>]). Put simply, parents mobilise their economic capital to invest cultural capital in their family social spaces. In doing so, they incubate a specific type of habitus in their children that functions as a disposition governing their reasoning abilities (Bourdieu, [<reflink idref="bib22" id="ref6">22</reflink>]) and ultimately arbitrates their learning outcomes (Bourdieu, [<reflink idref="bib14" id="ref7">14</reflink>], [<reflink idref="bib18" id="ref8">18</reflink>]). Because economic capital is a precondition for purchasing cultural capital, high‐socioeconomic status (SES) parents are capable of establishing an academic environment for their children, so this student group is likely to develop an academic habitus, leading to excellent learning outcomes (Choi et al., [<reflink idref="bib34" id="ref9">34</reflink>]; Hek & Kraaykamp, [<reflink idref="bib50" id="ref10">50</reflink>]; Katartzi & Hayward, [<reflink idref="bib58" id="ref11">58</reflink>]; Lancee & Werfhorst, [<reflink idref="bib61" id="ref12">61</reflink>]; Møllegaard & Jæger, [<reflink idref="bib75" id="ref13">75</reflink>]). In contrast, their low‐SES peers reside in a manual‐based social space in which cultural capital is severely lacking, due to their parents' poor financial circumstances. As a result, these children often possess a manual habitus that hampers their academic learning (Bourdieu, [<reflink idref="bib14" id="ref14">14</reflink>], [<reflink idref="bib19" id="ref15">19</reflink>], [<reflink idref="bib22" id="ref16">22</reflink>]). While the Bourdieusian school creates powerful insights into the phenomenon of cultural reproduction, its structural approach hasn't clearly recognised the influence of agency on people's actions and consequences, which is mediated by their visions, determined will (Chiang et al., [<reflink idref="bib31" id="ref17">31</reflink>]), capabilities (MacKenzie et al., [<reflink idref="bib71" id="ref18">71</reflink>]) and planned actions (Chiang, Toh et al., [<reflink idref="bib32" id="ref19">32</reflink>]; Tran, [<reflink idref="bib107" id="ref20">107</reflink>]). Another shortcoming is the need to address the situation of students in dual forms of social space, namely primary social spaces (families with insufficient cultural capital) and secondary social spaces (educational institutes, particularly schools equipped with sufficient educational resources) (Chiang & Trezise, [<reflink idref="bib33" id="ref21">33</reflink>]). The secondary social space substantially facilitates the ability of talented low‐SES students to attain excellent learning outcomes when they effectively exercise individualised agency by exploiting such resources available in schools (Chiang et al., [<reflink idref="bib31" id="ref22">31</reflink>]). Unfortunately, this individualised agency may not enable these students to take full advantage of such resources because they remain in the minority. However, we can infer that if the individualised agency is remoulded into an institutionalised form, empowering schools to transform their educational resources into compensatory cultural capital for low‐SES students, enshrined within the secondary social space, the phenomenon of cultural reproduction can be ameliorated to some extent. Put simply, its solution can be found in dual forms of social space, referring to families and schools, respectively (Chiang & Trezise, [<reflink idref="bib33" id="ref23">33</reflink>]), because it is reported that when low‐SES students can make good use of educational resources available in school social spaces, the structural constraints of their family social space, notably the shortage of cultural capital, can be ameliorated (Chiang et al., [<reflink idref="bib31" id="ref24">31</reflink>]). Unfortunately, very few such students tend to become successful academically through this path. However, it can be posited that when this agency is embedded not in individual students but in schools themselves, the activation of resources can comprehensively level up low‐SES students' learning outcomes. Although schools are the primary social institutes for distributing educational resources, the advantages of such resources haven't yet been fully unleashed. This predicament can be relieved through reading, which research shows can significantly improve students' language (Wasik et al., [<reflink idref="bib114" id="ref25">114</reflink>]), literacy (Buchon‐Stoll, [<reflink idref="bib26" id="ref26">26</reflink>]; Griffin et al., [<reflink idref="bib47" id="ref27">47</reflink>]; Silverman, [<reflink idref="bib102" id="ref28">102</reflink>]; Thurston, Cockerill et al., [<reflink idref="bib105" id="ref29">105</reflink>]; Wasik et al., [<reflink idref="bib114" id="ref30">114</reflink>]), comprehension and cognitive abilities (Dickinson & Anastasopoulos, [<reflink idref="bib40" id="ref31">40</reflink>]; Hulme & Snowling, [<reflink idref="bib53" id="ref32">53</reflink>]; Whorrall & Cabell, [<reflink idref="bib116" id="ref33">116</reflink>]). However, these findings are mainly based on premises drawn on psychology without accommodating theories of cultural capital. From this perspective, reading resources enshrined within school social spaces may be able to compensate for the shortage of cultural capital in low‐SES family social spaces if their advantages are unlocked through the scientific implementation of reading plans or projects in schools. This compensation may further alleviate the phenomenon of cultural reproduction. Considering the above possible correlations, this study sets out to explore the compensatory function of reading, and its correlations with achievement in such academic disciplines as mathematics and science, along with ways of reducing the interwoven impact of SES and cultural capital on student reading performance. In short, the structural orbit of cultural capital has failed to address the influence of agency on educational results. It has also overlooked dual forms of social space, in which compensatory cultural capital is enshrined within school social space rather than family social space. This compensatory cultural capital can be unpacked when schools systematically introduce reading intervention programmes. This study further attempts to uncover which factors affect the contribution of reading to students' academic attainment.</p> <hd id="AN0187391082-3">THEORETICAL CONCEPTS</hd> <p></p> <hd id="AN0187391082-4">Cultural capital</hd> <p>Bourdieu ([<reflink idref="bib18" id="ref34">18</reflink>]) argues that symbolic violence, essentially based on cultural legitimacy that transmits knowledge of bodies to steer people's perceptions and daily philosophy (Bourdieu, [<reflink idref="bib13" id="ref35">13</reflink>], [<reflink idref="bib22" id="ref36">22</reflink>]), has become a crucial conduit for reproducing the class power structure through its two core components—economic capital and cultural capital. According to the economic‐cultural‐capital configuration, sufficient financial resources assist high‐SES parents to provide their children with educational resources, termed cultural capital (Bourdieu, [<reflink idref="bib22" id="ref37">22</reflink>]; Choi et al., [<reflink idref="bib34" id="ref38">34</reflink>]; Hek & Kraaykamp, [<reflink idref="bib50" id="ref39">50</reflink>]; Katartzi & Hayward, [<reflink idref="bib58" id="ref40">58</reflink>]; Lancee & Werfhorst, [<reflink idref="bib61" id="ref41">61</reflink>]; Møllegaard & Jæger, [<reflink idref="bib75" id="ref42">75</reflink>]; Pishghadam, [<reflink idref="bib92" id="ref43">92</reflink>]; Xu & Hampden‐Thompson, [<reflink idref="bib117" id="ref44">117</reflink>]). Bourdieu ([<reflink idref="bib20" id="ref45">20</reflink>]) demarcates three forms of cultural capital—objectified, embodied and institutionalised—referring to substantial educational objects, manner/conduct and educational certificates, respectively. These three forms tend to be integrated into a single entity through schools, particularly reputed higher education institutes, in which the statuses of high‐brow culture, taste and logical engagement are cemented and aligned through school clubs and academic activities (Börjesson et al., [<reflink idref="bib11" id="ref46">11</reflink>]). This script reflects that schools function as a proxy linking scholastic capital to economic capital, enabling high‐SES groups' influence to penetrate throughout political, economic and cultural fields (Bourdieu, [<reflink idref="bib16" id="ref47">16</reflink>]), so when their culture acquires the status of orthodoxy (Bourdieu, [<reflink idref="bib12" id="ref48">12</reflink>]), the class power structure turns into a self‐proliferating circle. In this way, power transmission and social control are no longer reliant upon wealth but cultural capital.</p> <hd id="AN0187391082-5">Social space and habitus</hd> <p>Notwithstanding the cultural capital perspective's powerful insights into the phenomenon of cultural reproduction, its theories inhabit the macro level of social structure, without explaining how it occurs at the micro level of families. This omission led Bourdieu to propose two related notions in his late works—habitus and social space—flagging disposition and its development, respectively. Habitus functions to outline academic aptitudes, primarily resulting from cultural capital (Bourdieu, [<reflink idref="bib17" id="ref49">17</reflink>], [<reflink idref="bib21" id="ref50">21</reflink>]), because it is hatched within a specific type of social space, the features of which often are manufactured by parents' cultural capital, endorsing positive educational actions for their children. Since habitus, like dispositions registered within children's schemes of perception, judgement and action (Bourdieu, [<reflink idref="bib22" id="ref51">22</reflink>]), governs their familiarity with texts, including textbook contents and pedagogical information, it serves as a meter for predicting students' educational attainment, and becomes the primary channel for engendering cultural reproduction (Bourdieu, [<reflink idref="bib14" id="ref52">14</reflink>], [<reflink idref="bib18" id="ref53">18</reflink>]). Since economic capital is a prerequisite for gaining cultural capital, and rational minds aid high‐SES parents in securing this prerequisite (Ball et al., [<reflink idref="bib5" id="ref54">5</reflink>]; Reay, [<reflink idref="bib95" id="ref55">95</reflink>]), their children are retained in academic contexts through which they can build their academic habitus (Bourdieu, [<reflink idref="bib17" id="ref56">17</reflink>]), meeting the requirements of academic curricula, addressing logical reasoning competences (Bourdieu, [<reflink idref="bib14" id="ref57">14</reflink>]). Conversely, low‐SES students' habitus remains in the sphere of manual experiences, due to their parents' poor financial conditions. It is reported that this situation generates learning difficulties for this student group, so they need to spend extra time completing their degrees in postsecondary institutes (Zarifa et al., [<reflink idref="bib120" id="ref58">120</reflink>]). The correspondence between school curricula and academic habitus consequently benefits high‐SES pupils in obtaining much higher academic achievements than their low‐SES student counterparts (Bourdieu, [<reflink idref="bib14" id="ref59">14</reflink>], [<reflink idref="bib19" id="ref60">19</reflink>], [<reflink idref="bib22" id="ref61">22</reflink>]). In a word, the difference between academic and manual habitus is fashioned by the economic‐cultural‐capital architecture, in which high‐SES parents are apt to convert economic capital into cultural capital by virtue of their income advantage (Blanden & Gregg, [<reflink idref="bib9" id="ref62">9</reflink>]; Bourdieu, [<reflink idref="bib21" id="ref63">21</reflink>]; Covay & Carbonaro, [<reflink idref="bib36" id="ref64">36</reflink>]; Jæger, [<reflink idref="bib54" id="ref65">54</reflink>]; Katartzi & Hayward, [<reflink idref="bib58" id="ref66">58</reflink>]; Lancee & Werfhorst, [<reflink idref="bib61" id="ref67">61</reflink>]; Lareau, [<reflink idref="bib62" id="ref68">62</reflink>]; Reay, [<reflink idref="bib95" id="ref69">95</reflink>]).</p> <hd id="AN0187391082-6">Critiques on the perspective of cultural reproduction</hd> <p>Basically, Bourdieu adopts structuralism to fabricate his cultural capital and habitus theories. Cultural capital, for example, is distributed according to one's position in the social division of labour (Bourdieu, [<reflink idref="bib22" id="ref70">22</reflink>]). A one‐way convertible relation from economic capital to cultural capital also echoes this structural characteristic because it is determined by social class (Bourdieu, [<reflink idref="bib15" id="ref71">15</reflink>]). This structuralism can also be clarified through the features of habitus and its development, since they are the result of complying with the imperatives of the context in which its possessors are situated (Bourdieu, [<reflink idref="bib17" id="ref72">17</reflink>], [<reflink idref="bib22" id="ref73">22</reflink>]). Another case is the operation of habitus, which regularly takes place unconsciously. Although rationality may arise when the gap between social spaces cannot support the existing habitus and may, in turn, revive its owner's analytical faculties to engage in habitus evolution (Bourdieu, [<reflink idref="bib22" id="ref74">22</reflink>]; Bourdieu & Wacquant, [<reflink idref="bib24" id="ref75">24</reflink>]; Davey, [<reflink idref="bib38" id="ref76">38</reflink>]; Jin & Ball, [<reflink idref="bib56" id="ref77">56</reflink>]; Lehmann, [<reflink idref="bib64" id="ref78">64</reflink>]; Reay et al., [<reflink idref="bib96" id="ref79">96</reflink>]), this situation rarely occurs. That is to say, habitus constantly instructs actors to unconsciously carry out the structural features of the context/social space in which they grow up. That is why the habitus type mirrors its holder's social origin (Bourdieu, [<reflink idref="bib17" id="ref80">17</reflink>], [<reflink idref="bib19" id="ref81">19</reflink>], [<reflink idref="bib22" id="ref82">22</reflink>]). The above scenarios imply that whereas Bourdieusian academics have unmasked how the power structure is relayed and regenerated through the implicit linkage between habitus and cultural capital, their structural orbit overemphasises the idea that cultural reproduction is an unavoidable consequence. This has evoked the critique that schools possess not only the potential to accelerate educational inequity but also to ameliorate it, as manifested in the phenomenon of summer loss—the increase in inequality between high‐SES and low‐SES students that is observed after summer holidays. This implies that schools do not worsen the SES gap in cognitive skills, but actually reduce this inequality by compensating for the learning opportunities of low‐SES families (Downey & Condron, [<reflink idref="bib41" id="ref83">41</reflink>]; von Hippel et al., [<reflink idref="bib112" id="ref84">112</reflink>]). While the above studies highlight that schools may no longer serve as a repressive apparatus in terms of cultural reproduction but as a social equaliser in progressing low‐SES students' cognitive skills, a more effective method for improving their learning outcomes through compensatory cultural capital systematically implemented by schools has rarely been discussed. This method is likely anchored in the structural perspective of cultural capital, which largely overlooks the influence of agency on people's actions and consequences, mediated by such ingredients as visions, determined will (Chiang et al., [<reflink idref="bib31" id="ref85">31</reflink>]), capabilities (MacKenzie et al., [<reflink idref="bib71" id="ref86">71</reflink>]) and planned actions (Tran, [<reflink idref="bib107" id="ref87">107</reflink>]). This situation implies that the boundary between structure and agency is not absolute, and the two may be compatible to some extent (Giddens, [<reflink idref="bib45" id="ref88">45</reflink>], [<reflink idref="bib46" id="ref89">46</reflink>]). In this sense, interactions between structure and agency may constantly occur, and then engender new social structures that not only instruct our minds but also stimulate our creative reactions towards their structural constraints (Archer, [<reflink idref="bib3" id="ref90">3</reflink>]) through internal conservation (Archer, [<reflink idref="bib4" id="ref91">4</reflink>]), self‐understanding and creative plans, as evident in the notion of identity development (Holland et al., [<reflink idref="bib52" id="ref92">52</reflink>]). Related studies further reveal that the practice of agency is heavily managed by educational training because it can nourish people's interpretative rationality (Beck & Beck‐Gernsheim, [<reflink idref="bib6" id="ref93">6</reflink>]; Notten et al., [<reflink idref="bib78" id="ref94">78</reflink>]) and, in turn, spur visionary agency. This is often exerted by high‐SES parents and enables educational resources, ideas and methods to be integrated into systematic action plans (Chiang, Toh et al., [<reflink idref="bib32" id="ref95">32</reflink>]). In contrast, educational resources in low‐SES families are insufficient, so if children in such families want to better their academic achievements, they need to mobilise this agency to search for educational resources available in educational institutes, rather than in their family social space. It was reported that excellent low‐SES students behaved as cultural capital constructors who were skilful in accessing educational resources not available in their primary social spaces (families) but in their secondary social spaces (e.g., school libraries, community libraries and significant others who offered them aids and resources) (Chiang et al., [<reflink idref="bib31" id="ref96">31</reflink>]). This denotes that low‐SES students are situated in dual forms of social space, namely primary family social spaces and secondary social spaces, particularly schools, which are significant due to their rich educational resources (Chiang & Trezise, [<reflink idref="bib33" id="ref97">33</reflink>]). If these students can exercise agency to mobilise such resources, they will no longer be constrained by cultural capital deficits in their primary social spaces. Disappointingly, such individualised agency cannot generally enable this student group to reap the advantages of the educational resources available in school social spaces because they are in the minority. Fortunately, the implication of dual forms of social space offers insights for improving cultural reproduction because the structural constraints of low‐SES family social spaces can be reduced when educational resources are unpacked and properly used by schools. To achieve this purpose requires remoulding the individualised agency into an institutionalised form, empowering schools to transform their educational resources into compensatory cultural capital for low‐SES students.</p> <p>Schools are the leading sites for allocating educational resources among these institutes, so they essentially enact another form of social space that can compensate for the resource constraints of the low‐SES family social space. In a sense, this framework may explain the phenomenon of summer loss. This suggests that schools do not worsen the SES gap in cognitive skills, but actually reduce this inequality by compensating for the cultural deficit of low‐SES families (Downey & Condron, [<reflink idref="bib41" id="ref98">41</reflink>]; von Hippel et al., [<reflink idref="bib112" id="ref99">112</reflink>]). At any rate, this possible compensation can be achieved only if educational resources are unpacked and properly used, as evidenced by the case of excellent working‐class students who behave as cultural capital constructors by accessing these resources actively through their individualised agency (Chiang et al., [<reflink idref="bib31" id="ref100">31</reflink>]). Disappointingly, as such students are in the minority, this kind of compensation hasn't yet been scientifically realised, as can be observed through the continuing prevalence of cultural reproduction.