Teacher Support for Students’ Reading Performance: The Mediating Role of Emotional and Cognitive Engagement

Abstract

The study examined the relationship between teacher support and reading performance via the multiple mediation effects of reading engagement (emotional engagement and cognitive engagement). Data were extracted from Programme for International Students Assessment (PISA) 2018, with a sample of 171,429 adolescents from 5,692 schools across 22 countries/regions. Using multilevel serial mediation analysis, the results confirmed that teacher support enhanced adolescents’ reading performance by fostering reading engagement. Emotional engagement and cognitive engagement worked in tandem and sequentially as serial mediators. Explicitly, the positive impact of teacher support on reading performance was positively mediated through emotional engagement solely, and serially mediated through emotional engagement and then cognitive engagement. Cognitive engagement also exerted significant mediating effects in most territories. The current study provides new findings in depicting the sole and simultaneous functions of emotional and cognitive engagement in relations with teacher support and reading performance with cross-national evidence. It also suggests the practical implications for instructors to apply various pedagogical interventions to support adolescents’ reading performance growth.

Keywords

teacher support PISA reading emotional engagement cognitive engagement digital reading

Introduction

Skilled reading performance is key to adolescents’ lifelong learning and full participation in modern societies. Among the various determinants, teacher support has proved to be significantly influential on reading achievement (Wang & Hu, 2022; Guay et al., 2019; Khine et al., 2023; Ma et al., 2021). If student perceived teacher support for them, they exhibited higher level of engagement and would be more likely to show positive emotions (e.g., enjoyment) and use metacognitive strategies in learning process, thus gain better academic achievement (Förtsch et al., 2016; Tao et al., 2022). According to the model of reading engagement (Guthrie & Wigfield, 2000), reading engagement is also pivotal in developing reading performance, and has been acknowledged to mediates the instructional effects on reading outcome (Wigfield et al., 2008). Nevertheless, few studies shed lights on how individual engagements would separately or jointly predict reading development and the factors that promote them (Jang et al., 2022). Especially, we have far less knowledge about the influences from the teacher-related factors (Gu & Lau, 2023).

Therefore, this study aims to reveal the synergistic effects on reading growth from two contributing dimensions: teachers who are important others for students in the school setting, and individual factors which are determining for student development. It also provides valuable insights to address the discrepancy by examining the impact of teacher support on reading performance through the multiple mediation effects of reading engagement (emotional and cognitive aspects). Specifically, this study investigates the serial mediating effect of emotional engagement and cognitive engagement, and how it relates to teacher support and reading performance. This study offers empirical evidence to the reading engagement theory by exploring the mediating role of reading engagement between teacher support and reading achievement from diverse cultural contexts. Meanwhile, it also provides pedagogical implications on reading in provoking students’ positive emotions and activating their cognitions, which contribute much to the educational psychology research. Based on above, the multilevel serial mediation is adopted in the sample of 15-year-old students to gain international evidence on the interrelationships among studied variables.

Literature Review

Teacher Support, Reading Engagement and Reading Performance

Teacher support refers to the extent to which students perceive the teachers’ help on their study (Organisation for Economic Cooperation and Development [OECD], 2019; Wentzel, 2016). An autonomy supportive teacher acknowledges students’ perspectives (Bureau et al., 2021), provides opportunities for choice (Reeve & Cheon, 2021), or motivates them to participate in the interesting learning activities (Wentzel, 2016). These instructional behaviors that promote student intrinsic motivation are termed as autonomy support (Jiang & Zhang, 2021; Tao et al., 2022). Extends beyond creating an emotionally supportive climate, teacher support also involves direct strategy instruction, which activates students’ cognition by relating to their own experience (Wentzel, 2016), building prior knowledge into new concepts (Kitsantas et al., 2021), relating facts and ideas (Förtsch et al., 2016) or using questioning techniques (Law, 2011). These instructional behaviors that support students in academics are termed as academic support (Patrick et al., 2007; Tao et al., 2022). In the reading context, the PISA-related study conceptualizes teacher supportive behaviors in encouraging students’ independent thinking and free expressions of their own opinions about a text (Yu et al., 2022), and explicit explanations on reading text, such as to relate the reading materials to real life or build the text information on previous knowledge (F. Zhang et al., 2024).