</p> <p>This circumstance calls for organised plans or programmes that effectuate this redemptive function, availing low‐SES students of the opportunity for improved learning outcomes. Related studies attest the status of reading as a form of embedded cultural capital in this regard. It is argued that written language can be translated into spoken language because reading comprehension requires decoding and language comprehension skills, which involve the activation of phonological and semantic pathways, and reading plays a key role in this translation (Hulme & Snowling, [<reflink idref="bib53" id="ref101">53</reflink>]). This theory has been well documented by relevant research, as shown in the findings that phonemic awareness, letter‐sound knowledge and rapid automatised naming serve as cognitive preconditions for learning to read (Dickinson & Anastasopoulos, [<reflink idref="bib40" id="ref102">40</reflink>]), along with language comprehension skills composed of vocabulary knowledge, syntactical knowledge and narrative awareness (Hulme & Snowling, [<reflink idref="bib53" id="ref103">53</reflink>]; Whorrall & Cabell, [<reflink idref="bib116" id="ref104">116</reflink>]). When linguistic comprehension gears reading comprehension, oral language (vocabulary) and decoding skills can be acquired through the intervention of phonology with reading (Bowyer‐Crane et al., [<reflink idref="bib25" id="ref105">25</reflink>]). This doctrine is consistent with other findings that engagement with reading significantly levels up vocabulary (Sullivan & Brown, [<reflink idref="bib103" id="ref106">103</reflink>]), boosting literacy ability and literacy skills (Buchon‐Stoll, [<reflink idref="bib26" id="ref107">26</reflink>]; Griffin et al., [<reflink idref="bib47" id="ref108">47</reflink>]; Silverman, [<reflink idref="bib102" id="ref109">102</reflink>]; Thurston, Cockerill et al., [<reflink idref="bib105" id="ref110">105</reflink>]; Wasik et al., [<reflink idref="bib114" id="ref111">114</reflink>]). It has been reported that the returns of reading on vocabulary progression are much higher than on mathematics attainment (Sullivan & Brown, [<reflink idref="bib103" id="ref112">103</reflink>]), so it can be inferred that vocabulary proficiency heavily governs literacy ability, which is a core competency for answering mathematics test questions correctly (Fuchs et al., [<reflink idref="bib43" id="ref113">43</reflink>]). This inference can be validated through rules of recognition and realisation, respectively, referring to the ability to understand textual meanings and respond to their demands correctly (Bernstein, [<reflink idref="bib7" id="ref114">7</reflink>]), as evidenced in mathematics tests (Chiang, Thurston & Cockerill, [<reflink idref="bib30" id="ref115">30</reflink>]; Chiang et al., [<reflink idref="bib29" id="ref116">29</reflink>]). Furthermore, science performance is reported to be highly affected by students' reading comprehension. In a sample of 2051 German Grade 10 and 11 students, after controlling for cognitive ability, gender, family home language and school track, multi‐level logistic regression indicated that there was a significant interaction between students' reading comprehension and science performance (Neri et al., [<reflink idref="bib77" id="ref117">77</reflink>]). Other researchers have reported that the skills required by scientists rely heavily on reading ability, and education systems must ensure students are literate enough to learn science (Yore et al., [<reflink idref="bib118" id="ref118">118</reflink>]). Failure to develop effective literacy and reading skills has been reported to result in underachievement in science by the end of primary school, and is still evident by the end of secondary school. A large‐scale secondary data analysis of over 640,000 students in English schools reported that gaps in science attainment, which could be linked to poor reading attainment, opened up by the age of 7 and widened as students approached school‐leaving age. By the time they left school, 37% of science performance variance could be explained by reading attainment (Nunes et al., [<reflink idref="bib79" id="ref119">79</reflink>]). Previous research has indicated that these relationships also exist within previous analyses of Programme for International Student Assessment (PISA) data. Analysis of the PISA 2000 dataset showed a 0.84 correlation between reading performance and science performance; the correlation was 0.81 for the 2003 dataset and 0.82 for the 2006 dataset (Cromley, [<reflink idref="bib37" id="ref120">37</reflink>]).</p> <p>Unfortunately, students demonstrate highly differentiated literacy abilities, which are mainly brought about by their family language environments and shaped by parents' attitudes towards and engagement in reading (Guevara et al., [<reflink idref="bib49" id="ref121">49</reflink>]; Vivas, [<reflink idref="bib111" id="ref122">111</reflink>]). High‐SES students dwell in an academic‐like environment in which books (Xu & Hampden‐Thompson, [<reflink idref="bib117" id="ref123">117</reflink>]) and reading activities (Bodovski et al., [<reflink idref="bib10" id="ref124">10</reflink>]; Loh & Sun, [<reflink idref="bib70" id="ref125">70</reflink>]) have been systematically supplied by their parents, so they often develop strong reading competencies (Araujo & Costa, [<reflink idref="bib2" id="ref126">2</reflink>]; Bodovski et al., [<reflink idref="bib10" id="ref127">10</reflink>]; Loh & Sun, [<reflink idref="bib70" id="ref128">70</reflink>]), which are the essential premise for understanding academic curriculum knowledge. In contrast, such activities are uncommon in low‐SES families, due to the shortage of economic capital noted previously, so most low‐SES children fail in school academically (Mikus et al., [<reflink idref="bib73" id="ref129">73</reflink>]). In response to the uneven distribution of home reading resources, which consolidates inequity in educational results, reading intervention programmes are expected to actualise a compensatory function. In fact, the positive returns of such interventions have been authenticated, as shown in the visible progressions in language abilities (Wasik et al., [<reflink idref="bib114" id="ref130">114</reflink>]), literacy knowledge (Buchon‐Stoll, [<reflink idref="bib26" id="ref131">26</reflink>]; Silverman, [<reflink idref="bib102" id="ref132">102</reflink>]), comprehension abilities (Thurston, Cockerill et al., [<reflink idref="bib105" id="ref133">105</reflink>]) and scientific knowledge (Pollard‐Durodola et al., [<reflink idref="bib93" id="ref134">93</reflink>]) of students who have participated in them.</p> <hd id="AN0187391082-7">RESEARCH DESIGN</hd> <p></p> <hd id="AN0187391082-8">Research questions and hypotheses</hd> <p>As outlined above, theories of cultural capital unveil how the phenomenon of cultural reproduction is rooted in low‐SES family social spaces, in which the shortage of educational resources constrains the children from developing academic habitus due to their family's poor economic capital. However, the structural orbit of these theories largely neglects the influence of agency on educational results, which may assist schools to improve the phenomenon of cultural reproduction. Unfortunately, while schools are the leading sites of distribution of educational resources, the advantages of such resources remain as embedded cultural capital because no systematic schemes have been implemented for turning this embedded cultural capital into a compensatory form for low‐SES students. Since excellent low‐SES students exercise agency to transform this embedded cultural capital into an active form, compensating for cultural capital deficits in their family social space, compensatory cultural capital can be systematically generated for low‐SES students when schools actively practice agency by unleashing the advantages of their educational resources, particularly the power of reading, which can contribute significantly to students' cognitive skills. These correlations suggest that if agency is properly performed through reading programmes in schools, embedded cultural capital can evolve into compensatory cultural capital effectively supplementing the paucity of reading resources/activities in low‐SES family social space. That is to say, if schools can actively bolster this student group with educational resources unavailable in their family social spaces, this agency will bring them a great harvest in terms of academic performance. The activation of this agency is highly feasible if schools systematically introduce reading programmes/interventions, which may serve to optimise the compensatory function of embedded cultural capital, in the form of reading resources, plans, activities and involvements inscribed within a school social space, such as the one described in this study. Such interventions assume that schools can perform as social equalisers rather than repressive apparatuses if the advantages of reading are unlocked in school. From this viewpoint, reading is defined in this study as a portal for improving low‐SES students' academic attainments and hence ameliorating the phenomenon of cultural reproduction, rather than a means to effect cognitive development. To examine whether schools can perform as great equalisers, at least two research questions need to be investigated, as follows:</p> <p></p> <ulist> <item> RQ1 What are the relations between students' scores in reading and those in mathematics and science on PISA 2018?</item> <p></p> <item> RQ2 What are the contributions of identified independent variables to students' reading scores on PISA 2018?</item> </ulist> <p>To answer the above two research questions, two paired hypotheses are proposed:</p> <hd id="AN0187391082-9">H1</hd> <p>Students' scores in reading on PISA 2018 effectively predict those in mathematics and science.</p> <hd id="AN0187391082-10">H2</hd> <p>Students' reading scores on PISA 2018 are significantly regulated by such independent variables as gender (H2a), economic, social and cultural status (ESCS) (H2b), cultural capital (H2c), reading engagement (H2d), teacher support in test language lessons (TEACHSUP) (H2e), teacher‐directed instruction (DIRINS) (H2f), teacher's stimulation of reading engagement as perceived by students (STIMREAD) (H2g), meta‐cognition: understanding and remembering (H2h), meta‐cognition: summarising (H2i), meta‐cognition: assess credibility (H2j) and time spent reading (H2k) on the basis of data collected in the PISA 2018 tests.</p> <hd id="AN0187391082-11">Subjects</hd> <p>The PISA 2018 dataset (OECD, [<reflink idref="bib83" id="ref135">83</reflink>]) is our data source for testing the above hypotheses due to its crucial advantages, including its excellent international reputation, quantitative data and systematically stratified random sampling, which is based on two steps—selection of schools according to PPS (probabilities proportional to their size) and then selection of students within the schools. This two‐stage sampling design calls for the use of student final weights and 80 replicate weights for calculating standard errors (SE) (OECD, [<reflink idref="bib80" id="ref136">80</reflink>]). Additionally, although PISA is comprised of three disciplines, namely science, mathematics and reading, reading literacy was the main domain of PISA 2018—the seventh assessment. These features render the PISA data perfectly suited to the purposes of this study.</p> <p>PISA tests, offered by the OECD (Organisation for Economic Co‐operation and Development) in a 3‐year cycle since 2000, set out to assess the knowledge and skills in reading, mathematics and science that are required by modern society and expected to have been learned by 15‐year‐old students (OECD, [<reflink idref="bib84" id="ref137">84</reflink>]: 3). The students from Taiwan who participated in PISA 2018 were our subjects. Taiwanese students have participated in six cycles of the PISA surveys and achieved leading results in mathematics and science tests since 2006. However, their attainment in reading tests has been much lower. This unsatisfactory situation can be attributed to disparities between schools caused by social stratification, which can affect students' learning opportunities and thus their learning outcomes. Education systems with lower inter‐school disparities often yield higher educational equity, as exemplified in the case of Finland, in which the inter‐school disparity in reading scores accounts for less than 10% of the total national variance (OECD, [<reflink idref="bib87" id="ref138">87</reflink>]). In Taiwan, most 15‐year‐old students belong to the 9th and 10th grades. About two‐thirds of them enrol in the 10th grade, which is part of the senior high school system, and for which students are selected through national examinations. The results of these examinations strongly influence participants' choice of either academic or vocational tracks. Other 15‐year‐old students are enrolled in the 9th grade, which is part of the junior high school system. Attendance at this level is mandatory in Taiwan, and government schools are required by law to recruit students based on residential districts. However, this policy doesn't apply to private schools and their score‐based approach has strongly attracted many middle‐class parents. These characteristics of the local system mean that students' learning outcomes correspond closely to schools' reputations, which, in a way, echoes the system of stratified schools. Regarding the sampling method, the schools in Taiwan were categorised into 36 strata, of which 192 schools were selected and 7243 respondents were randomly sampled from an estimated total population of 226,698 students. In PISA 2018, the mean scores in reading, mathematics and science were 487, 489 and 489, respectively. Proficiency scores in reading, science and mathematics were classified into six levels. However, level 1 of reading was further divided into three sub‐levels (OECD, [<reflink idref="bib86" id="ref139">86</reflink>]).</p> <hd id="AN0187391082-12">Variables</hd> <p>In the first stage, which probed the contribution of reading to mathematics and science, students' scores in reading on PISA 2018 were defined as an independent variable for predicting the dependent variables, which were their scores in mathematics and science. As individual student participants of PISA were required to answer different combinations of questionnaire items, 10 PVs (plausible values) were designed to indicate the range of possible values by which a student's ability could be identified (OECD, [<reflink idref="bib80" id="ref140">80</reflink>]: 96). In this regard, according to OECD ([<reflink idref="bib85" id="ref141">85</reflink>]) and statisticians (Rutkowski et al., [<reflink idref="bib97" id="ref142">97</reflink>]), the estimation of each PISA scale, the handling of variables and weights are required for replicate analysis and plausible value analysis in secondary data analysis. This approach is crucial because it allows for a nuanced understanding of the reading proficiency levels of students as assessed by the PISA framework. The reading scales were thus analysed 10 times separately, with each iteration corresponding to one of the plausible values generated within the dataset. The results from these iterations were averaged to produce a final set of outcomes. The above processes help mitigate any anomalies that may arise from individual analyses, thereby providing a more reliable estimate of the underlying trends in the data.</p> <p>We also calculated sampling weights because any significant tests performed on the results necessitated careful adjustments. This adjustment process accounted for the inherent variations that might occur across the initial 10 sets of results. Such adjustments are vital in ensuring that the findings are statistically valid and reflect actual differences rather than fluctuations due to sampling variability (Laukaityte & Wiberg, [<reflink idref="bib63" id="ref143">63</reflink>]). According to the PISA 2015 technical report (OECD, [<reflink idref="bib81" id="ref144">81</reflink>]), within‐school weights correspond to student final weights (variable W_FSTUWT), rescaled to amount to the sample size within each school. Between‐school weights correspond to the sum of student final weights within each school. Therefore, in this study, all analyses were conducted with careful consideration of the data weighting, specifically utilising the adjusted student weights identified by the variable W_FSTUWT. This approach is critical for ensuring that the results are representative of the broader population, as these weights adjust for sampling design and non‐response biases (Little & Rubin, [<reflink idref="bib67" id="ref145">67</reflink>]).</p> <p>As the second stage sought to specify what factors contributed to student performance in the reading test (dependent variable) of PISA 2018, the technique of HLM (hierarchical linear modelling) was performed in this study, which is perfectly suited for educational data with the feature of hierarchical structures (Raudenbush & Bryk, [<reflink idref="bib94" id="ref146">94</reflink>]). More concretely, the within‐school weights were defined as the student weights that had been rescaled. This rescaling was essential to reflect the sample size within individual schools accurately and then enhance the validity of the findings. Conversely, the between‐school weights were calculated as the aggregated sum of student weights from each school, meaning that the above analysis accounted for the individual student weights and the collective representation of students across different schools, which is crucial for understanding broader trends and outcomes in educational performance (OECD, [<reflink idref="bib81" id="ref147">81</reflink>]).</p> <p>Considering the complex sampling design, we employed the technique of balanced repeated replication (BRR), or balanced half‐samples used for calculating sampling variances for PISA estimates. A particular variant known as Fay's method was used. The major advantage of BRR over the jackknife method is that the jackknife is inappropriate for use with non‐differentiable functions of the quantified survey data, and it is thus difficult to ensure accurate estimators of variance (Judkins, [<reflink idref="bib57" id="ref148">57</reflink>]). More specifically, the BRR employed by this study can yield more reliable estimates by utilising the 80 replicate weights that are included in the PISA dataset. By leveraging these replicate weights, we were able to conduct our analyses with enhanced precision, thereby improving the reliability of the results. The variance estimator for a specific analysed statistic, named <emph>X</emph>*, from the full sample is given as <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0001" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>V</mi><mi>BRR</mi></msub><mfenced close=")" open="("><msup><mi>X</mi><mo>*</mo></msup></mfenced><mo linebreak="goodbreak">=</mo><mn>0.05</mn><munderover><mo>∑</mo><mrow><mi>t</mi><mo>=</mo><mn>1</mn></mrow><mn>80</mn></munderover><mfenced close="}" open="{"><msup><mfenced close=")" open="("><mrow><msubsup><mi>X</mi><mi>t</mi><mo>*</mo></msubsup><mo linebreak="goodbreak">−</mo><msup><mi>X</mi><mo>*</mo></msup></mrow></mfenced><mn>2</mn></msup></mfenced></mrow></semantics></math> </ephtml> with <emph>t</emph> = 1, ..., 80 being the number of replicates. <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0002" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msubsup><mi>X</mi><mi>t</mi><mo>*</mo></msubsup></mrow></semantics></math> </ephtml> results in the <emph>t</emph>‐th estimation of this statistic with the <emph>t</emph>‐th replication weights combination.</p> <p>The BRR method allows for the estimation of variances in a way that accounts for the stratification and clustering inherent in the sample design. This is very important in educational research, where data are often collected in nested structures, such as students within schools. Implementing the BRR method ensures that the estimates derived from our analyses are robust and reflective of the true characteristics of the population being studied. In this analysis, gender, ESCS, cultural capital, reading engagement, TEACHSUP, DIRINS, STIMREAD, meta‐cognition: understanding and remembering, meta‐cognition: summarising, meta‐cognition: assess credibility and time spent reading were selected as independent variables. In terms of gender, there were only minor differences between the percentages of male and female subjects, as shown in Table 1.</p> <p>1 TABLE The number and percentage of participants by gender in Taiwan.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Gender</th><th align="left">Respondent</th><th align="left">Estimated number of student population aged 15</th></tr><tr><th align="left"><italic>N</italic></th><th align="left">%</th><th align="left"><italic>n</italic></th><th align="left">%</th></tr></thead><tbody valign="top"><tr><td align="left">Female</td><td align="char" char=".">3624</td><td align="char" char=".">50.0</td><td align="char" char=".">112,859</td><td align="char" char=".">49.8</td></tr><tr><td align="left">Male</td><td align="char" char=".">3619</td><td align="char" char=".">50.0</td><td align="char" char=".">113,839</td><td align="char" char=".">50.2</td></tr></tbody></table> </ephtml> </p> <p>The OECD‐formulated ESCS scale is comprised of parental qualifications, occupations and home possessions, which include study space and available educational resources, such as internet and books. In order to achieve the same weights on the scale across PISA member countries/regions, its standard deviation (SD) and mean are set at 1 and 0, respectively (OECD, [<reflink idref="bib87" id="ref149">87</reflink>]: 52). The constitution of other independent variables is detailed in Table 2. As the construct of cultural capital was comprised of selected items from PISA 2018, it is necessary to rationalise its composition. According to Bourdieu, there are three forms of cultural capital: objectified, embodied and institutionalised. However, related test items on PISA 2018 alluded only to the objectified and embodied forms. More specifically, as demonstrated in Table 2, ST011Q10 (books to help with your school work), ST011Q11 (technical reference books), ST011Q12 (a dictionary) and ST013Q01 (the number of books) are associated with the objectified type, while ST011Q07 (classic literature), ST011Q08 (books of poetry), ST011Q09 (works of art), ST011Q16 (books on art, music or design) and ST012Q09 (musical instruments) focus on the embodied genre (OECD, [<reflink idref="bib88" id="ref150">88</reflink>]). In order to meet the definition of cultural capital, parental qualifications (ST005, ST006, ST007 and ST008), viewed as the institutionalised form of cultural capital, were included in this composition.</p> <p>2 TABLE The question items of PISA 2018 associated with independent variables.</p> <p> <ephtml> <table><tbody valign="top"><tr><td align="left">Cultural capitalCronbach's alpha 0.74</td></tr><tr><td align="left">ST011</td><td align="left">Which of the following are in your home?</td><td align="left">Q07</td><td align="left">Classic literature (e.g., Shakespeare)</td></tr><tr><td align="left">Q08</td><td align="left">Books of poetry</td></tr><tr><td align="left">Q09</td><td align="left">Works of art (e.g., paintings)</td></tr><tr><td align="left">Q10</td><td align="left">Books to help with your school work</td></tr><tr><td align="left">Q11</td><td align="left">Technical reference books</td></tr><tr><td align="left">Q12</td><td align="left">A dictionary</td></tr><tr><td align="left">Q16</td><td align="left">Books on art, music or design</td></tr><tr><td align="left">ST012</td><td align="left">How many of these are there at your home?</td><td align="left">Q09</td><td align="left">Musical instruments (e.g., guitar, piano)</td></tr><tr><td align="left">ST013</td><td align="left">How many books are there in your home?</td><td align="left">Q01</td><td align="left">0–10, 11–25, 26–100, 101–200, 201–500, more than 500 books</td></tr><tr><td align="left">Reading engagementCronbach's alpha 0.84</td></tr><tr><td align="left">ST160</td><td align="left">How much do you agree or disagree with these statements about reading?</td><td align="left">Q01</td><td align="left">I read only if I have to</td></tr><tr><td align="left">Q02</td><td align="left">Reading is one of my favourite hobbies</td></tr><tr><td align="left">Q03</td><td align="left">I like talking about books with other people</td></tr><tr><td align="left">Q04</td><td align="left">For me, reading is a waste of time</td></tr><tr><td align="left">Q05</td><td align="left">I read only to get information that I need</td></tr><tr><td align="left">Teacher support in test language lessonsCronbach's alpha 0.90</td></tr><tr><td align="left">ST100</td><td align="left">How often do these things happen in your test language lessons?</td><td align="left">Q01</td><td align="left">The teacher shows an interest in every student's learning</td></tr><tr><td align="left">Q02</td><td align="left">The teacher gives extra help when students need it</td></tr><tr><td align="left">Q03</td><td align="left">The teacher helps students with their learning</td></tr><tr><td align="left">Q04</td><td align="left">The teacher continues teaching until the students understand</td></tr><tr><td align="left">Teacher‐directed instructionCronbach's alpha 0.86</td></tr><tr><td align="left">ST102</td><td align="left">How often do these things happen in your test language lessons?</td><td align="left">Q01</td><td align="left">The teacher sets clear goals for our learning</td></tr><tr><td align="left">Q02</td><td align="left">The teacher asks questions to check whether we have understood what was taught</td></tr><tr><td align="left">Q03</td><td align="left">At the beginning of a lesson, the teacher presents a short summary of the previous lesson</td></tr><tr><td align="left">Q04</td><td align="left">The teacher tells us what we have to learn</td></tr><tr><td align="left">Teacher's stimulation of reading engagement as perceived by studentsCronbach's alpha 0.91</td></tr><tr><td align="left">ST152</td><td align="left">In your test language lessons, how often does the following occur?</td><td align="left">Q05</td><td align="left">The teacher encourages students to express their opinion about a text</td></tr><tr><td align="left">Q06</td><td align="left">The teacher helps students relate the stories they read to their lives</td></tr><tr><td align="left">Q07</td><td align="left">The teacher shows students how the information in texts builds on what they already know</td></tr><tr><td align="left">Q08</td><td align="left">The teacher poses questions that motivate students to participate actively</td></tr><tr><td align="left">Meta‐cognition: understanding and rememberingCronbach's alpha is inapplicable due to this not being an interval scale</td></tr><tr><td align="left">ST164</td><td align="left">How do you rate the usefulness of the following strategies for understanding and memorising the text?</td><td align="left">Q01</td><td align="left">I concentrate on the parts of the text that are easy to understand</td></tr><tr><td align="left">Q02</td><td align="left">I quickly read through the text twice</td></tr><tr><td align="left">Q03</td><td align="left">After reading the text, I discuss its content with other people</td></tr><tr><td align="left">Q04</td><td align="left">I underline important parts of the text</td></tr><tr><td align="left">Q05</td><td align="left">I summarise the text in my own words</td></tr><tr><td align="left">Q06</td><td align="left">I read the text aloud to another person</td></tr><tr><td align="left">Meta‐cognition: summarisingCronbach's alpha is inapplicable due to this not being an interval scale</td></tr><tr><td align="left">ST165</td><td align="left">How do you rate the usefulness of the following strategies for writing a summary of this two‐page text?</td><td align="left">Q01</td><td align="left">I write a summary then I check that each paragraph is covered in the summary, because the content of each paragraph should be included</td></tr><tr><td align="left">Q02</td><td align="left">I try to copy out accurately as many sentences as possible</td></tr><tr><td align="left">Q03</td><td align="left">Before writing the summary, I read the text as many times as possible</td></tr><tr><td align="left">Q04</td><td align="left">I carefully check whether the most important facts in the text are represented in the summary</td></tr><tr><td align="left">Q05</td><td align="left">I read through the text, underlining the most important sentences. Then I write them in my own words as a summary</td></tr><tr><td align="left">Meta‐cognition: assess credibilityCronbach's alpha is inapplicable due to this not being an interval scale</td></tr><tr><td align="left">ST166</td><td align="left">In your opinion, how appropriate are the following strategies in reaction to this email?</td><td align="left">Q01</td><td align="left">Answer the email and ask for more information about the smartphone</td></tr><tr><td align="left">Q02</td><td align="left">Check the sender's email address</td></tr><tr><td align="left">Q03</td><td align="left">Click on the link to fill out the form as soon as possible</td></tr><tr><td align="left">Q04</td><td align="left">Delete the email without clicking on the link</td></tr><tr><td align="left">Q05</td><td align="left">Check the website of the mobile phone operator to see whether the smartphone offer is mentioned</td></tr><tr><td align="left">Time spent readingCronbach's alpha is inapplicable due to this not being an interval scale</td></tr><tr><td align="left">ST175</td><td align="left">About how much time do you usually spend reading for enjoyment?</td><td align="left">Q01</td><td align="left">(1) I do not read for enjoyment(2) 30 min or less a day(3) More than 30 min to less than 60 min a day(4) 1–2 h a day(5) More than 2 h a day</td></tr></tbody></table> </ephtml> </p> <p>Regarding coding, questions ST005 and ST006 were designed to measure mothers' highest qualifications, which were categorised into seven scales, coded from 0 to 6 (0: None, 1: ISCED 1, 2: ISCED 2, 3: ISCED 3B, 4: ISCED 3A/4, 5: ISCED 5B, 6: ISCED 5A/6). This coding method was also applied to questions ST006 and ST008, which assessed fathers' highest qualifications. All items of question ST011 were dichotomous (yes or no), so 1 was for yes and 0 for no. Questions ST012 and ST013 were polytomous items with four (none, one, two, three or more) and six (0–10, 11–25, 26–100, 101–200, 201–500, more than 500) categories, and they were coded from 0 to 3 and 0 to 5, respectively. Based on IRT (item response theory) scaling methodology, questions ST160, ST100, ST102 and ST152 were set as four scale items. The method of reverse scoring was applied to questions ST100, ST102, ST160Q01, ST160Q04 and ST160Q05. Questions ST164, ST165 and ST166 had six‐point scales, from 1 to 6. Each of the student responses to these questions needed to be validated by the approval of 80% of PISA's experts. Ratios were further calculated on the basis of predetermined criteria. For instance, the criteria of ST166 were Q02 > Q01, Q02 > Q03, Q04 > Q01, Q04 > Q03, Q05 > Q01 and Q05 > Q03, so if a student's responses to ST166 only met three of these criteria, his/her score was 0.5. Students were weighted using the final student weight (W_FSTUWT) (OECD, [<reflink idref="bib88" id="ref151">88</reflink>]).</p> <p>The data were analysed in SPSS for Windows Release 23 (IBM, Armonk, USA) to produce basic descriptive statistics and regression analysis, and in HLM6 (Scientific Software International) to conduct the HLM analysis. Table 3 details descriptive statistics of the independent variables noted above.</p> <p>3 TABLE Descriptive statistics for independent variables of student scores in the reading of PISA 2018.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left" /><th align="left"><italic>N</italic></th><th align="left">Mean</th><th align="left">SD</th></tr></thead><tbody valign="top"><tr><td align="left">Reading literacy</td><td align="char" char=".">7243</td><td align="char" char=".">502.60</td><td align="char" char=".">101.73</td></tr><tr><td align="left">Mathematics literacy</td><td align="char" char=".">7243</td><td align="char" char=".">531.14</td><td align="char" char=".">99.70</td></tr><tr><td align="left">Science literacy</td><td align="char" char=".">7243</td><td align="char" char=".">515.75</td><td align="char" char=".">99.28</td></tr><tr><td align="left">ESCS</td><td align="char" char=".">7172</td><td align="char" char=".">−0.32</td><td align="char" char=".">0.92</td></tr><tr><td align="left">Cultural capital</td><td align="char" char=".">7004</td><td align="char" char=".">0.02</td><td align="char" char=".">0.99</td></tr><tr><td align="left">Reading engagement</td><td align="char" char=".">7152</td><td align="char" char=".">0.34</td><td align="char" char=".">0.93</td></tr><tr><td align="left">Teacher support in test language lessons</td><td align="char" char=".">7162</td><td align="char" char=".">0.04</td><td align="char" char=".">0.92</td></tr><tr><td align="left">Teacher‐directed instruction</td><td align="char" char=".">7161</td><td align="char" char=".">0.08</td><td align="char" char=".">1.04</td></tr><tr><td align="left">Teacher's stimulation of reading engagement</td><td align="char" char=".">7155</td><td align="char" char=".">0.14</td><td align="char" char=".">1.01</td></tr><tr><td align="left">Meta‐cognition: understanding and remembering</td><td align="char" char=".">6972</td><td align="char" char=".">−0.38</td><td align="char" char=".">0.91</td></tr><tr><td align="left">Meta‐cognition: summarising</td><td align="char" char=".">7010</td><td align="char" char=".">−0.51</td><td align="char" char=".">0.99</td></tr><tr><td align="left">Meta‐cognition: assess credibility</td><td align="char" char=".">7032</td><td align="char" char=".">−0.35</td><td align="char" char=".">1.00</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note</emph>: Because the technique of HLM automatically discards those cases with missing values, the number of subjects constantly changed for individual variables.</p> <hd id="AN0187391082-13">Mathematical expressions of the HLM model</hd> <p>Model 1 is the null model, testing whether the technique of HLM is suitable for analysing the data noted previously, which may have the effect of nested data (Raudenbush & Bryk, [<reflink idref="bib94" id="ref152">94</reflink>]). The effect of the level 2 variables on the dependent variables is also equivalent to the intra‐class correlation coefficient (ICC). The ICC is the proportion of the variance in the dependent variable (students' scores in reading on PISA 2018) that is attributed to the school effect. It is calculated as the variance of the school effect <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0003" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msubsup><mi>δ</mi><mtext>cons</mtext><mn>2</mn></msubsup></mrow></semantics></math> </ephtml> divided by the variance of the total effect ( <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0004" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msubsup><mi>δ</mi><mtext>cons</mtext><mn>2</mn></msubsup><mo>+</mo><msubsup><mi>δ</mi><mtext>residual</mtext><mn>2</mn></msubsup></mrow></semantics></math> </ephtml> ). ICC, calculated as <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0005" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msubsup><mi>δ</mi><mrow><mi>μ</mi><mn>0</mn><mi>j</mi></mrow><mn>2</mn></msubsup><mo>/</mo><msubsup><mi>δ</mi><mrow><mi>μ</mi><mn>0</mn><mi>j</mi></mrow><mn>2</mn></msubsup><mo>+</mo><msubsup><mi>δ</mi><mi mathvariant="italic">rij</mi><mn>2</mn></msubsup></mrow></semantics></math> </ephtml> , is the key index to distinguish how much the variation of the reading score is nested at the school level or, say, between‐group effect. The null model is expressed as follows:</p> <hd1 id="AN0187391082-14">Model 1 (null model)</hd1> <p>Level 1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0006" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>READ</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">=</mo><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">+</mo><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml></p> <p>Level 2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0007" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <p> <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0008" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>j</mi></mrow></semantics></math> </ephtml> labels different schools; <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0009" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>i</mi></mrow></semantics></math> </ephtml> refers to different students; <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0010" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>READ</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> signifies the outcome variable (the PISA reading score of student <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0011" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>i</mi></mrow></semantics></math> </ephtml> in school <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0012" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>j</mi></mrow></semantics></math> </ephtml> ); <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0013" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> represents the mean of the students' reading scores in school <emph>j</emph>; <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0014" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> is the deviation of the reading score of student <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0015" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>i</mi></mrow></semantics></math> </ephtml> from school <emph>j</emph>; the variance of <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0016" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> expresses the random error at the student level (i.e., intra‐school variation); <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0017" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>γ</mi><mn>00</mn></msub></mrow></semantics></math> </ephtml> is associated with the grand mean (the mean of the scores of all the students); <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0018" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> details the difference between the grand mean and the school mean; and variance of <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0019" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> describes the random error at the school level (i.e., inter‐school variation).