Grounded in the model of reading engagement, individual factors hold distinctive importance on reading development. It also assumed that this decisive impact could be influenced by classroom instruction, combining motivational enhancing practices, and reading strategy instruction (Guthrie & Wigfield, 2000). A supportive teacher plays a vital role in providing such classroom-based instructional practices. Therefore, the model underpins the theoretical foundation of our conceptual framework in which we aimed to explore the relationship among support from teacher (who provides classroom instruction), reading engagement and reading performance. As the model posited, reading engagement incorporated various aspects (Guthrie & Wigfield, 2000). According to the Programme for International Assessment (PISA) 2018, reader’s emotional and cognitive aspects of engagement are the consistently robust predictors of reading performance (OECD, 2019). Substantial PISA-related studies have conceived of reading engagement a joint functioning of motivation and metacognitive strategies (Gu & Lau, 2023; Jang et al., 2022). Followed the PISA 2018 framework and previous studies, we explore reading engagement as a two-dimensional construct, namely, emotional engagement, and cognitive engagement and will further clarify these conceptual terms in the following sections.

The effect of teacher support on reading achievement has been extensively identified but yielded inconsistent results (Guthrie & Klauda, 2014; Kikas et al., 2015; Wang & Hu, 2022). When teachers utilized motivational strategies and provided cognitive support, students exhibited higher reading proficiency in Hong Kong (Law, 2011). However, in a longitudinal study (Silinskas et al., 2016), 13% of Finnish-speaking children’s reading skills showed no associations with teacher support, while 49% of children acquired reading skills slowly, irrespective of teachers’ support in reading. Kikas et al. (2015) discovered a negative correlation between teacher individualized support and reading proficiency of Estonian children. The reasons behind the inconsistency might be attributed to the multifaceted nature of teacher support and the differentiation of research types. In the previous studies, teacher support was evaluated either on classroom level (e.g., Law, 2011) or individual level (e.g., Kikas et al., 2015). Some longitudinal studies are restricted to the sample size, thus cannot involves an extensive sampled population. Therefore, this study adopted a secondary analysis extracting massive quantitative data from a large-scale assessment PISA that involves different countries/regions, to gain more empirical evidence on the inconsistent results. In this study, we examined the teacher support for the entire class and hypothesized it can facilitate successful reading performance by fostering student engagement across countries/regions.

Teacher Support, Emotional Engagement, and Reading Performance

Emotional engagement refers to the emotional responses and motivational commitment to reading (Gu & Lau, 2023; Ho & Lau, 2018). Reading enjoyment is a vital form of intrinsic reading motivation pertaining to the emotional aspect of reading engagement (Mak et al., 2016), which is one of the most frequently investigated variable that measures positive and enjoyable feelings about reading. Considerable studies have verified that emotional engagement can be affected by teacher support (Guay et al., 2019; Khine et al., 2023). Teacher autonomy support contributed to student intrinsic motivation and reading achievement in a large longitudinal study from Canada (Guay et al., 2019). Additionally, reading enjoyment is better predictive of reading performance than other moderators, such as extrinsic motivation or amount of reading (Taboada et al., 2008; Wang & Guthrie, 2004), with significant cross-country difference (Cheema, 2018; Ho & Lau, 2018). Students who read for enjoyment or are intrinsically motivated to read appear to be more skilled in reading (Mullis et al., 2007). Some researchers have further confirmed that teacher support can also boost reading improvement by fostering academic enjoyment in the mainland China (Hu & Wang, 2022a; Ma et al., 2021). The above cited studies provide the impetus for the present study, unfortunately, only the Chinese sample were examined. The present study aimed to make an extension of their studies with a larger size and more various sample to get a broader view of this issue. Therefore, it is hypothesized that emotional engagement could mediate the relationship between teacher support and reading performance across countries/regions.