</p> <hd1 id="AN0187391082-15">Model 2</hd1> <p>Level 1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0020" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>READ</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">=</mo><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><msub><mtext>GENDER</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><msub><mtext>ESCS</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml></p> <p>Level 2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0021" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0022" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>10</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0023" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>20</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <hd1 id="AN0187391082-16">Model 3</hd1> <p>Level 1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0024" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>READ</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">=</mo><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><msub><mtext>GENDER</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><msub><mtext>ESCS</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><msub><mi>CC</mi><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml></p> <p>Level 2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0025" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0026" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>10</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0027" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>20</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0028" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>30</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <hd1 id="AN0187391082-17">Model 4</hd1> <p>Level 1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0029" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>READ</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">=</mo><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><msub><mtext>GENDER</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><msub><mtext>ESCS</mtext><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>β</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><msub><mi>CC</mi><mi mathvariant="italic">ij</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>r</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml></p> <p>Level 2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0030" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>02</mn></msub><msub><mtext>SCTSUP</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>03</mn></msub><msub><mtext>SCDIRIN</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>04</mn></msub><msub><mtext>SCSTIMR</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <p>(Model 4‐1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0031" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo>=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo>+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>γ</mi><mn>02</mn></msub><msub><mtext>SCTSUP</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> )</p> <p>(Model 4‐2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0032" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo>=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo>+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>γ</mi><mn>03</mn></msub><msub><mtext>SCDIRIN</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> )</p> <p>(Model 4‐3: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0033" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo>=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo>+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>γ</mi><mn>04</mn></msub><msub><mtext>SCSTIMR</mtext><mi>j</mi></msub><mo>+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml> ) <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0034" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>10</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0035" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>20</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0036" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>30</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <hd1 id="AN0187391082-18">Model 5</hd1> <p>Level 1: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0037" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML">READij=β0j+β1jGENDERij+β2jESCSij+β3jCCij+β4jENGAGEij+β5jTIMEij+β6jUNDREMij+β7jMETASUMij+β8jMETASPAMij+rij</math> </ephtml></p> <p>Level 2: <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0038" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>00</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>01</mn></msub><msub><mtext>SCESCS</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>02</mn></msub><msub><mtext>SCTSUP</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>03</mn></msub><msub><mtext>SCDIRIN</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>γ</mi><mn>04</mn></msub><msub><mtext>SCSTIMR</mtext><mi>j</mi></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>0</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0039" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>10</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>1</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0040" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>20</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0041" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>30</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0042" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>4</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>40</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>4</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0043" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>5</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>50</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>5</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0044" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>6</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>60</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>6</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0045" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>7</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>70</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>7</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml><ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0046" display="block" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>β</mi><mrow><mn>8</mn><mi>j</mi></mrow></msub><mo linebreak="goodbreak">=</mo><msub><mi>γ</mi><mn>80</mn></msub><mo linebreak="goodbreak">+</mo><msub><mi>u</mi><mrow><mn>8</mn><mi>j</mi></mrow></msub></mrow></semantics></math> </ephtml></p> <p> <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0047" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>GENDER</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0048" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>ESCS</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0049" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>CC</mi><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0050" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>ENGAGE</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0051" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>TIME</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0052" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>UNDREM</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0053" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>METASUM</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> and <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0054" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>METASPAM</mtext><mi mathvariant="italic">ij</mi></msub></mrow></semantics></math> </ephtml> indicate gender, ESCS, cultural capital, reading engagement, time spent reading, respectively, and PISA three indices of reading meta‐cognition (understanding and remembering, summarising, assess credibility) of student <emph>i</emph> from school <emph>j</emph>. <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0055" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>SCESCS</mtext><mi>j</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0056" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>SCTSUP</mtext><mi>j</mi></msub></mrow></semantics></math> </ephtml> , <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0057" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>SCDIRIN</mtext><mi>j</mi></msub></mrow></semantics></math> </ephtml> and <ephtml> <math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0058" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mtext>SCSTIMR</mtext><mi>j</mi></msub></mrow></semantics></math> </ephtml> correspondingly present average ESCS, teacher support in test language lessons, teacher‐directed instruction, teacher's stimulation of reading engagement perceived by students of each school <emph>j</emph>.</p> <hd id="AN0187391082-19">FINDINGS</hd> <p></p> <hd id="AN0187391082-20">The results of regression analysis</hd> <p>H1 was strongly confirmed through the results of regression analysis, as displayed in Table 4, in which reading functions as a powerful predictor of student performance in mathematics and science on PISA 2018, as evidenced by its overwhelming correlations with mathematics (<emph>r</emph><sups>2</sups> = 0.69) and with science (<emph>r</emph><sups>2</sups> = 0.78).</p> <p>4 TABLE Reading as the predictor of student scores in mathematics and science on PISA 2018.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Dependent variable</th><th align="left">Independent variable</th><th align="left">Unstandardised coefficients</th><th align="left">Standardised coefficients</th><th align="left"><italic>t</italic></th><th align="left">Sig.</th><th align="left">Adjusted <italic>R</italic><sup>2</sup></th></tr><tr><th align="left">B</th><th align="left">SE</th><th align="left">Beta</th></tr></thead><tbody valign="top"><tr><td align="left">Mathematics</td><td align="left">(Constant)</td><td align="char" char=".">123.096</td><td align="char" char=".">5.208</td><td align="left" /><td align="char" char=".">23.638</td><td align="char" char=".">0.000***</td><td align="char" char=".">0.686</td></tr><tr><td align="left">Reading</td><td align="char" char=".">0.812</td><td align="char" char=".">0.010</td><td align="char" char=".">0.828</td><td align="char" char=".">82.853</td><td align="char" char=".">0.000***</td><td align="left" /></tr><tr><td align="left">Science</td><td align="left">(Constant)</td><td align="char" char=".">81.672</td><td align="char" char=".">3.765</td><td align="left" /><td align="char" char=".">21.690</td><td align="char" char=".">0.000***</td><td align="char" char=".">0.783</td></tr><tr><td align="left">Reading</td><td align="char" char=".">0.864</td><td align="char" char=".">0.008</td><td align="char" char=".">0.885</td><td align="char" char=".">102.271</td><td align="char" char=".">0.000***</td><td align="left" /></tr></tbody></table> </ephtml> </p> <p>2 <emph>Note</emph>: *<emph>p</emph> < 0.05, **<emph>p</emph> < 0.01, ***<emph>p</emph> < 0.001.</p> <hd id="AN0187391082-21">HLM results</hd> <p></p> <hd id="AN0187391082-22">Null model</hd> <p>The ICC of Model 1 in Table 5 is 0.2912, meaning that 29.12% of the variation in student reading performance in PISA 2018 is explained by the effect of the school level. This is much higher than 0.138, which is defined as a strong criterion for the need to perform HLM (Cohen, [<reflink idref="bib35" id="ref153">35</reflink>]).</p> <p>5 TABLE Determinants of the PISA reading scores (1/3).</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left" /><th align="left">Reading score</th><th align="left">Model 1 (null model)</th><th align="left">Model 2</th><th align="left">Model 3</th></tr><tr><th align="left" /><th align="left">Fixed effects</th><th align="left">Coefficient</th><th align="left">Coefficient</th><th align="left">Coefficient</th></tr></thead><tbody valign="top"><tr><td align="left">Level 1</td><td align="left">Intercept <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0059" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>00</semantics></math></p></td><td align="left">506.16***(3.54)[142.81]{0.000}</td><td align="left">530.97***(3.01)[176.53]{0.000}</td><td align="left">532.14***(2.96)[179.50]{0.000}</td></tr><tr><td align="left">Gender <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0060" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>10</semantics></math></p></td><td align="left" /><td align="left">17.53***(2.34)[7.48]{0.000}</td><td align="left">15.27***(2.39) [6.39]{0.000}</td></tr><tr><td align="left">ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0061" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>20</semantics></math></p></td><td align="left" /><td align="left">17.53***(1.38)[12.69]{0.000}</td><td align="left">8.00***(1.87) [4.28]{0.000}</td></tr><tr><td align="left">Cultural capital <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0062" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>30</semantics></math></p></td><td align="left" /><td align="left" /><td align="left">14.26***(1.71) [8.32]{0.000}</td></tr><tr><td align="left">Reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0063" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>40</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Time spent reading <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0064" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>50</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: understanding and remembering <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0065" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>60</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: summarising <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0066" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>70</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: assess credibility <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0067" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>80</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Level 2</td><td align="left">School average ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0068" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>01</semantics></math></p></td><td align="left" /><td align="left">106.55***(4.58) [23.28]{0.000}</td><td align="left">106.65***(4.59) [23.21]{0.000}</td></tr><tr><td align="left">School average teacher support <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0069" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>02</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">School average direct instruction <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0070" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>03</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">School average teacher's stimulation of reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0071" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>04</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Random effects</td><td align="left">School level, var. (<p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0072" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>u</mi>0<mi>j</mi></semantics></math></p>)</td><td align="left">2926.11</td><td align="left">738.02</td><td align="left">741.98</td></tr><tr><td align="left">Student level, <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0073" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi mathvariant="normal">σ</mi>2</semantics></math></p></td><td align="left">7121.73</td><td align="left">6861.54</td><td align="left">6751.21</td></tr><tr><td align="left">Obs./groups</td><td align="left">6564/192</td><td align="left">6564/192</td><td align="left">6564/192</td></tr><tr><td align="left">Chi‐square</td><td align="left">2848.12***</td><td align="left">862.98***</td><td align="left">877.52***</td></tr><tr><td align="left">ICC</td><td align="left">0.2912</td><td align="left">0.0971</td><td align="left">0.0990</td></tr><tr><td align="left">% of explained inter‐school variation</td><td align="left" /><td align="left">75%</td><td align="left">75%</td></tr><tr><td align="left">% of explained intra‐school variation</td><td align="left" /><td align="left">4%</td><td align="left">5%</td></tr></tbody></table> </ephtml> </p> <ulist> <item>3 <emph>Note</emph>: Values in parentheses are standard error; values in square brackets are <emph>t</emph>‐ratio; values in curly brackets are the corresponding <emph>p</emph>‐values. *<emph>p</emph> < 0.05, **<emph>p</emph> < 0.01, ***<emph>p</emph> < 0.001.</item> <item>6 TABLE Determinants of the PISA reading scores (2/3).</item> </ulist> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left" /><th align="left">Reading score</th><th align="left">Model 4</th><th align="left">Model 4‐1</th><th align="left">Model 4‐2</th><th align="left">Model 4‐3</th></tr><tr><th align="left" /><th align="left">Fixed effects</th><th align="left">Coefficient</th><th align="left">Coefficient</th><th align="left">Coefficient</th><th align="left">Coefficient</th></tr></thead><tbody valign="top"><tr><td align="left">Level 1</td><td align="left">Intercept <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0074" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>00</semantics></math></p></td><td align="left">525.99***(3.16)[166.66]{0.000}</td><td align="left">529.96***(3.01)[175.87]{0.000}</td><td align="left">531.45***(3.01)[176.69]{0.000}</td><td align="left">525.80***(3.16)[166.02]{0.000}</td></tr><tr><td align="left">Gender <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0075" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>10</semantics></math></p></td><td align="left">15.07***(2.40)[6.29]{0.000}</td><td align="left">15.31***(2.40)[6.39]{0.000}</td><td align="left">15.28***(2.39)[6.39]{0.000}</td><td align="left">15.07***(2.38)[6.33]{0.000}</td></tr><tr><td align="left">ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0076" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>20</semantics></math></p></td><td align="left">7.99***(1.87)[4.27]{0.000}</td><td align="left">8.01***(1.87)[4.28]{0.000}</td><td align="left">8.00***(1.87)[4.28]{0.000}</td><td align="left">7.99***(1.87)[4.28]{0.000}</td></tr><tr><td align="left">Cultural capital <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0077" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>30</semantics></math></p></td><td align="left">14.27***(1.71)[8.33]{0.000}</td><td align="left">14.26***(1.71)[8.32]{0.000}</td><td align="left">14.26***(1.71)[8.32]{0.000}</td><td align="left">14.27***(1.71)[8.33]{0.000}</td></tr><tr><td align="left">Reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0078" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>40</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Time spent reading <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0079" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>50</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: understanding and remembering <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0080" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>60</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: summarising <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0081" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>70</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Meta‐cognition: assess credibility <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0082" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>80</semantics></math></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left">Level 2</td><td align="left">School average ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0083" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>01</semantics></math></p></td><td align="left">95.70***(4.60)[20.80]{0.000}</td><td align="left">104.56***(4.51)[23.17]{0.000}</td><td align="left">107.23***(4.54)[23.64]{0.000}</td><td align="left">100.80***(4.68)[21.53]{0.000}</td></tr><tr><td align="left">School average teacher support <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0084" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>02</semantics></math></p></td><td align="left">48.03**(16.88)[2.85]{0.005}</td><td align="left">35.94***(8.38)[4.29]{0.000}</td><td align="left" /><td align="left" /></tr><tr><td align="left">School average direct instruction <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0085" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>03</semantics></math></p></td><td align="left">−48.70***(13.84)[−3.52]{0.000}</td><td align="left" /><td align="left">11.49(8.89)[1.29]{0.198}</td><td align="left" /></tr><tr><td align="left">School average teacher's stimulation of reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0086" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>04</semantics></math></p></td><td align="left">33.19**(10.81)[3.07]{0.002}</td><td align="left" /><td align="left" /><td align="left">34.04***(6.90)[4.93]{0.000}</td></tr><tr><td align="left">Random effects</td><td align="left">School level, var. (<p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0087" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>u</mi>0<mi>j</mi></semantics></math></p>)</td><td align="left">619.70</td><td align="left">688.13</td><td align="left">740.75</td><td align="left">659.95</td></tr><tr><td align="left">Student level, <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0088" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi mathvariant="normal">σ</mi>2</semantics></math></p></td><td align="left">6751.57</td><td align="left">6751.48</td><td align="left">6751.12</td><td align="left">6751.13</td></tr><tr><td align="left">Obs./groups</td><td align="left">6564/192</td><td align="left">6564/192</td><td align="left">6564/192</td><td align="left">6564/192</td></tr><tr><td align="left">Chi‐square</td><td align="left">751.93***</td><td align="left">823.67***</td><td align="left">872.65***</td><td align="left">798.57***</td></tr><tr><td align="left">ICC</td><td align="left">0.0841</td><td align="left">0.0925</td><td align="left">0.0989</td><td align="left">0.0890</td></tr><tr><td align="left">% of explained inter‐school variation</td><td align="left">79%</td><td align="left">75%</td><td align="left">76%</td><td align="left">75%</td></tr><tr><td align="left">% of explained intra‐school variation</td><td align="left">5%</td><td align="left">5%</td><td align="left">5%</td><td align="left">5%</td></tr></tbody></table> </ephtml> </p> <p>4 <emph>Note</emph>: Values in parentheses are standard error; values in square brackets are <emph>t</emph>‐ratio; values in curly brackets are the corresponding <emph>p</emph>‐values. *<emph>p</emph> < 0.05, **<emph>p</emph> < 0.01, ***<emph>p</emph> < 0.001.</p> <hd id="AN0187391082-23">Determinants of students' reading scores</hd> <p>Tables 5–7 list the results of HLM analysis, showing that H2 was partially authenticated. Basically, the independent variables of this analysis can be classified into two categories. The first category of variables is student background conditions (e.g., gender, ESCS, school average ESCS and cultural capital). The second category of variables has two sub‐groups: school support and student personal efforts/learning strategies. School support is comprised of teacher support in test language lessons, teacher's stimulation of reading engagement perceived by students and teacher‐directed instruction. Student personal efforts/learning strategies are made up of reading engagement, time spent reading, meta‐cognition: understanding and remembering, meta‐cognition: summarising and meta‐cognition: assess credibility.</p> <p>7 TABLE Determinants of the PISA reading scores (3/3).</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left" /><th align="left">Reading score</th><th align="left">Model 5</th><th align="left">Model 5‐1</th><th align="left">Model 5‐2</th></tr><tr><th align="left" /><th align="left">Fixed effects</th><th align="left">Coefficient</th><th align="left">Coefficient</th><th align="left">Coefficient</th></tr></thead><tbody valign="top"><tr><td align="left">Level 1</td><td align="left">Intercept <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0089" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>00</semantics></math></p></td><td align="left">524.57***(3.20)[164.18]{0.000}</td><td align="left">524.06***(3.25)[161.06]{0.000}</td><td align="left">493.07***(3.76)[131.06]{0.000}</td></tr><tr><td align="left">Gender <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0090" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>10</semantics></math></p></td><td align="left">1.38(1.90)[0.73]{0.468}</td><td align="left">1.67(1.88)[0.89]{0.376}</td><td align="left">0.32(1.83)[0.18]{0.860}</td></tr><tr><td align="left">ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0091" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>20</semantics></math></p></td><td align="left">7.84***(1.45)[5.41]{0.000}</td><td align="left">10.24***(1.14)[9.01]{0.000}</td><td align="left" /></tr><tr><td align="left">Cultural capital <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0092" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>30</semantics></math></p></td><td align="left">3.84**(1.41)[2.72]{0.007}</td><td align="left" /><td align="left">7.97***(1.12)[7.11]{0.000}</td></tr><tr><td align="left">Reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0093" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>40</semantics></math></p></td><td align="left">16.57***(1.24)[13.34]{0.000}</td><td align="left">17.22***(1.20)[14.40]{0.000}</td><td align="left">16.49***(1.25)[13.22]{0.000}</td></tr><tr><td align="left">Time spent reading <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0094" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>50</semantics></math></p></td><td align="left">4.42***(0.64)[6.90]{0.000}</td><td align="left">4.62***(0.66)[7.03]{0.000}</td><td align="left">4.40***(0.65)[6.81]{0.000}</td></tr><tr><td align="left">Meta‐cognition: understanding and remembering <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0095" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>60</semantics></math></p></td><td align="left">9.45***(1.19)[7.95]{0.000}</td><td align="left">9.54***(1.19)[8.01]{0.000}</td><td align="left">9.45***(1.19)[7.93]{0.000}</td></tr><tr><td align="left">Meta‐cognition: summarising <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0096" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>70</semantics></math></p></td><td align="left">16.77***(0.94)[17.79]{0.000}</td><td align="left">16.79***(0.94)[17.80]{0.000}</td><td align="left">16.89***(0.95)[17.74]{0.000}</td></tr><tr><td align="left">Meta‐cognition: assess credibility <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0097" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>80</semantics></math></p></td><td align="left">23.50***(1.01)[23.22]{0.000}</td><td align="left">23.50***(1.02)[23.13]{0.000}</td><td align="left">23.51***(1.01)[23.11]{0.000}</td></tr><tr><td align="left">Level 2</td><td align="left">School average ESCS <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0098" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>01</semantics></math></p></td><td align="left">94.56***(4.60)[20.55]{0.000}</td><td align="left">94.50***(4.60)[20.54]{0.000}</td><td align="left" /></tr><tr><td align="left">School average teacher support <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0099" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>02</semantics></math></p></td><td align="left">46.59**(17.04)[2.73]{0.007}</td><td align="left">46.63**(17.03)[2.74]{0.007}</td><td align="left">97.78**(30.46)[3.21]{0.002}</td></tr><tr><td align="left">School average direct instruction <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0100" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>03</semantics></math></p></td><td align="left">−49.13***(13.61)[−3.61]{0.000}</td><td align="left">−49.12***(13.60)[−3.61]{0.000}</td><td align="left">−142.43***(23.93)[−5.95]{0.000}</td></tr><tr><td align="left">School average teacher's stimulation of reading engagement <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0101" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>γ</mi>04</semantics></math></p></td><td align="left">33.76**(10.91)[3.10]{0.002}</td><td align="left">33.64**(10.89)[3.09]{0.002}</td><td align="left">86.60***(19.10)[4.53]{0.000}</td></tr><tr><td align="left">Random effects</td><td align="left">School level, var. (<p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0102" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi>u</mi>0<mi>j</mi></semantics></math></p>)</td><td align="left">662.58</td><td align="left">661.32</td><td align="left">2085.02</td></tr><tr><td align="left">Student level, <p><math altimg="urn:x-wiley:01411926:media:berj4157:berj4157-math-0103" display="inline" overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><semantics xmlns=""><mi mathvariant="normal">σ</mi>2</semantics></math></p></td><td align="left">4987.12</td><td align="left">4994.15</td><td align="left">5011.58</td></tr><tr><td align="left">Obs./groups</td><td align="left">6564/192</td><td align="left">6564/192</td><td align="left">6564/192</td></tr><tr><td align="left">Chi‐square</td><td align="left">1002.15***</td><td align="left">999.54***</td><td align="left">2817.09***</td></tr><tr><td align="left">ICC</td><td align="left">0.1173</td><td align="left">0.1169</td><td align="left">0.2938</td></tr><tr><td align="left">% of explained inter‐school variation</td><td align="left">77%</td><td align="left">77%</td><td align="left">29%</td></tr><tr><td align="left">% of explained intra‐school variation</td><td align="left">30%</td><td align="left">30%</td><td align="left">30%</td></tr></tbody></table> </ephtml> </p> <p>5 <emph>Note</emph>: Values in parentheses are standard error; values in square brackets are <emph>t</emph>‐ratio; values in curly brackets are the corresponding <emph>p</emph>‐values. *<emph>p</emph> < 0.05, **<emph>p</emph> < 0.01, ***<emph>p</emph> < 0.001.</p> <p>While gender is statistically significant in Models 2 (<emph>γ</emph><subs>10</subs> = 17.53, <emph>p</emph> < 0.001), 3 (<emph>γ</emph><subs>10</subs> = 15.27, <emph>p</emph> < 0.001), 4 (<emph>γ</emph><subs>10</subs> = 15.07, <emph>p</emph> < 0.001), 4‐1 (<emph>γ</emph><subs>10</subs> = 15.31, <emph>p</emph> < 0.001), 4‐2 (<emph>γ</emph><subs>10</subs> = 15.28, <emph>p</emph> < 0.001) and 4‐3 (<emph>γ</emph><subs>10</subs> = 15.07, <emph>p</emph> < 0.001), it becomes insignificant in Model 5 (<emph>γ</emph><subs>10</subs> = 1.38, <emph>p</emph> = 0.47). Unlike this trend, the influence of ESCS on student reading performance remains prominent across Models 2 (<emph>γ</emph><subs>20</subs> = 17.53, <emph>p</emph> < 0.001), 3 (<emph>γ</emph><subs>20</subs> = 8.00, <emph>p</emph> < 0.001), 4 (<emph>γ</emph><subs>20</subs> = 7.99, <emph>p</emph> < 0.001), 4‐1 (<emph>γ</emph><subs>20</subs> = 8.01, <emph>p</emph> < 0.001), 4‐2 (<emph>γ</emph><subs>20</subs> = 8.00, <emph>p</emph> < 0.001), 4‐3 (<emph>γ</emph><subs>20</subs> = 7.99, <emph>p</emph> < 0.001) and 5 (<emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001). This influence further extends to schools, as is evident in the coefficients of school average ESCS throughout Models 2 (<emph>γ</emph><subs>01</subs> = 106.55, <emph>p</emph> < 0.001), 3 (<emph>γ</emph><subs>01</subs> = 106.65, <emph>p</emph> < 0.001), 4 (<emph>γ</emph><subs>01</subs> = 95.70, <emph>p</emph> < 0.001), 4‐1 (<emph>γ</emph><subs>01</subs> = 104.56, <emph>p</emph> < 0.001), 4‐2 (<emph>γ</emph><subs>01</subs> = 107.23, <emph>p</emph> < 0.001), 4‐3 (<emph>γ</emph><subs>01</subs> = 100.80, <emph>p</emph> < 0.001) and 5 (<emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001). Even if the influence of cultural capital on student reading performance is visible, as showcased in its coefficients of Models 3 (<emph>γ</emph><subs>30</subs> = 14.26, <emph>p</emph> < 0.001), 4 (<emph>γ</emph><subs>30</subs> = 14.27, <emph>p</emph> < 0.001), 4‐1 (<emph>γ</emph><subs>30</subs> = 14.26, <emph>p</emph> < 0.001), 4‐2 (<emph>γ</emph><subs>30</subs> = 14.26, <emph>p</emph> < 0.001) and 4‐3 (<emph>γ</emph><subs>30</subs> = 14.27, <emph>p</emph> < 0.001), its effect declines sharply in Model 5 (<emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.01).</p> <p>The above findings denote that notwithstanding the fact that these variables constantly change their influences on student reading achievement on PISA 2018, their effects generally project a descending trend, as is observable from the first models in which they are calculated (gender <emph>γ</emph><subs>10</subs> = 17.53, <emph>p</emph> < 0.001, Model 2; ESCE <emph>γ</emph><subs>20</subs> = 17.53, <emph>p</emph> < 0.001, Model 2; school average ESCS <emph>γ</emph><subs>01</subs> = 106.55, <emph>p</emph> < 0.001, Model 2; cultural capital <emph>γ</emph><subs>30</subs> = 14.26, <emph>p</emph> < 0.001, Model 3) to Model 5 (gender <emph>γ</emph><subs>10</subs> = 1.38, <emph>p</emph> = 0.47; ESCE <emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001; school average ESCS <emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001; cultural capital <emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.01). This situation points to the exciting connotation that even though gender, ESCS, school average ESCS and cultural capital are, in a sense, unchangeable independent variables, their influences on reading performance can be considerably reduced when other selected variables are considered. More precisely, this reduction is attributable to the introduction of student variables associated with student personal efforts/learning strategies, as observed from Model 5, in which reading engagement (<emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001), time spent reading (<emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001), meta‐cognition: understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001), meta‐cognition: summarising (<emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001) and meta‐cognition: assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001) play a vital role in this aspect. The importation of school variables also ushers in a similar picture, in which school average teacher support visibly contributes to this result, as manifested in its coefficients across Models 4 (<emph>γ</emph><subs>02</subs> = 48.03, <emph>p</emph> < 0.01), 4‐1 (<emph>γ</emph><subs>02</subs> = 35.94, <emph>p</emph> < 0.01) and 5 (<emph>γ</emph><subs>02</subs> = 46.59, <emph>p</emph> < 0.05). Similar contributions made by school average teacher's stimulation of reading engagement are certified by Models 4 (<emph>γ</emph><subs>04</subs> = 33.19, <emph>p</emph> < 0.01), 4‐3 (<emph>γ</emph><subs>04</subs> = 34.04, <emph>p</emph> < 0.001) and 5 (<emph>γ</emph><subs>04</subs> = 33.76, <emph>p</emph> < 0.01). Considering ESCS, which includes cultural capital, as incarnated in one of its constituents, cultural possessions, when their contributions to students' reading scores are calculated separately, the descending pattern is similar to the above picture. More concretely, in Model 5‐1, the accumulated effect of student personal efforts/learning strategies such as reading engagement (<emph>γ</emph><subs>40</subs> = 17.22, <emph>p</emph> < 0.001), time spent reading (<emph>γ</emph><subs>50</subs> = 4.62, <emph>p</emph> < 0.001), understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.54, <emph>p</emph> < 0.001), summarising (<emph>γ</emph><subs>70</subs> = 16.79, <emph>p</emph> < 0.001) and assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001), and school variables such as school average teacher support (<emph>γ</emph><subs>02</subs> = 46.63, <emph>p</emph> < 0.01), school average teacher's simulation of reading (<emph>γ</emph><subs>04</subs> = 33.64, <emph>p</emph> < 0.01) is 151.94. This effect is much higher than 106.41, the accumulated effect of the unchangeable independent variables in Model 5‐1, including gender (<emph>γ</emph><subs>10</subs> = 1.67, <emph>p</emph> = 0.38), ESCE (<emph>γ</emph><subs>20</subs> = 10.24, <emph>p</emph> < 0.001) and school average ESCS (<emph>γ</emph><subs>01</subs> = 94.50, <emph>p</emph> < 0.001). In this respect, the accumulated effect of gender (<emph>γ</emph><subs>10</subs> = 0.32, <emph>p</emph> = 0.86) and cultural capital (<emph>γ</emph><subs>30</subs> = 7.97, <emph>p</emph> < 0.001) becomes minor in Model 5‐2.</p> <p>These findings strongly support our conjecture that in terms of inequity in educational results, or at least in reading, schools can perform as great equalisers rather than as repressive devices, maintaining the phenomenon of cultural reproduction, if school reading programmes/interventions emphasising school support and student personal efforts/learning strategies are practiced. Nevertheless, such interventions need to avoid the use of school average direct instruction (ST012) due to its strong negative correlations with student performance in reading, as evident in Models 4 (<emph>γ</emph><subs>03</subs> = −48.70, <emph>p</emph> < 0.01) and 5 (<emph>γ</emph><subs>03</subs> = −49.13, <emph>p</emph> < 0.001). This is mainly inasmuch as the questions of ST012 relate to teacher‐based instruction devaluing the status of student‐based learning, which is the core element in implementing reading activities/strategies and teacher support/simulation, as manifested in the above variables. According to the positive effects in Model 5 (full model), the importance of the variables can be ranked from school average ESCS (<emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001) to teacher support (<emph>γ</emph><subs>02</subs> = 46.59, <emph>p</emph> < 0.01), teacher's stimulation (<emph>γ</emph><subs>04</subs> = 33.76, <emph>p</emph> < 0.01), meta‐cognition: assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001), meta‐cognition: summarising (<emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001), reading engagement (<emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001), meta‐cognition: understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001), ESCS (<emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001), time spent reading (<emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001) and finally cultural capital (<emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.01).</p> <hd id="AN0187391082-24">DISCUSSION</hd> <p>The strong findings of our regression analysis solidly substantiate RQ1 and H1, illuminating that reading acts as a reliable predictor of student performance in mathematics (<emph>r</emph><sups>2</sups> = 0.69) and science (<emph>r</emph><sups>2</sups> = 0.78) on the PISA 2018. This finding resonates with related findings on mathematics (Lambdin, [<reflink idref="bib60" id="ref154">60</reflink>]; Ombra, [<reflink idref="bib89" id="ref155">89</reflink>]; Vilenius‐Tuohimaa et al., [<reflink idref="bib110" id="ref156">110</reflink>]) and on science (Jang & Hand, [<reflink idref="bib55" id="ref157">55</reflink>]; Pearson et al., [<reflink idref="bib90" id="ref158">90</reflink>]; Webb, [<reflink idref="bib115" id="ref159">115</reflink>]). It thus contributes to the universal principle that learning outcomes are implicitly appropriated by the interplay between linguistic ability (Chiang, Thurston & Cockerill, [<reflink idref="bib30" id="ref160">30</reflink>]; Chiang et al., [<reflink idref="bib29" id="ref161">29</reflink>]), reading comprehension ability (Bowyer‐Crane et al., [<reflink idref="bib25" id="ref162">25</reflink>]; Hulme & Snowling, [<reflink idref="bib53" id="ref163">53</reflink>]) and logical reasoning ability (Can, [<reflink idref="bib27" id="ref164">27</reflink>]; Lambdin, [<reflink idref="bib60" id="ref165">60</reflink>]; Schoenfeld, [<reflink idref="bib99" id="ref166">99</reflink>]; Segers & Verhoeven, [<reflink idref="bib100" id="ref167">100</reflink>]). At this point, this discovery accounts for why reading benefits students' cognitive development. At the centre of this are elements of internal reasoning, logic and use of language proposed by Vygotsky ([<reflink idref="bib113" id="ref168">113</reflink>]). Most mathematics problems that require logic and reasoning include word descriptions as part of the question (Verschaffel et al., [<reflink idref="bib109" id="ref169">109</reflink>]). It is reported that to solve such problems requires word knowledge that includes reading comprehension (Fuchs et al., [<reflink idref="bib43" id="ref170">43</reflink>]). This is in addition to mathematics skills (Fuchs et al., [<reflink idref="bib44" id="ref171">44</reflink>]).</p> <p>The above evidence provides us with a hard bedrock upon which to investigate RQ2 and H2 further, concentrating on the contributions of identified independent variables to students' reading scores on PISA 2018. Broadly speaking, such contributions are validated by the estimates of HLM analysis. While the influence of background variables on student reading performance on PISA 2018 is overwhelming throughout Models 2, 3, 4, 4‐1, 4‐2 and 4‐3, it shrinks sharply in Model 5, in which the variables of school support and student personal efforts/learning strategies are included. The coefficient for gender, for instance, changes from significant in Model 2 (<emph>γ</emph><subs>10</subs> = 17.53, <emph>p</emph> < 0.001) to insignificant in Model 5 (<emph>γ</emph><subs>10</subs> = 1.38, <emph>p</emph> = 0.47). A more striking breakthrough is that this reduction crosses over cultural capital. Though its influence in the first five models is robust and stable, it deflates 73.07% from Model 3 (<emph>γ</emph><subs>30</subs> = 14.26, <emph>p</emph> < 0.001) to Model 5 (<emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.05), albeit remaining statistically significant, and this deflation becomes 44.11% in Model 5‐2 (<emph>γ</emph><subs>30</subs> = 7.97, <emph>p</emph> < 0.001). This descending pattern also occurs on ESCS, the coefficient of which contracts 223.60% from Model 2 (<emph>γ</emph><subs>20</subs> = 17.53, <emph>p</emph> < 0.001) to Model 5 (<emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001). This contraction is 41.41% in Model 5‐1 (<emph>γ</emph><subs>20</subs> = 10.27, <emph>p</emph> < 0.001). Surprisingly, in this respect, school average ESCS is notable across all models we have tested, and its influence only diminishes 11.25% from Model 2 (<emph>γ</emph><subs>01</subs> = 106.55, <emph>p</emph> < 0.001) to Model 5 (<emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001), and this situation can apply to Model 5‐1 because its coefficient (<emph>γ</emph><subs>01</subs> = 94.50, <emph>p</emph> < 0.001) is almost the same as that in Model 5. Concerning the effects of ESCS, the school level markedly surpasses the respondent level across all models, as exemplified by Model 2 (<emph>γ</emph><subs>01</subs> = 106.55, <emph>p</emph> < 0.001 vs. <emph>γ</emph><subs>20</subs> = 17.53, <emph>p</emph> < 0.001) and Model 5 (<emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001 vs. <emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001). Comparing the effects of ESCS and cultural capital, the former (<emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001) is stronger than the latter (<emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.001). When their effects are calculated separately, this principle is confirmed again, as manifested in the coefficients in Model 5‐1 (<emph>γ</emph><subs>20</subs> = 10.24, <emph>p</emph> < 0.001) and Model 5‐2 (<emph>γ</emph><subs>20</subs> = 7.97, <emph>p</emph> < 0.001). This scenario can be attributed to the complicated constitution of SES (Caro et al., [<reflink idref="bib28" id="ref172">28</reflink>]), which covers cultural capital, so its influence is superior to that of cultural capital. Therefore, the cause of this is likely derived from the constitution of ESCS, which comprises parental qualifications, occupations and educational resources—the critical element of cultural capital. The first two components heavily regulate household incomes (Hoff, [<reflink idref="bib51" id="ref173">51</reflink>]; National Institute for Education Statistics, [<reflink idref="bib76" id="ref174">76</reflink>]), which is the yardstick for scaling SES and children's academic achievements (Matherly et al., [<reflink idref="bib72" id="ref175">72</reflink>]; Saraví, [<reflink idref="bib98" id="ref176">98</reflink>]). When this formation is configured through the combination of SES and cultural capital, the above finding in turn confirms Bourdieu's cultural capital theory, delineating the one‐way convertible relationship from economic capital to cultural capital (Bourdieu, [<reflink idref="bib19" id="ref177">19</reflink>], [<reflink idref="bib20" id="ref178">20</reflink>]; Bourdieu & Passeron, [<reflink idref="bib23" id="ref179">23</reflink>]; Chiang, Toh et al., [<reflink idref="bib32" id="ref180">32</reflink>]; Katartzi & Hayward, [<reflink idref="bib58" id="ref181">58</reflink>]; Lareau, [<reflink idref="bib62" id="ref182">62</reflink>]).