Teacher Support, Cognitive Engagement, and Reading Performance

Cognitive engagement is learners’ self-regulated psychological efforts in learning process. Readers are cognitively engaged when they exhibit deliberate, selective, and sustained attention to comprehend reading tasks (Guthrie et al., 2012). While reading printed or digital information, readers are required to retrieve the target information, make inferences, or stay selective and assess the sources of information to minimize their visits to irrelevant pages (OECD, 2011). These abilities draw on the readers’ involvement of metacognitive strategies, that is, understanding of literal meaning, summary of main idea as well as the assessment of text quality and credibility (OECD, 2011). The cognitive engagement is widely operationalized as learning strategies (Appleton et al., 2006; Tao et al., 2022; Walker et al., 2005), correspondingly metacognitive reading strategies in PISA studies (Jang et al., 2022; Mak et al., 2016). Therefore, the three facets of PISA metacognitive reading strategies lie in well with the manifestation of reading engagement, which are integral for reading growth, and underlie the construct of cognitive engagement in this study.

The perceived teacher support is identified to have beneficial influence on cognitive engagement (Lau & Chen, 2013). Young (2005) also demonstrated that US students employ learning strategies more extensively when taught by caring teachers who give supportive feedback. Numerous studies have also discovered a positive correlation between metacognitive knowledge and reading performance (Gu & Lau, 2023; Jang et al., 2022; Trigueros et al., 2020). Prominent and self-regulated readers can effectively interact with a text by using more appropriate reading strategies, thus are more high-achieving in the reading test (L. Zhang et al., 2014).

To conclude, previous studies have supported that cognitive engagement is significantly associated with both teacher support and reading proficiency. However, little is known about whether teacher support can enhance students reading performance by promoting their cognitive engagement. The similar interrelations were merely investigated in chemistry or mathematics (González & Paoloni, 2015; Yıldırım, 2012). The scarcity of studies has probed into the multilevel mediation hypothesis of these variables in terms of adolescents’ reading. To fill this gap, this study hypothesized that cognitive engagement could mediate the relationship between teacher support and reading performance.

The Associations Between Emotional and Cognitive Engagement

Motivation researchers have delved into how motivation affect metacognition. Pekrun (2006) highlighted that those academic emotions (e.g., enjoyment) influence various cognitive processes and metacognitive knowledge. Intrinsic motivation is viewed as an energizer that helps students cognitively engaged during reading. For instance, emotionally engaged readers, who have stronger desire to read, will get deeper impression of what they have read and are better at connecting text will prior knowledge in future reading (Taboada et al., 2008). Empirical evidence also indicated that that some motivational attributes affect how readers approach (by use of strategy and metacognition) reading text (Wigfield & Tonks, 2004) or influence depth of cognitive processing (Schiefele, 1999) to promote text comprehension.

Recent researchers explored whether the impact of emotional engagement on cognitive engagement will subsequently be passed on the reading performance (Miyamoto et al., 2019). In German, sufficient empirical evidence confirmed the mediating role of metacognitive strategies in the relationship between intrinsic motivation and reading comprehension (Schiefele et al., 2012; van Kraayenoord & Schneider, 1999). Based on the above literature, emotional engagement can exert either direct effect on reading performance, or indirect effect through cognitive engagement. Given the solid bond of emotional engagement with teacher support, it is assumed that teacher support could enhance emotional engagement and this enhancement would have a chain effect on reading performance through cognitive engagement. As a result, we conjectured that the effect of teacher support on reading performance was serially mediated by emotional engagement and then cognitive engagement.

The Present Study

In light of previous literature review, perceived support from teacher holds distinctive importance on student emotional engagement (Guay et al., 2019; Hu & Wang, 2022b; Ma et al., 2021) and cognitive engagement (Lau & Chen, 2013), and the two aspects of reading engagement also significantly related to reading performance (Cheema, 2018; Gu & Lau, 2023; Ho & Lau, 2018; Jang et al., 2022; Trigueros et al., 2020; L. Zhang et al., 2014). Despite these direct effects, the indirect effects of teacher support on reading performance have not yet been fully explored. Although the model of reading engagement has highlighted the mediating role of reading engagement, the existing studies mainly focused on emotional aspect of reading engagement as a mediator (Ma et al., 2021). The cognitive factors are primarily explored in the science-related subjects (González & Paoloni, 2015; Yıldırım, 2012). Recent evidence has suggested the mediating role of cognitive factors (such as cognitive flexibility) that could benefit reading performance (Hu & Hu, 2024) and captured the interplay between emotional and cognitive processes (Miyamoto et al., 2019; Schiefele et al., 2012). Therefore, it is assumed that the subtypes of reading engagement function simultaneously to impact the relationship between teacher support and reading performance. To my knowledge, the internal mechanism behind teacher support, emotional engagement, cognitive engagement and reading performance remains an open question, which constitute the major motivation of our inquiry. To address these gaps, we proposed the conceptual framework, to reveal the under-explored relationships to enrich the existing model of reading engagement, which also offers compelling empirical evidence by delineating the constructs in the context of PISA data. Thus, we developed our hypothesized serial mediation model (see Figure 1) to depict the mechanism of how teacher support predicts reading performance. As perceived teacher support is measured at the student level, while it may vary across classes or schools. This study attempts to reduce common method variance by employing a multilevel mediation analysis. To summarize, our research will address three hypotheses as follows:

H1: The effect of teacher support on reading performance is solely mediated by emotional engagement across countries/regions.

H2: The effect of teacher support on reading performance is mediated by cognitive engagement across countries/regions.

H3: The effect of teacher support on reading performance is serially mediated by emotional engagement and then cognitive engagement across countries/regions

Figure 1.

The serial mediation model of how perceived teacher support predicts reading performance through emotional engagement and cognitive engagement.

Method

Data Source

The data were obtained from PISA 2018 database, which is the latest survey with reading as the major domain. PISA is the most influential large-scale survey administered every 3 years to assess the performance of 15-year-old students. The massive PISA data provides the most comprehensive and rigorous cross-national evidence for educational research and policymaking, thus often serves as an authoritative benchmark for evaluating the quality of educational systems worldwide. After eliminating the countries whose studied variables are not significantly correlated in the data preprocessing, we selected all the other 22 countries/regions (See details in Table 1) that performed above OECD average in PISA 2018, including a whole population of 171,429 (female = 84,905) participants from 5,692 schools, to draw on the experiences of their successful educational systems.

Table 1.

Demographic Information of Samples and Their Mean Score in Reading.

Countries/Regions	N	Percentage of female	Number of schools	Mean score
Australia	14,273	49.6	763	503
Belgium	8,475	50.4	288	493
Canada	22,653	49.9	232	520
China (B-S-J-Z)	12,058	47.9	361	555
Chinese Taipei	7,243	50.0	192	503
Denmark	7,657	49.8	348	501
Estonia	5,316	49.9	230	523
Finland	5,649	49.1	214	520
France	6,308	48.8	252	493
Germany	5,451	46.3	223	498
Hong Kong	6,034	48.9	152	524
Ireland	5,519	49.8	157	518
Japan	6,019	51.1	183	504
Macao	3,775	49.3	45	525
New Zealand	6,137	51.1	192	506
Norway	5,813	49.5	251	499
Poland	5,625	50.8	240	512
Singapore	6,648	49.1	166	549
Slovenia	6,401	46.8	345	495
Sweden	5,504	50.2	223	506
United Kingdom	13,818	50.6	471	504
United States	4,838	49.1	164	505

Source. PISA 2018 database.

Variables

As an independent variable, four items were used to assess teacher support (TS) from students’ perception. The adolescents were asked to rate how frequently their teacher exhibited a particular behavior, for example, “encourages students to express their opinion about a text (ST152Q05IA)” represented autonomy support, “helps students relate the stories they read to their lives (ST152Q06IA)” represented academic support (The Cronbach’s α values ≥ .807 in all territories).

As a dependent variable, the students’ reading performance (RP) was measured by scores of the computer-based reading assessment. During the assessment, students were required to click on links, browse the website, evaluate documents and retrieve relevant information. The assessment adjusts its difficulty by dynamically extracting portions of the item pool based on how the students performed in the previous stage.

Two mediating variables were selected to capture the two subcomponents of reading engagement. Five items were used to measure emotional engagement (coded as “JOY”), for example, “Reading is one of my favourite hobbies (ST160Q02IA)” (The Cronbach’s α values ≥ .812 in all territories); Cognitive engagement was a latent variable measured by three variables of metacognitive strategies, understanding and remembering (META1), summarizing (META2) and assessing credibility (META3). META3 was newly assessed to access information credibility and quality, concerning the emerging genres of digital-based reading, such as e-mail. For each scale, students are asked to evaluate a set of strategies regarding their usefulness for solving a given reading task. Six items were used to assess META1, five items each for META2 and META3, for example, “I underline important parts of the text (ST164Q04IA)” (The Cronbach’s α values ≥ .804 in all territories).