</p> <p>Apparently, with respect to the coefficient values, there is an ascending trajectory from Model 1 to Model 5, commonly existing between the background variables, such as gender, cultural capital, ESCS and school average ESCS. These layouts denote promising implications for relieving inequity in educational results, conceptualised as cultural reproduction in this study. More specifically, when selected variables at level 2 are considered, students' scores on reading tests on the PISA 2018 visibly increase, as witnessed by the encouraging coefficients in Model 5. In this aspect, the variables, related to student personal efforts/learning strategies, such as reading engagement (<emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001), time spent reading (<emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001), meta‐cognition: understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001), meta‐cognition: summarising (<emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001) and meta‐cognition: assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001) are pivotal factors. A similar picture can be seen from the variables related to school support, such as school average teacher support (<emph>γ</emph><subs>02</subs> = 46.59, <emph>p</emph> < 0.01) and school average teacher's stimulation of reading engagement (<emph>γ</emph><subs>04</subs> = 33.76, <emph>p</emph> < 0.01). In light of their contributions, the above independent variables can be graded in order from school average teacher support (<emph>γ</emph><subs>02</subs> = 46.59, <emph>p</emph> < 0.05) to school average teacher's stimulation of reading engagement (<emph>γ</emph><subs>04</subs> = 33.76, <emph>p</emph> < 0.01), meta‐cognition: assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001), meta‐cognition: summarising (<emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001), reading engagement (<emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001), meta‐cognition: understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001) and finally time spent reading (<emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001). This stratum portrays that the accumulated effects of school average teacher support and school average teacher's stimulation of reading engagement can elevate scores by up to 80.35 when they operate on a unit simultaneously, which is higher than those of student personal efforts/learning strategy (70.71) in the same condition. Therefore, to improve students' reading performance, we need to establish a scaffolding environment whereby teachers are encouraged to support and stimulate students' reading engagement. This goal needs to be embraced by teachers through a strong commitment to promote student reading activities, which is likely to be fortified when they witness the impact of improved reading on students' learning outcomes. Furthermore, the features of students' personal lives can be shaped through supportive teaching actions, creating a scaffolding environment that can contribute significantly to their output. These correlations highlight the fact that teachers play a key role in ensuring the contribution of reading to academic results, so education policy needs to provide teachers with incentives and establish a supportive environment for them. Education policy must also provide teachers with sufficient resources for reading instruction and systematically implement reading programmes, based on the school and student variables mentioned above. Even though these results highlight the critical leverage of school support in improving student reading outcomes, if we want to further improve such outcomes, it is necessary to eschew the use of school average direct instruction due to its strong negative coefficient (<emph>γ</emph><subs>03</subs> = −49.13, <emph>p</emph> < 0.01). This negative impact may come from its coercive and imposing features, incompatible with or even contrary to the active nature of efficient reading, which needs to be boosted by commitment, initiative, integration, analytical reasoning, and so on (Bilton & Duff, [<reflink idref="bib8" id="ref183">8</reflink>]; Shanahan, [<reflink idref="bib101" id="ref184">101</reflink>]; Thurston et al., [<reflink idref="bib106" id="ref185">106</reflink>]).</p> <p>Meanwhile, the outputs of student personal efforts/learning strategies cannot be underestimated, given their accumulated effects (70.71) noted above. Surprisingly, in comparison with the contribution of reading engagement (<emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001), time spent reading (<emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001) plays a relatively minor role in student reading performance. In this sense, diligence demands support of engagement; otherwise, the outcome of personal efforts will be fragile. This principle is in tune with related studies which reveal that this integration yields better learning outcomes than their separated operations (Locher & Pfost, [<reflink idref="bib68" id="ref186">68</reflink>]; Mol & Bus, [<reflink idref="bib74" id="ref187">74</reflink>]; Taylor et al., [<reflink idref="bib104" id="ref188">104</reflink>]). Moreover, the cumulative consequences of student personal efforts are much smaller than those of student personal learning strategies, termed meta‐cognition. Since assess credibility (<emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001), summarising (<emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001) and understanding and remembering (<emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001) are meta‐cognitive engagements, belonging to high‐order abstract reasoning activities (Korotaeva, [<reflink idref="bib59" id="ref189">59</reflink>]; Pehlivanoglu et al., [<reflink idref="bib91" id="ref190">91</reflink>]; Yu, [<reflink idref="bib119" id="ref191">119</reflink>]), it can be inferred that the three forms of meta‐cognition heavily mediate the outputs of reading, thereby functioning as a gateway for enhancing student reading performance. Fortunately, it has been found that these three forms of meta‐cognition are learnable through schooling (Dunlosky et al., [<reflink idref="bib42" id="ref192">42</reflink>]; Logan et al., [<reflink idref="bib69" id="ref193">69</reflink>]), so schools may accomplish their acquisition by students. In addition, because the realisation of these meta‐activities is essentially based on mental abilities empowered with voluntarism or auto‐involvement, this student‐based orientation bolstered by self‐regulation or auto‐initiative can be verified through the strong negative effect of school average direct instruction (<emph>γ</emph><subs>03</subs> = −49.13, <emph>p</emph> < 0.01), the delivery of which mainly calls upon the implicit coercive force engraved within direct instruction, functioning similarly to orders or commands.</p> <p>The total effects of school support (teacher support <emph>γ</emph><subs>02</subs> = 46.59, <emph>p</emph> < 0.01; teacher's stimulation of reading engagement <emph>γ</emph><subs>04</subs> = 33.76, <emph>p</emph> < 0.01) and student personal efforts/learning strategies (reading engagement <emph>γ</emph><subs>40</subs> = 16.57, <emph>p</emph> < 0.001; time spent reading <emph>γ</emph><subs>50</subs> = 4.42, <emph>p</emph> < 0.001; assess credibility <emph>γ</emph><subs>80</subs> = 23.50, <emph>p</emph> < 0.001; summarising <emph>γ</emph><subs>70</subs> = 16.77, <emph>p</emph> < 0.001; understanding and remembering <emph>γ</emph><subs>60</subs> = 9.45, <emph>p</emph> < 0.001) is 151.06, which is much bigger than those of background variables (106.24), consisting of ESCS (<emph>γ</emph><subs>20</subs> = 7.84, <emph>p</emph> < 0.001), school average ESCS (<emph>γ</emph><subs>01</subs> = 94.56, <emph>p</emph> < 0.001) and cultural capital (<emph>γ</emph><subs>30</subs> = 3.84, <emph>p</emph> < 0.01). This scenario implies that when schools engage in appropriate tactics, such as school support and student personal efforts/learning strategies, student reading performance is no longer dominated by their family background variables, the effects of which may still be statistically visible. This encouraging shift articulates a hope that when schools carry out positive educational actions like reading programmes/interventions, the structural impacts of ESCS and cultural capital will be effectively mitigated. This finding somehow fills the gap in past studies, which have consistently addressed the positive consequences of reading interventions (Shanahan, [<reflink idref="bib101" id="ref194">101</reflink>]; Silverman, [<reflink idref="bib102" id="ref195">102</reflink>]; Thurston, Cockerill et al., [<reflink idref="bib105" id="ref196">105</reflink>]; Tymms et al., [<reflink idref="bib108" id="ref197">108</reflink>]; Wasik et al., [<reflink idref="bib114" id="ref198">114</reflink>]) without including sociological factors, such as ESCS and cultural capital, as depicted above. Since reading can substantially compensate for the shortage of cultural capital in low‐SES family social spaces, which has been consistently spotlighted in past studies (Dickinson & Anastasopoulos, [<reflink idref="bib40" id="ref199">40</reflink>]; Guevara et al., [<reflink idref="bib49" id="ref200">49</reflink>]), this uncovering explicates how cultural capital can be transmitted through reading activities (Lareau, [<reflink idref="bib62" id="ref201">62</reflink>]) in high‐SES families (Bodovski et al., [<reflink idref="bib10" id="ref202">10</reflink>]; Loh & Sun, [<reflink idref="bib70" id="ref203">70</reflink>]) and why children in such families often exhibit excellent reading competence (Araujo & Costa, [<reflink idref="bib2" id="ref204">2</reflink>]; Bodovski et al., [<reflink idref="bib10" id="ref205">10</reflink>]). As the returns of reading need to be evoked through the positive educational actions specified above, these vantages are, in a sense, embedded within the schools. Accordingly, reading can be viewed as an embedded sub‐type of cultural capital inscribed within the school social space, although it still belongs to the objectified form of cultural capital (Bourdieu, [<reflink idref="bib20" id="ref206">20</reflink>]). Furthermore, since this embedded cultural capital is mainly activated and relayed within the school social space, the scaffolding of reading activities/interventions initiated by schools enables low‐SES students to become embedded cultural capital receivers who can overcome the structural limitations of their family social space brought about by the scarcity of cultural capital. This cogent evidence thus validates the concept of dual forms of social space (Chiang & Trezise, [<reflink idref="bib33" id="ref207">33</reflink>]), referring to families and schools, respectively. In this sense, if the scaffolding of reading is appropriately realised in school social spaces, the individualised agency occasionally exercised by excellent low‐SES students by acquiring educational resources from outside their family social spaces (Chiang et al., [<reflink idref="bib31" id="ref208">31</reflink>]) will change into institutionalised agency achieved by schools to ameliorate the phenomenon of cultural reproduction.</p> <p>Previous studies using PISA data have tended to look at patterns in the data or create 'league tables' of country performance. For instance, Cromley's analysis of PISA 2000 looked at correlations between reading performance and science performance (Cromley, [<reflink idref="bib37" id="ref209">37</reflink>]). Akbaşlı et al. ([<reflink idref="bib1" id="ref210">1</reflink>]) picked out five top, five mid and five bottom‐performing countries and linked data to economic performance. There is significant literature on how reading is measured using PISA, but this tends not to include sections on how schools could usefully use PISA data to inform policy and practice (e.g., OECD, [<reflink idref="bib82" id="ref211">82</reflink>]). Some governments have identified the opportunity for using PISA data to inform policy and improve reading attainment (Department for Education, [<reflink idref="bib39" id="ref212">39</reflink>]) but even here, the interrogation of data within the PISA dataset on what effective schools do to promote literacy is limited. Previously reported research on Taiwan's PISA performance has tended to focus on historical aspects of Taiwan's low reading attainment performance, in comparison to higher performance in science and mathematics, and the cultural reasons that may have a causal effect on this pattern (Lin et al., [<reflink idref="bib65" id="ref213">65</reflink>]). Alternatively, there have been attempts to analyse data in regional groups in order to build league tables for groups of countries with shared ethnic histories and traditions (Gu & Lau, [<reflink idref="bib48" id="ref214">48</reflink>]). Therefore, this study offers insight into a new relationship with PISA data, whereby countries/regions may not use it to categorise and grade school performance but use the data to help improve school performance.</p> <hd id="AN0187391082-25">CONCLUSION</hd> <p>Regarding student scores on PISA 2018, reading ability reliably forecasts mathematics and science levels. This causal relation results from the interwoven relationships between the selected variables of reading, specified as follows. Based on the effects, the three forms of meta‐cognition can be ranked in order, from assess credibility to summarising and, finally, understanding and remembering. This stratified edifice serves as a guide for perfecting student reading achievements. Nevertheless, in this regard, teacher support and stimulation are more critical than students' personal efforts and learning strategies, including the above three forms of meta‐cognition, reading engagement and time spent. Above all, school average ESCS is the most decisive determinant in Model 5, referring to the boundary between schools, which is mainly fabricated by ESCS. Although this boundary is established primarily by unchangeable features of demographic variables, including ESCS, school average ESCS and cultural capital, its impacts can be attenuated when selected variables—such as school support and student personal efforts/learning strategies—are taken into action, as witnessed by their total effect (151.06), which is much higher than those of background variables (106.24). This anecdote illustrates that if positive educational actions are adopted in schools, the constraints of unchangeable variables can be weakened. More specifically, the strategies correlated to school support and student personal efforts/learning strategies function as a foundation for these actions because their effects efficiently alleviate the structural impact of weak cultural capital on low‐SES students, which occurs as a result of the limited opportunities to acquire cultural capital within their family social space. The above educational actions become workable since school support and student personal learning strategies are potentially acquirable and learnable in schools.</p> <p>Inasmuch as the rewards of strong reading ability can be attained through strategic actions initiated by schools, reading, by its nature, can be regarded as a sub‐genre of objectified cultural capital embedded within educational institutes, particularly schools. Accordingly, reading can be characterised as compensatory cultural capital encapsulated within school social spaces, rather than family social spaces. When schools exert institutionalised agency through reading programmes/interventions, integrating school support and student personal efforts/learning strategies, such reading interventions help low‐SES students utilise embedded cultural capital available within school social spaces, so they are no longer restricted by the structural inhibition of their family social space. These inspiring facts not only explain how reading contributes to cognitive development but also how reading assists schools in transforming themselves into social equalisers rather than repressive apparatuses. At this point, we believe that our convincing findings enable us to claim some contributions to the elaboration of theories of cultural capital, as follows.</p> <p>Firstly, our robust evidence distinguishes two genres of social space not addressed by Bourdieusian academics, namely a primary and a secondary arena, referring to families and educational institutes, respectively. Within the secondary type, schools are the most important because they are the leading sites for distributing educational resources, which can compensate for deficits in cultural capital in low‐SES family social spaces. Since low‐SES students are situated in the two genres simultaneously, this synchronicity offers them potential opportunities to access compensatory cultural capital supplied by schools. Secondly, this possibility can come to fruition when educational resources inscribed within school social spaces become compensatory cultural capital available for low‐SES students. Disappointingly, such resources may remain embedded within institutions because no systematic schemes have been implemented to unpack their advantages for this student group. Fortunately, our compelling findings indicate that reading programmes can transform this embedded cultural capital into compensatory cultural capital.</p> <p>Thirdly, this transformation becomes attainable through the practice of institutionalised agency rather than that of individualised agency. Even if excellent low‐SES students behave as cultural capital constructors by using creative actions to uncover and transform the embedded cultural capital into compensatory cultural capital for themselves, individualised agency restricts the returns of compensatory cultural capital to a very narrow personal domain, as is prominent in these students as the minority. Conversely, institutionalised agency empowers schools to comprehensively fulfil the above transformation, through which many low‐SES students can benefit in terms of learning outcomes. Since this empowerment is heavily regulated by the integration of institutionalised agency and school social spaces that govern innovative actions and educational resources, respectively, this integration, in a sense, is mainly generated by education policy, as manifested in the reading programme noted previously. Finally, because institutionalised agency assists schools in evoking compensatory cultural capital that supplements the paucity of cultural capital in low‐SES family social spaces, this situation can improve low‐SES students' academic performance. This improvement substantially reduces the influence of the structural constraints of unchangeable variables, particularly SES and cultural capital, on educational results. This scenario implies that when schools properly exert agency by generating compensatory cultural capital, inequity in educational results can be effectively harnessed, and the influence of cultural capital won't be perpetuated. This principle hereby evolves cultural capital theories from a purely structural approach to a mix of structuralism and agency. Considering the overwhelming gains brought about by compensatory cultural capital on educational results, academics should shift their research focus from structuralism towards agency, leading to innovative ideas and actions that amplify the utility of compensatory cultural capital in school social spaces for low‐SES students.</p> <p>Although the findings provide us with robust evidence of the advantages of reading being unpacked by proactive reading programmes in schools, the statistical results found in this study are based solely on Taiwanese students' performance in PISA 2018. Since the results are derived from a single region, more studies examining other PISA regions are needed to cement the validity of our argument. Furthermore, while the advantage of statistical techniques is their ability to uncover the phenomenon or universal principles related to the contribution of reading programmes to mitigating educational inequity, it is meaningful to understand contextual and subjective meanings behind this phenomenon. This situation indicates a need for qualitative studies on this topic, the findings of which will complement the quantitative approach adopted by this research.</p> <hd id="AN0187391082-26">FUNDING INFORMATION</hd> <p>This study has received no funding.</p> <hd id="AN0187391082-27">CONFLICT OF INTEREST STATEMENT</hd> <p>The authors declare no conflict of interest.</p> <hd id="AN0187391082-28">DATA AVAILABILITY STATEMENT</hd> <p>The PISA 2018 dataset, as our data source, is open to the public (https://<ulink href="http://www.oecd.org/pisa/data/2018database/">www.oecd.org/pisa/data/2018database/</ulink>).</p> <hd id="AN0187391082-29">ETHICS STATEMENT</hd> <p>This study complied with all necessary ethics requirements.</p> <hd id="AN0187391082-30">PATIENT CONSENT STATEMENT</hd> <p>Not applicable.</p> <hd id="AN0187391082-31">PERMISSION TO REPRODUCE MATERIAL FROM OTHER SOURCES</hd> <p>The PISA 2018 dataset, as our data source, is open to the public (https://<ulink href="http://www.oecd.org/pisa/data/2018database/">www.oecd.org/pisa/data/2018database/</ulink>).</p> <hd id="AN0187391082-32">CLINICAL TRIAL REGISTRATION</hd> <p>Not applicable.</p> <ref id="AN0187391082-33"> <title> REFERENCES </title> <blist> <bibl id="bib1" idref="ref210" type="bt">1</bibl> <bibtext> Akbaşlı, S., Şahin, M., & Yaykiran, Z. (2016). The effects of reading comprehension on performance in science and mathematics. Journal of Education and Practice, 7 (16), 108 – 121.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref126" type="bt">2</bibl> <bibtext> Araujo, L., & Costa, P. (2015). Home book reading and reading achievement in EU countries: The Progress in International Reading Literacy Study 2011 (PIRLS). Educational Research and Evaluation, 21, 422 – 438.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref90" type="bt">3</bibl> <bibtext> Archer, M. (1995). Realist social theory: The morphogenetic approach. Cambridge University Press.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref91" type="bt">4</bibl> <bibtext> Archer, M. (2003). Structure, agency and the internal conversation. Cambridge University Press.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref54" type="bt">5</bibl> <bibtext> Ball, S. J., Bowe, R., & Gewirtz, S. (1997). Circuits of schooling: A sociological exploration of parental choice of school in social‐class context. In A. H. Halsey, H. Lauder, P. Brown, & A. S. Wells (Eds.), Education, culture, economy, and society (pp. 409 – 421). Oxford University Press.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref93" type="bt">6</bibl> <bibtext> Beck, U., & Beck‐Gernsheim, E. (2006). Beyond status and class. In H. Lauder, P. Brown, A. Dillabough, & A. H. Halsey (Eds.), Education, globalization, and social change (pp. 143 – 151). Oxford University Press.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref114" type="bt">7</bibl> <bibtext> Bernstein, B. (1996). Pedagogy, symbolic control and identity. Taylor & Francis.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref183" type="bt">8</bibl> <bibtext> Bilton, C., & Duff, A. (2024). Improving literacy in Key Stage 2: Guidance report. https://d2tic4wvo1iusb.cloudfront.net/production/eef‐guidance‐reports/literacy‐ks2/EEF‐Improving‐literacy‐in‐key‐stage‐2‐report‐Second‐edition.pdf?v=1714992663</bibtext> </blist> <blist> <bibl id="bib9" idref="ref62" type="bt">9</bibl> <bibtext> Blanden, J., & Gregg, P. (2004). Family income and educational attainment: A review of approaches and evidence for Britain. Oxford Review of Economic Policy, 20 (2), 245 – 263.</bibtext> </blist> <blist> <bibtext> Bodovski, K., Jeon, H., & Byun, S. Y. (2017). Cultural capital and academic achievement in post‐socialist Eastern Europe. British Journal of Sociology of Education, 38 (6), 887 – 907.</bibtext> </blist> <blist> <bibtext> Börjesson, M., Broady, D., Roux, B. L., Lidegran, I., & Palme, M. (2016). Cultural capital in the elite subfield of Swedish higher education. Poetics, 56, 15 – 34.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1971a). Systems of education and systems of thought. In M. F. D. Young (Ed.), Knowledge and control (pp. 189 – 207). Collier‐Macmillan.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1971b). Intellectual field and creative project. In M. F. D. Young (Ed.), Knowledge and control (pp. 161 – 188). Collier‐Macmillan.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1973). Cultural reproduction and social reproduction. In R. Brown (Ed.), Knowledge, education and cultural change (pp. 71 – 84). Tavistock.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1977). Outline of a theory of practice. Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1990a). Photography: A middle‐brow art. Stanford University Press.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1990b). In other words: Essays towards a reflective sociology. Stanford University Press.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1992). Reproduction in education, society and culture. Sage.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1993). The field of cultural production: Essays on art and literature. Polity.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1997). The forms of capital. In A. H. Halsey, H. Lauder, P. Brown, & A. S. Wells (Eds.), Education, culture, economy, and society (pp. 241 – 258). Oxford University Press.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (1998). Practical reason: On the theory of action. Polity.</bibtext> </blist> <blist> <bibtext> Bourdieu, P. (2000). Pascalian meditations. Polity.</bibtext> </blist> <blist> <bibtext> Bourdieu, P., & Passeron, C. (1977). Reproduction in education, society and culture. Sage.</bibtext> </blist> <blist> <bibtext> Bourdieu, P., & Wacquant, L. J. D. (1992). An invitation to reflective sociology. The University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Bowyer‐Crane, C., Snowling, M. J., Duff, F. J., Fieldsend, E., Carroll, J. M., Miles, J., Götz, K., & Hulme, C. (2008). Improving early language and literacy skills: Differential effects of an oral language versus a phonology with reading intervention. Journal of Child Psychology and Psychiatry, 49 (4), 422 – 432.</bibtext> </blist> <blist> <bibtext> Buchon‐Stoll, B. (2002). The influence of storybook reading on language development. Language Matters: Studies in the Language of Africa, 33 (1), 25 – 48.</bibtext> </blist> <blist> <bibtext> Can, D. (2020). The mediator effect of reading comprehension in the relationship between logical reasoning and word problem solving. Participatory Educational Research, 7 (3), 230 – 246.</bibtext> </blist> <blist> <bibtext> Caro, D. H., Sandoval‐Hernández, A., & Lüdtke, O. (2014). Cultural, social, and economic capital constructs in international assessments: An evaluation using exploratory structural equation modeling. School Effectiveness and School Improvement, 25 (3), 433 – 450.</bibtext> </blist> <blist> <bibtext> Chiang, T. H., Ma, X., Zhang, R., Thurston, A., Jiang, S., Lin, S. W., & Cockerill, M. (2024). Examining the inner logic of students' coding orientations and the internal structure of written language in math tests from Basil Bernstein's code theory. International Journal of Educational Research, 124, 102307.</bibtext> </blist> <blist> <bibtext> Chiang, T. H., Thurston, A., & Cockerill, M. (2022). Examining Basil Bernstein's rules of recognition and realization in the case of underachieving students in math tests. International Journal of Educational Research, 115, 102021.</bibtext> </blist> <blist> <bibtext> Chiang, T. H., Thurston, A., & Lin, H. C. (2020). How the excellent working‐class student becomes a cultural capital constructor: Reflections on the theories of cultural reproduction. International Journal of Educational Research, 101, 103625.</bibtext> </blist> <blist> <bibtext> Chiang, T. H., Toh, C. W., Zhang, R., Thurston, A., & MacKenzie, A. (2022). The embeddedness of visionary agency within the economic‐cultural‐capital formation: A case of Sino‐Malaysian high school graduates. International Journal of Educational Research, 112, 101946.</bibtext> </blist> <blist> <bibtext> Chiang, T. H., & Trezise, D. (2017). Practicing duality in cultural capital constructors, China. In M. Peters (Ed.), Encyclopedia of educational philosophy and theory (pp. 1 – 7). Springer. https://doi.org/10.1007/978‐981‐287‐532‐7_626‐1</bibtext> </blist> <blist> <bibtext> Choi, Y., Kim, S. W., & Hong, W. P. (2019). Is the role of cultural capital in student achievement in South Korea different? A systematic review. British Journal of Sociology of Education, 40 (6), 776 – 794.</bibtext> </blist> <blist> <bibtext> Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Erlbaum.</bibtext> </blist> <blist> <bibtext> Covay, E., & Carbonaro, W. (2010). After the bell: Participation in extracurricular activities, classroom behavior, and academic achievement. Sociology of Education, 83 (1), 20 – 45.</bibtext> </blist> <blist> <bibtext> Cromley, J. (2009). Reading achievement and science proficiency: International comparisons from the Programme on International Student Assessment. Reading Psychology, 30, 89 – 118. https://doi.org/10.1080/02702710802274903</bibtext> </blist> <blist> <bibtext> Davey, G. (2009). Using Bourdieu's concept of habitus to explore narratives of transition. European Educational Research Journal, 8 (2), 276 – 284.</bibtext> </blist> <blist> <bibtext> Department for Education. (2021). Using PISA 2018 to inform policy: Learning from the Republic of Ireland. Her Majesty's Stationary Office. https://dera.ioe.ac.uk/id/eprint/37975/1/Using_PISA_2018_to_inform_policy_learning_from_the_republic_of_Ireland_research_brief.pdf</bibtext> </blist> <blist> <bibtext> Dickinson, D. K., & Anastasopoulos, L. (2003). The comprehensive language approach to early literacy: The interrelationships among vocabulary, phonological sensitivity, and print knowledge among preschool‐aged children. Journal of Educational Psychology, 95 (3), 465 – 481.</bibtext> </blist> <blist> <bibtext> Downey, D. B., & Condron, D. J. (2016). Fifty years since the Coleman Report: Rethinking the relationship between schools and inequity. Sociology of Education, 89 (3), 207 – 220.</bibtext> </blist> <blist> <bibtext> Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14 (1), 4 – 58.</bibtext> </blist> <blist> <bibtext> Fuchs, L. S., Fuchs, D., Compton, D. L., Hamlett, C. L., & Wang, A. Y. (2015). Is word‐problem solving a form of text comprehension? Scientific Studies of Reading, 19, 204 – 223.</bibtext> </blist> <blist> <bibtext> Fuchs, L. S., Fuchs, D., Stuebing, K., Fletcher, J. M., Hamlett, C. L., & Lambert, W. E. (2008). Problem‐solving and computation skills: Are they shared or distinct aspects of mathematical cognition? Journal of Educational Psychology, 100, 30 – 47.</bibtext> </blist> <blist> <bibtext> Giddens, A. (1979). Central problems in social theory: Action, structure and contradiction in social analysis. Macmillan.</bibtext> </blist> <blist> <bibtext> Giddens, A. (1984). The constitution of society. University of California Press.</bibtext> </blist> <blist> <bibtext> Griffin, T. M., Hemphill, L., Camp, L., & Wolf, D. P. (2004). Oral discourse in the preschool years and later literacy skills. First Language, 24 (2), 123 – 147.</bibtext> </blist> <blist> <bibtext> Gu, Y., & Lau, K. (2023). Reading instruction and reading engagement and their relationship with Chinese students' PISA reading performance: Evidence from B‐S‐J‐Z, Hong Kong, and Chinese Taipei. International Journal of Educational Research, 120, 102202. https://doi.org/10.1016/j.ijer.2023.102202</bibtext> </blist> <blist> <bibtext> Guevara, J. P., Erkoboni, D., Gerdes, M., Winston, S., Sands, D., Rogers, K., Haecker, T., Jimenez, M. E., & Mendelsohn, A. L. (2020). Effects of early literacy promotion on child language development and home reading environment: A randomized controlled trial. The Journal of Pediatrics, X (2), 100020.</bibtext> </blist> <blist> <bibtext> Hek, M. V., & Kraaykamp, G. (2015). How do parents affect cultural participation of their children? Testing hypotheses on the importance of parental example and active parental guidance. Poetics, 52, 124 – 138.</bibtext> </blist> <blist> <bibtext> Hoff, E. (2012). Causes and consequences of SES‐related differences in parent‐to‐child speech. In M. H. Bornstein & R. H. Bradley (Eds.), Socioeconomic status, parenting, and child development (pp. 147 – 160). Routledge.</bibtext> </blist> <blist> <bibtext> Holland, D., Lachicotte, W. S. J., Skinner, D., & Cain, C. (2001). Identity and agency in cultural worlds. Harvard University Press.</bibtext> </blist> <blist> <bibtext> Hulme, C., & Snowling, M. J. (2014). The interface between spoken and written language: Developmental disorders. Philosophical Transactions of the Royal Society, 369 (1634), 20120395.</bibtext> </blist> <blist> <bibtext> Jæger, M. M. (2011). Does cultural capital really affect academic achievement? New evidence from combined sibling and panel data. Sociology of Education, 84 (4), 281 – 298.</bibtext> </blist> <blist> <bibtext> Jang, J.‐Y., & Hand, B. (2017). Examining the value of a scaffolded critique framework to promote argumentative and explanatory writings within an argument‐based inquiry approach. Research in Science Education, 47 (6), 1213 – 1231.</bibtext> </blist> <blist> <bibtext> Jin, J., & Ball, S. (2021). Precarious success and the conspiracy of reflexivity: Questioning the 'habitus transformation' of working‐class students at elite universities. Critical Studies in Education, 62 (5), 608 – 623.</bibtext> </blist> <blist> <bibtext> Judkins, D. R. (1990). Fay's method for variance estimation. Journal of Statistics, 6 (3), 223 – 229.</bibtext> </blist> <blist> <bibtext> Katartzi, E., & Hayward, G. (2020). Conceptualizing transitions from vocational to higher education: Bringing together Bourdieu and Bernstein. British Journal of Sociology of Education, 41 (3), 299 – 314.</bibtext> </blist> <blist> <bibtext> Korotaeva, I. (2012). Metacognitive strategies in reading comprehension of education majors. Procedia ‐ Social and Behavioral Sciences, 69, 1895 – 1900.</bibtext> </blist> <blist> <bibtext> Lambdin, D. V. (2003). Benefits of teaching through problem solving. In F. Lester (Ed.), Research and issues in teaching mathematics through problem solving (pp. 2 – 15). National Council for Curriculum and Assessment.</bibtext> </blist> <blist> <bibtext> Lancee, B., & Werfhorst, H. V. (2012). Income inequality and participation: A comparison of 24 European countries. Social Science Research, 41 (5), 1166 – 1178.</bibtext> </blist> <blist> <bibtext> Lareau, A. (2015). Cultural knowledge and social inequality. American Sociological Review, 80 (1), 1 – 27.</bibtext> </blist> <blist> <bibtext> Laukaityte, I., & Wiberg, M. (2017). Using plausible values in secondary analysis in large‐scale assessments. Communications in Statistics ‐ Theory and Methods, 46 (22), 11341 – 11357.</bibtext> </blist> <blist> <bibtext> Lehmann, W. (2013). Habitus transformation and hidden injuries: Successful working‐class university students. Sociology of Education, 87 (1), 1 – 15.</bibtext> </blist> <blist> <bibtext> Lin, S. W., Tzou, H. I., Lu, I. C., & Hung, P. H. (2021). Taiwan: Performance in the Programme for International Student Assessment. In N. Crato (Ed.), Improving a country's education. Springer. https://doi.org/10.1007/978‐3‐030‐59031‐4_10</bibtext> </blist> <blist> <bibtext> Linberg, T., Schneider, T., Waldfogel, J., & Wang, Y. (2019). Socioeconomic status gaps in child cognitive development in Germany and the United States. Social Science Research, 79, 1 – 31.</bibtext> </blist> <blist> <bibtext> Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data. Wiley.</bibtext> </blist> <blist> <bibtext> Locher, F. M., & Pfost, M. (2020). The relationship between time spent reading and reading comprehension throughout lifecourse. Journal of Research in Reading, 43 (1), 57 – 77.</bibtext> </blist> <blist> <bibtext> Logan, J. M., Castel, A. D., Haber, S., & Viehman, E. J. (2012). Metacognition and the spacing effect: The role of repetition, feedback, and instruction on judgments of learning for massed and spaced rehearsal. Metacognition and Learning, 7 (3), 175 – 195.</bibtext> </blist> <blist> <bibtext> Loh, C. E., & Sun, B. (2020). Cultural capital, habitus and reading futures: Middle‐class adolescent students' cultivation of reading disposition in Singapore. British Journal of Sociology of Education, 41 (2), 234 – 252.</bibtext> </blist> <blist> <bibtext> MacKenzie, A., Chiang, T. H., & Thurston, A. (2022). New insights on the persistence and reproduction of educational inequality and injustice: Towards a synthesis of Nussbaum's capabilities approach and Bourdieu's theories. International Journal of Educational Research, 115, 102032.</bibtext> </blist> <blist> <bibtext> Matherly, L. L., Amin, N., & Nahyan, A. S. K. A. (2017). The impact of generation and socioeconomic status on the value of higher education in the UAS: A longitudinal study. International Journal of Educational Development, 55, 1 – 10.</bibtext> </blist> <blist> <bibtext> Mikus, K., Tieben, N., & Schober, P. S. (2020). Children's conversion of cultural capital into educational success: The symbolic and skill‐generating functions of cultural capital. British Journal of Sociology of Education, 41 (2), 197 – 217.</bibtext> </blist> <blist> <bibtext> Mol, S. E., & Bus, A. G. (2011). To read or not to read: A meta‐analysis of print exposure from infancy to early adulthood. Psychological Bulletin, 137 (2), 267 – 296.</bibtext> </blist> <blist> <bibtext> Møllegaard, S., & Jæger, M. M. (2015). The effect of grandparents' economic, cultural, and social capital on grandchildren's educational success. Research in Social Stratification and Mobility, 42, 11 – 19.</bibtext> </blist> <blist> <bibtext> National Institute for Education Statistics. (2022). Socioeconomic status (SES). Institute of Education Sciences, U.S. Department of Education. https://web.archive.org/web/20081210031419/<ulink href="http://nces.ed.gov/programs/coe/glossary/s.asp">http://nces.ed.gov/programs/coe/glossary/s.asp</ulink></bibtext> </blist> <blist> <bibtext> Neri, N. C., Guill, K., & Retelsdorf, J. (2021). Language in science performance: Do good readers perform better? European Journal of Psychology of Education, 36, 45 – 61.</bibtext> </blist> <blist> <bibtext> Notten, N., Lancee, B., Werfhorst, H. G. V., & Ganzeboom, H. B. G. (2015). Educational stratification in cultural participation: Cognitive competence or status motivation? Journal of Cultural Economics, 39 (2), 177 – 203.</bibtext> </blist> <blist> <bibtext> Nunes, T., Bryant, B., Strand, S., Hiller, J., Baros, R., & Miller‐Friedmann, J. (2017). Review of SES and science learning in formal educational settings. University of Oxford. https://royalsociety.org/‐/media/policy/topics/education‐skills/education‐research/evidence‐review‐eef‐royalsociety‐22‐09‐2017.pdf</bibtext> </blist> <blist> <bibtext> OECD. (2009). PISA data analysis manual: SPSS (2nd ed.). OECD Publishing. https://doi.org/10.1787/9789264056275‐en</bibtext> </blist> <blist> <bibtext> OECD. (2017). PISA 2015 technical report. OECD Publishing. https://<ulink href="http://www.oecd.org/content/dam/oecd/en/about/programmes/edu/pisa/publications/technical‐report/PISA2015%5fTechRep%5fFinal.pdf">www.oecd.org/content/dam/oecd/en/about/programmes/edu/pisa/publications/technical‐report/PISA2015%5fTechRep%5fFinal.pdf</ulink></bibtext> </blist> <blist> <bibtext> OECD. (2018). PISA for development reading framework. In PISA for development assessment and analytical framework: Reading, mathematics and science. OECD Publishing. https://doi.org/10.1787/9789264305274‐4‐en</bibtext> </blist> <blist> <bibtext> OECD. (2019a). PISA 2018 database. https://<ulink href="http://www.oecd.org/pisa/data/2018database/">www.oecd.org/pisa/data/2018database/</ulink></bibtext> </blist> <blist> <bibtext> OECD. (2019b). PISA 2018: Insights and interpretations. https://<ulink href="http://www.oecd.org/pisa/PISA%202018%20Insights%20and%20Interpretations%20FINAL%20PDF.pdf">www.oecd.org/pisa/PISA%202018%20Insights%20and%20Interpretations%20FINAL%20PDF.pdf</ulink></bibtext> </blist> <blist> <bibtext> OECD. (2019c). PISA 2018 assessment and analytical framework. OECD Publishing. https://doi.org/10.1787/b25efab8‐en</bibtext> </blist> <blist> <bibtext> OECD. (2019d). PISA 2018 results: Vol. 1. What students know and can do. OECD Publishing. https://doi.org/10.1787/5f07c754‐en</bibtext> </blist> <blist> <bibtext> OECD. (2019e). PISA 2018 results: Vol. 2. Where all students can succeed. OECD Publishing. https://doi.org/10.1787/b5fd1b8f‐en</bibtext> </blist> <blist> <bibtext> OECD. (2019f). PISA 2018 technical report. OECD Publishing. https://<ulink href="http://www.oecd.org/pisa/data/pisa2018technicalreport/">www.oecd.org/pisa/data/pisa2018technicalreport/</ulink></bibtext> </blist> <blist> <bibtext> Ombra, A. I. (2016). Effects of reading skills on students' performance in science and mathematics in public and private secondary schools. Journal of Education and Learning, 10 (2), 177 – 186.