Gender and ESCS were selected as control variables to account for individual variations. As a dichotomous variable, gender was dummy recoded as female = 1, male = 2 (Hardy, 1993).

Data Preprocessing

The data preprocessing consists of imputation, dummy coding, normalization, and correlation analysis. First, missing values were imputed by K-nearest neighbor (KNN). KNN imputation uses values of the neighbors and the weighted average of their values to fill in the unknown value, which has been widely used in to predict the missing data of large-scale assessment datasets (Hu & Yu, 2022). This was performed in R (R Core Team, 2019) using the KnnImputation function in the DMwR2 package (Torgo, 2020). Next, dummy coding was used to convert the gender variable into a dummy variable, as noted above. Data normalization was in accordance with centering and variance scaling principles. Correlation analysis was conducted to capture significant relationships among the variables. The correlations between variables for all 22 territories are calculated (see Supplemental Materials). As suggested, all the variables are significantly correlated with each other. The PISA data are highly clustered in nature, with students nested in schools and schools nested in countries. To account for the hierarchical structure of the data, intraclass correlation coefficients (ICCs) were calculated to assess the need for multilevel analysis.

Data Analysis

The present study adopted a multilevel serial multiple mediator mediation model, as there are two mediators and it assumes potential chain effects among mediators, as well as accounting for the inherent feature of hierarchical data. The statistical model can be classified as a 1-1-1 multilevel mediation model (Krull & MacKinnon, 2001), as all variables were measured at level-1 (student level). The hypothesized model was based on the multilevel structural equation model (MSEM) approach, with a measurement model, a structural model and model fit as necessary components.

The statistical model was performed with the cfa and sem functions in the lavaan package (Rosseel, 2012) using R. Initially, confirmatory factor analysis (CFA) was conducted to confirm the validity of the latent construct, that is, cognitive engagement based on three indicators. CFA was also accommodated within the multilevel framework, which permits elements of some coefficient matrices to vary at the cluster level. As for structural model, the total, indirect and direct effects were calculated to display the paths and quantify the mediation effects. To take higher-level effects into account, the scores of variables at the student level were clustered at the school level, the of which was to be analyzed at the school level. The multilevel mediation model was run at the within-and between-level spontaneously to quantify the variance of different levels. Regarding the effect size of the total indirect effects, the proportion mediated (PM) was computed as the effect size measure.

The model fits of the CFA model and mediation model were both calculated. The assessed fit indices include the comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). As a rule of thumb, CFI greater than 0.90, RMSEA lower than 0.08 and SRMR lower than 0.1 indicate an acceptable model fit (Hu & Bentler, 1999).

Results

Preliminary Analysis

The ICC above 0.059 (see Table 2 in bold) suggests that a multilevel model should be preferred (Cohen, 1988) and the students’ reading proficiency presents discrepancies across schools for all datasets. Table 2 also presents the standardized factor loadings for the latent variable. All the indicators significantly loaded on their latent variable (95% CI excluding zero) and above the cutoff value (above 0.40) of accepted standardized factor loadings.

Table 2.

ICC for Each Territory and Standardized Factor Loadings of the Indicators for Latent Variables.

Countries/Regions	Standardized factor loadings [95% CI] of META			ICC
Countries/Regions	META1	META2	META3	ICC
Australia	0.563 [0.521, 0.557]	0.667 [0.620, 0.656]	0.629 [0.583, 0.618]	0.181
Belgium	0.557 [0.490, 0.537]	0.663 [0.576, 0.623]	0.498[0.429, 0.477]	0.468
Canada	0.548 [0.514, 0.544]	0.688 [0.650, 0.680]	0.576 [0.540, 0.569]	0.153
China (B-S-J-Z)	0.600 [0.552, 0.593]	0.645 [0.586, 0.628]	0.524 [0.463, 0.505]	0.472
Chinese Taipei	0.592 [0.543, 0.595]	0.651 [0.591, 0.644]	0.513 [0.460, 0.513]	0.320
Denmark	0.631 [0.589, 0.636]	0.680 [0.636, 0.683]	0.652 [0.602, 0.648]	0.173
Estonia	0.550 [0.499, 0.559]	0.659 [0.606, 0.668]	0.575 [0.529, 0.590]	0.202
Finland	0.602 [0.543, 0.599]	0.663 [0.597, 0.653]	0.573 [0.510, 0.566]	0.082
France	0.561 [0.496, 0.552]	0.640 [0.556, 0.612]	0.506 [0.438, 0.495]	0.511
Germany	0.535 [0.450, 0.507]	0.626 [0.520, 0.575]	0.598 [0.501, 0.556]	0.501
Hong Kong	0.602 [0.543, 0.599]	0.663 [0.597, 0.653]	0.573 [0.510, 0.566]	0.324
Ireland	0.478 [0.436, 0.502]	0.591 [0.543, 0.610]	0.562 [0.518, 0.581]	0.141
Japan	0.567 [0.513, 0.569]	0.669 [0.586, 0.640]	0.634 [0.558, 0.613]	0.393
Macao	0.524 [0.471, 0.550]	0.637 [0.570, 0.654]	0.476 [0.423, 0.506]	0.325
New Zealand	0.566 [0.523, 0.578]	0.676 [0.623, 0.678]	0.596 [0.546, 0.600]	0.165
Norway	0.638 [0.591, 0.643]	0.719 [0.669, 0.721]	0.628 [0.586, 0.639]	0.095
Poland	0.559 [0.517, 0.576]	0.712 [0.656, 0.715]	0.534 [0.487, 0.546]	0.184
Singapore	0.527 [0.478, 0.532]	0.659 [0.594, 0.648]	0.609 [0.543, 0.596]	0.292
Slovenia	0.562 [0.492, 0.548]	0.629 [0.543, 0.600]	0.528 [0.462, 0.519]	0.484
Sweden	0.648 [0.597, 0.650]	0.718 [0.660, 0.713]	0.666 [0.612, 0.665]	0.165
United Kingdom	0.523 [0.488, 0.526]	0.639 [0.596, 0.634]	0.605 [0.562, 0.600]	0.178
United States	0.591 [0.478, 0.532]	0.674 [0.594, 0.648]	0.569 [0.543, 0.596]	0.171

Notes. 95% CI = 95% bias-corrected confidence intervals. 95% CI that exclude zero indicate statistically significant results. ICC values greater than 0.1 indicates the need for a multilevel model.

Multilevel Mediation Analysis

The fit indices for CFA model and multilevel mediation model are displayed in Table 3. Every model fit the data well following the commonly adopted criteria (Hu & Bentler, 1999).

Table 3.

Fit Indices of the CFA Model and Multilevel Mediation Model.

Countries/Regions	CFA model				Multilevel mediation model
Countries/Regions	CFI	RMSEA	SRMR with	SRMR between	CFI	RMSEA	SRMR with	SRMR between
Australia	0.961	0.060	0.026	0.042	0.952	0.051	0.022	0.038
Belgium	0.950	0.057	0.025	0.050	0.949	0.070	0.023	0.048
Canada	0.933	0.074	0.034	0.083	0.936	0.064	0.029	0.066
China (B-S-J-Z)	0.947	0.060	0.033	0.046	0.950	0.050	0.027	0.040
Chinese Taipei	0.958	0.055	0.030	0.012	0.960	0.046	0.023	0.021
Denmark	0.976	0.048	0.020	0.051	0.959	0.055	0.020	0.060
Estonia	0.960	0.056	0.028	0.053	0.962	0.049	0.024	0.054
Finland	0.966	0.064	0.026	0.031	0.958	0.064	0.023	0.105
France	0.962	0.052	0.029	0.014	0.954	0.052	0.026	0.023
Germany	0.973	0.047	0.023	0.023	0.958	0.054	0.024	0.021
Hong Kong	0.966	0.048	0.024	0.025	0.963	0.043	0.020	0.028
Ireland	0.937	0.068	0.037	0.024	0.938	0.060	0.030	0.032
Japan	0.946	0.067	0.033	0.023	0.948	0.055	0.026	0.023
Macao	0.928	0.059	0.034	0.053	0.923	0.053	0.028	0.079
New Zealand	0.955	0.064	0.030	0.019	0.955	0.057	0.027	0.022
Norway	0.987	0.037	0.017	0.031	0.970	0.049	0.018	0.090
Poland	0.979	0.040	0.020	0.046	0.964	0.048	0.019	0.060
Singapore	0.954	0.059	0.028	0.025	0.959	0.048	0.023	0.024
Slovenia	0.987	0.029	0.016	0.014	0.979	0.032	0.015	0.020
Sweden	0.976	0.049	0.023	0.010	0.974	0.045	0.022	0.038
United Kingdom	0.950	0.062	0.029	0.043	0.952	0.053	0.024	0.040
United States	0.944	0.068	0.032	0.070	0.943	0.060	0.029	0.067