</bibtext> </blist> <blist> <bibtext> Pearson, P. D., Moje, E., & Greenleaf, C. (2010). Literacy and science: Each in the service of the other. Science, 328 (5977), 459 – 463.</bibtext> </blist> <blist> <bibtext> Pehlivanoglu, D., Lin, T., Deceus, F., Heemskerk, A., Ebner, N., & Cahill, B. (2021). The role of analytical reasoning and source credibility on the evaluation of real and fake full‐length news articles. Cognitive Research: Principles and Implications, 6, 24.</bibtext> </blist> <blist> <bibtext> Pishghadam, R. (2011). Parental education and social and cultural capital in academic achievement. International Journal of English Linguistics, 1 (2), 50 – 57.</bibtext> </blist> <blist> <bibtext> Pollard‐Durodola, S. D., Gonzalez, J. E., Saenz, L., Soares, D., Resendez, N., Kwork, O., Davis, H., & Zhu, L. (2016). The effects of content‐related shared book reading on the language development of preschool dual language learners. Early Childhood Research Quarterly, 36, 106 – 121.</bibtext> </blist> <blist> <bibtext> Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Sage.</bibtext> </blist> <blist> <bibtext> Reay, D. (1998). Cultural reproduction: Mothers' involvement in their children's primary schooling. In M. Grenfell & D. James (Eds.), Bourdieu and education: Acts of practical theory (pp. 55 – 71). Falmer.</bibtext> </blist> <blist> <bibtext> Reay, D., Crozier, G., & Clayton, J. (2009). 'Strangers in paradise'? Working‐class students in elite universities. Sociology, 43 (6), 1103 – 1121.</bibtext> </blist> <blist> <bibtext> Rutkowski, L., Gonzalez, E., Joncas, M., & Von Davier, M. (2010). International large‐scale assessment data: Issues in secondary analysis and reporting. Educational Researcher, 39 (2), 142 – 151. https://doi.org/10.3102/0013189X10363170</bibtext> </blist> <blist> <bibtext> Saraví, G. A. (2015). Unequal inclusion: Experiences and meanings of school segmentation in Mexico. International Journal of Educational Development, 45, 152 – 160.</bibtext> </blist> <blist> <bibtext> Schoenfeld, A. H. (1985). Mathematical problem solving. Academic Press.</bibtext> </blist> <blist> <bibtext> Segers, E., & Verhoeven, L. (2016). How logical reasoning mediates the relation between lexical quality and reading comprehension. Reading and Writing, 29, 577 – 590.</bibtext> </blist> <blist> <bibtext> Shanahan, T. (2019). Improving reading comprehension in primary classes. National Council for Curriculum and Assessment.</bibtext> </blist> <blist> <bibtext> Silverman, R. D. (2007). Vocabulary development of English‐language and English‐only learners in kindergarten. The Elementary School Journal, 107 (4), 365 – 383.</bibtext> </blist> <blist> <bibtext> Sullivan, A., & Brown, M. (2015). Reading for pleasure and progress in vocabulary and mathematics. British Educational Research Journal, 41 (6), 971 – 991.</bibtext> </blist> <blist> <bibtext> Taylor, B. M., Frye, B. J., & Maruyama, G. M. (1990). Time spent reading and reading growth. American Educational Research Journal, 27 (2), 351 – 362.</bibtext> </blist> <blist> <bibtext> Thurston, A., Cockerill, M., Chiang, T.‐H., Taylor, A., & O'Keeffe, J. (2020). An efficacy randomized controlled trial of reciprocal reading in secondary schools. International Journal of Educational Research, 104, 101626.</bibtext> </blist> <blist> <bibtext> Thurston, A., Tymms, P., Merrell, C., & Conlin, N. (2012). Improving attainment across an education authority. Better, 9, 9 – 10.</bibtext> </blist> <blist> <bibtext> Tran, L. T. (2016). Mobility as 'becoming': A Bourdieuian analysis of the factors shaping international student mobility. British Journal of Sociology of Education, 37 (8), 1268 – 1289.</bibtext> </blist> <blist> <bibtext> Tymms, P., Merrell, C., Thurston, A., Andor, J., Topping, K. J., & Miller, D. J. (2011). Improving attainment across a whole district: Peer tutoring in a randomised controlled trial. School Effectiveness and School Improvement, 22 (3), 265 – 289.</bibtext> </blist> <blist> <bibtext> Verschaffel, L., Greer, B., & De Corte, E. (2000). Making sense of word problems. Swets & Zeitlinger.</bibtext> </blist> <blist> <bibtext> Vilenius‐Tuohimaa, P. M., Aunola, K., & Nurmi, J. (2008). The association between mathematical word problems and reading comprehension. Educational Psychology, 28 (4), 409 – 426.</bibtext> </blist> <blist> <bibtext> Vivas, E. (1996). Effects of story reading on language. Language Learning, 46 (2), 189 – 216.</bibtext> </blist> <blist> <bibtext> von Hippel, P. T., Workman, J., & Downey, D. B. (2018). Inequality in reading and math skills forms mainly before kindergarten: A replication, and partial correction, of "are schools the great equalizer?". Sociology of Education, 91 (4), 323 – 357.</bibtext> </blist> <blist> <bibtext> Vygotsky, L. (1985). Mind in society. Continumum.</bibtext> </blist> <blist> <bibtext> Wasik, B. A., Bond, M. A., & Hindman, A. (2006). The effects of language and literacy intervention on Head Start children and teachers. Journal of Educational Psychology, 98, 63 – 74.</bibtext> </blist> <blist> <bibtext> Webb, P. (2010). Science education and literacy: Imperatives for the developed and developing world. Science, 328 (5977), 448 – 450.</bibtext> </blist> <blist> <bibtext> Whorrall, J., & Cabell, S. Q. (2016). Supporting children's oral language development in the preschool classroom. Early Childhood Education Journal, 44 (4), 335 – 341.</bibtext> </blist> <blist> <bibtext> Xu, J., & Hampden‐Thompson, G. (2012). Cultural reproduction, cultural mobility, cultural resources, or trivial effect? A comparative approach to cultural capital and educational performance. Comparative Education Review, 56 (1), 98 – 124.</bibtext> </blist> <blist> <bibtext> Yore, L. D., Hand, B., Goldman, S. R., Hildebrand, G. M., Osborne, J. F., Treagust, D. F., & Wallace, C. S. (2004). New directions in language and science education research. Reading Research Quarterly, 39, 347 – 352. https://doi.org/10.1598/RRQ.39.3.8</bibtext> </blist> <blist> <bibtext> Yu, H. (2021). Cross‐country comparability and validity of metacognition of assessing online information credibility in PISA 2018: Evidence from 37 OCED countries. International Journal of Social Science and Education Research, 4 (12), 509 – 519.</bibtext> </blist> <blist> <bibtext> Zarifa, D., Kim, J., Seward, B., & Walters, D. (2018). What's taking you so long? Examining the effects of social class on completing a bachelor's degree in four years. Sociology of Education, 91 (4), 290 – 322.</bibtext> </blist> </ref> <aug> <p>By Tien‐Hui Chiang; Pei‐Ming Chiang; Su‐Wei Lin and Allen Thurston</p> <p>Reported by Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib66" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib75" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib72" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib17" firstref="ref4"></nolink> <nolink nlid="nl5" bibid="bib21" firstref="ref5"></nolink> <nolink nlid="nl6" bibid="bib22" firstref="ref6"></nolink> <nolink nlid="nl7" bibid="bib14" firstref="ref7"></nolink> <nolink nlid="nl8" bibid="bib18" firstref="ref8"></nolink> <nolink nlid="nl9" bibid="bib34" firstref="ref9"></nolink> <nolink nlid="nl10" bibid="bib50" firstref="ref10"></nolink> <nolink nlid="nl11" bibid="bib58" firstref="ref11"></nolink> <nolink nlid="nl12" bibid="bib61" firstref="ref12"></nolink> <nolink nlid="nl13" bibid="bib19" firstref="ref15"></nolink> <nolink nlid="nl14" bibid="bib31" firstref="ref17"></nolink> <nolink nlid="nl15" bibid="bib71" firstref="ref18"></nolink> <nolink nlid="nl16" bibid="bib32" firstref="ref19"></nolink> <nolink nlid="nl17" bibid="bib107" firstref="ref20"></nolink> <nolink nlid="nl18" bibid="bib33" firstref="ref21"></nolink> <nolink nlid="nl19" bibid="bib114" firstref="ref25"></nolink> <nolink nlid="nl20" bibid="bib26" firstref="ref26"></nolink> <nolink nlid="nl21" bibid="bib47" firstref="ref27"></nolink> <nolink nlid="nl22" bibid="bib102" firstref="ref28"></nolink> <nolink nlid="nl23" bibid="bib105" firstref="ref29"></nolink> <nolink nlid="nl24" bibid="bib40" firstref="ref31"></nolink> <nolink nlid="nl25" bibid="bib53" firstref="ref32"></nolink> <nolink nlid="nl26" bibid="bib116" firstref="ref33"></nolink> <nolink nlid="nl27" bibid="bib13" firstref="ref35"></nolink> <nolink nlid="nl28" bibid="bib92" firstref="ref43"></nolink> <nolink nlid="nl29" bibid="bib117" firstref="ref44"></nolink> <nolink nlid="nl30" bibid="bib20" firstref="ref45"></nolink> <nolink nlid="nl31" bibid="bib11" firstref="ref46"></nolink> <nolink nlid="nl32" bibid="bib16" firstref="ref47"></nolink> <nolink nlid="nl33" bibid="bib12" firstref="ref48"></nolink> <nolink nlid="nl34" bibid="bib95" firstref="ref55"></nolink> <nolink nlid="nl35" bibid="bib120" firstref="ref58"></nolink> <nolink nlid="nl36" bibid="bib36" firstref="ref64"></nolink> <nolink nlid="nl37" bibid="bib54" firstref="ref65"></nolink> <nolink nlid="nl38" bibid="bib62" firstref="ref68"></nolink> <nolink nlid="nl39" bibid="bib15" firstref="ref71"></nolink> <nolink nlid="nl40" bibid="bib24" firstref="ref75"></nolink> <nolink nlid="nl41" bibid="bib38" firstref="ref76"></nolink> <nolink nlid="nl42" bibid="bib56" firstref="ref77"></nolink> <nolink nlid="nl43" bibid="bib64" firstref="ref78"></nolink> <nolink nlid="nl44" bibid="bib96" firstref="ref79"></nolink> <nolink nlid="nl45" bibid="bib41" firstref="ref83"></nolink> <nolink nlid="nl46" bibid="bib112" firstref="ref84"></nolink> <nolink nlid="nl47" bibid="bib45" firstref="ref88"></nolink> <nolink nlid="nl48" bibid="bib46" firstref="ref89"></nolink> <nolink nlid="nl49" bibid="bib52" firstref="ref92"></nolink> <nolink nlid="nl50" bibid="bib78" firstref="ref94"></nolink> <nolink nlid="nl51" bibid="bib25" firstref="ref105"></nolink> <nolink nlid="nl52" bibid="bib103" firstref="ref106"></nolink> <nolink nlid="nl53" bibid="bib43" firstref="ref113"></nolink> <nolink nlid="nl54" bibid="bib30" firstref="ref115"></nolink> <nolink nlid="nl55" bibid="bib29" firstref="ref116"></nolink> <nolink nlid="nl56" bibid="bib77" firstref="ref117"></nolink> <nolink nlid="nl57" bibid="bib118" firstref="ref118"></nolink> <nolink nlid="nl58" bibid="bib79" firstref="ref119"></nolink> <nolink nlid="nl59" bibid="bib37" firstref="ref120"></nolink> <nolink nlid="nl60" bibid="bib49" firstref="ref121"></nolink> <nolink nlid="nl61" bibid="bib111" firstref="ref122"></nolink> <nolink nlid="nl62" bibid="bib10" firstref="ref124"></nolink> <nolink nlid="nl63" bibid="bib70" firstref="ref125"></nolink> <nolink nlid="nl64" bibid="bib73" firstref="ref129"></nolink> <nolink nlid="nl65" bibid="bib93" firstref="ref134"></nolink> <nolink nlid="nl66" bibid="bib83" firstref="ref135"></nolink> <nolink nlid="nl67" bibid="bib80" firstref="ref136"></nolink> <nolink nlid="nl68" bibid="bib84" firstref="ref137"></nolink> <nolink nlid="nl69" bibid="bib87" firstref="ref138"></nolink> <nolink nlid="nl70" bibid="bib86" firstref="ref139"></nolink> <nolink nlid="nl71" bibid="bib85" firstref="ref141"></nolink> <nolink nlid="nl72" bibid="bib97" firstref="ref142"></nolink> <nolink nlid="nl73" bibid="bib63" firstref="ref143"></nolink> <nolink nlid="nl74" bibid="bib81" firstref="ref144"></nolink> <nolink nlid="nl75" bibid="bib67" firstref="ref145"></nolink> <nolink nlid="nl76" bibid="bib94" firstref="ref146"></nolink> <nolink nlid="nl77" bibid="bib57" firstref="ref148"></nolink> <nolink nlid="nl78" bibid="bib88" firstref="ref150"></nolink> <nolink nlid="nl79" bibid="bib35" firstref="ref153"></nolink> <nolink nlid="nl80" bibid="bib60" firstref="ref154"></nolink> <nolink nlid="nl81" bibid="bib89" firstref="ref155"></nolink> <nolink nlid="nl82" bibid="bib110" firstref="ref156"></nolink> <nolink nlid="nl83" bibid="bib55" firstref="ref157"></nolink> <nolink nlid="nl84" bibid="bib90" firstref="ref158"></nolink> <nolink nlid="nl85" bibid="bib115" firstref="ref159"></nolink> <nolink nlid="nl86" bibid="bib27" firstref="ref164"></nolink> <nolink nlid="nl87" bibid="bib99" firstref="ref166"></nolink> <nolink nlid="nl88" bibid="bib100" firstref="ref167"></nolink> <nolink nlid="nl89" bibid="bib113" firstref="ref168"></nolink> <nolink nlid="nl90" bibid="bib109" firstref="ref169"></nolink> <nolink nlid="nl91" bibid="bib44" firstref="ref171"></nolink> <nolink nlid="nl92" bibid="bib28" firstref="ref172"></nolink> <nolink nlid="nl93" bibid="bib51" firstref="ref173"></nolink> <nolink nlid="nl94" bibid="bib76" firstref="ref174"></nolink> <nolink nlid="nl95" bibid="bib98" firstref="ref176"></nolink> <nolink nlid="nl96" bibid="bib23" firstref="ref179"></nolink> <nolink nlid="nl97" bibid="bib101" firstref="ref184"></nolink> <nolink nlid="nl98" bibid="bib106" firstref="ref185"></nolink> <nolink nlid="nl99" bibid="bib68" firstref="ref186"></nolink> <nolink nlid="nl100" bibid="bib74" firstref="ref187"></nolink> <nolink nlid="nl101" bibid="bib104" firstref="ref188"></nolink> <nolink nlid="nl102" bibid="bib59" firstref="ref189"></nolink> <nolink nlid="nl103" bibid="bib91" firstref="ref190"></nolink> <nolink nlid="nl104" bibid="bib119" firstref="ref191"></nolink> <nolink nlid="nl105" bibid="bib42" firstref="ref192"></nolink> <nolink nlid="nl106" bibid="bib69" firstref="ref193"></nolink> <nolink nlid="nl107" bibid="bib108" firstref="ref197"></nolink> <nolink nlid="nl108" bibid="bib82" firstref="ref211"></nolink> <nolink nlid="nl109" bibid="bib39" firstref="ref212"></nolink> <nolink nlid="nl110" bibid="bib65" firstref="ref213"></nolink> <nolink nlid="nl111" bibid="bib48" firstref="ref214"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1480416
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: The Effects of Schools Providing Compensatory Cultural Capital on Student Reading in the Case of Taiwanese Students Participating in PISA
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Tien-Hui+Chiang%22">Tien-Hui Chiang</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-8255-1049">0000-0001-8255-1049</externalLink>)<br /><searchLink fieldCode="AR" term="%22Pei-Ming+Chiang%22">Pei-Ming Chiang</searchLink><br /><searchLink fieldCode="AR" term="%22Su-Wei+Lin%22">Su-Wei Lin</searchLink><br /><searchLink fieldCode="AR" term="%22Allen+Thurston%22">Allen Thurston</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-9518-3257">0000-0001-9518-3257</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22British+Educational+Research+Journal%22"><i>British Educational Research Journal</i></searchLink>. 2025 51(4):1820-1852.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 33
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Secondary+Education%22">Secondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Cultural+Capital%22">Cultural Capital</searchLink><br /><searchLink fieldCode="DE" term="%22Socioeconomic+Status%22">Socioeconomic Status</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Achievement+Tests%22">Achievement Tests</searchLink><br /><searchLink fieldCode="DE" term="%22International+Assessment%22">International Assessment</searchLink><br /><searchLink fieldCode="DE" term="%22Secondary+School+Students%22">Secondary School Students</searchLink><br /><searchLink fieldCode="DE" term="%22Reading%22">Reading</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Habits%22">Reading Habits</searchLink><br /><searchLink fieldCode="DE" term="%22Cognitive+Development%22">Cognitive Development</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+Strategies%22">Learning Strategies</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Taiwan%22">Taiwan</searchLink>
– Name: SubjectThesaurus
  Label: Assessment and Survey Identifiers
  Group: Su
  Data: <searchLink fieldCode="SU" term="%22Program+for+International+Student+Assessment%22">Program for International Student Assessment</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1002/berj.4157
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0141-1926<br />1469-3518
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The structural approach of cultural capital theories neglects the idea that the impact of structural constraints on educational results can be reduced by agency. This predicament can be solved when the advantages of abundant educational resources available in schools are unpacked, since doing so can compensate for the paucity of such resources often seen in the low-socioeconomic status (SES) family social space. Although the gains available from such resources remain embedded in the school social space, it can be assumed that their compensatory function can be activated through reading activities that contribute significantly to students' cognitive development. This situation prompts two research questions related to the contributions of reading to academic disciplines such as mathematics and science, and the amelioration of the structural impact of SES and cultural capital on low-SES students' learning outcomes. To explore these questions, we used regression analysis and hierarchical linear modelling (HLM) to analyse data from the stratified random sample of Taiwanese students (n = 7342) contained in the Programme for International Student Assessment (PISA) 2018 dataset. The estimates of regression analysis showed that reading ability functioned as a reliable indicator of Taiwanese students' mathematics and science scores on PISA 2018. The results of HLM analysis further demonstrated that the predominant influence of economic, social and cultural status and cultural capital can be attenuated significantly when the independent variables of school support and student personal efforts/learning strategies are considered. Accordingly, reading resources can be regarded as a compensatory genre of cultural capital embedded within the school social space, at least in the case of Taiwan, as the benefits they create need to be achieved through reading plans/projects scientifically implemented by schools.
– 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: EJ1480416
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1480416
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1002/berj.4157
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 33
        StartPage: 1820
    Subjects:
      – SubjectFull: Cultural Capital
        Type: general
      – SubjectFull: Socioeconomic Status
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Achievement Tests
        Type: general
      – SubjectFull: International Assessment
        Type: general
      – SubjectFull: Secondary School Students
        Type: general
      – SubjectFull: Reading
        Type: general
      – SubjectFull: Reading Habits
        Type: general
      – SubjectFull: Cognitive Development
        Type: general
      – SubjectFull: Learning Strategies
        Type: general
      – SubjectFull: Taiwan
        Type: general
      – SubjectFull: Program for International Student Assessment
        Type: general
    Titles:
      – TitleFull: The Effects of Schools Providing Compensatory Cultural Capital on Student Reading in the Case of Taiwanese Students Participating in PISA
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Tien-Hui Chiang
      – PersonEntity:
          Name:
            NameFull: Pei-Ming Chiang
      – PersonEntity:
          Name:
            NameFull: Su-Wei Lin
      – PersonEntity:
          Name:
            NameFull: Allen Thurston
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 08
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0141-1926
            – Type: issn-electronic
              Value: 1469-3518
          Numbering:
            – Type: volume
              Value: 51
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: British Educational Research Journal
              Type: main
ResultId 1