The total effects and direct effects of teacher support on reading performance were estimated in Table 4, together with the total indirect effects. Table 5 lists all specific indirect effects of teacher support on reading performance. Only unstandardized coefficients are provided, as well as 95% bias-corrected confidence intervals (95% CI), which can be used in the significance test to substitute the traditional p value (Chen & Hu, 2020). The figures in bold indicate statistically nonsignificant results.

Table 4.

Total Indirect Effects, Total Effects and Direct Effects of Teacher Support on Reading Performance.

Countries/Regions	Total indirect effects (TS-RE-RP)		Total effects (TS-RP)		Direct effects (TS-RP)		Effect size
Countries/Regions	ab [95% CI]	SE	c [95% CI]	SE	c′ [95% CI]	SE	Effect size
Australia	0.0550 [0.0429, 0.0670]	0.0061	0.0600 [0.0452, 0.0756]	0.0078	0.0057 [−0.0081, 0.0191]	0.0069	0.917
Belgium	0.0313 [0.0199, 0.0427]	0.0058	0.0213 [0.0049, 0.0377]	0.0084	−0.0100 [−0.0254, 0.0054]	0.0079	1.470
Canada	0.0662 [0.0572, 0.0752]	0.0046	0.0594 [0.0472, 0.0716]	0.0062	−0.0070 [−0.0178, 0.0043]	0.0056	1.115
China (B-S-J-Z)	0.0360 [0.0262, 0.0453]	0.0049	0.0610 [0.0476, 0.0753]	0.0070	0.0257 [0.0123, 0.0391]	0.0068	0.590
Chinese Taipei	0.0340 [0.0204, 0.0484]	0.0072	0.0340 [0.0283, 0.0672]	0.0099	0.0133 [−0.0045, 0.0312]	0.0091	0.720
Denmark	0.1110 [0.0950, 0.1270]	0.0082	0.1550 [0.1340, 0.1750]	0.0104	0.0438 [0.0254, 0.0621]	0.0094	0.717
Estonia	0.0633 [0.0444, 0.0821]	0.0096	0.0874 [0.0624, 0.1120]	0.0128	0.0024 [0.0009, 0.0475]	0.0119	0.723
Finland	0.0566 [0.0423, 0.0709]	0.0073	0.0834 [0.0620, 0.1050]	0.0110	0.0268 [0.0068, 0.0468]	0.0102	0.679
France	0.0569 [0.0448, 0.0689]	0.0062	0.0463 [0.0285, 0.0641]	0.0091	−0.0105 [−0.0273, 0.0062]	0.0085	1.230
Germany	0.0489 [0.0332, 0.0646]	0.0080	0.0243 [0.0046, 0.0441]	0.0101	−0.0246 [−0.0432, −0.0061]	0.0095	2.010
Hong Kong	0.0570 [0.0423, 0.0709]	0.0073	0.0830 [0.0620, 0.1050]	0.0109	0.0268 [0.00682, 0.0468]	0.0102	0.679
Ireland	0.1010 [0.0807, 0.1210]	0.0102	0.1100 [0.0853, 0.1340]	0.0125	0.0090 [−0.0138, 0.0317]	0.0116	0.918
Japan	0.0370 [0.0219, 0.0513]	0.0075	0.0390 [0.0184, 0.0588]	0.0103	0.0020 [−0.0159, 0.0199]	0.0092	0.949
Macao	0.0783 [0.0605, 0.0961]	0.0091	0.0520 [0.0245, 0.0796]	0.0141	−0.00263 [−0.0529, 0.0004]	0.0136	1.505
New Zealand	0.0685 [0.0502, 0.0868]	0.0093	0.0714 [0.0486, 0.0942]	0.0116	0.0029 [−0.0174, 0.0232]	0.0104	0.959
Norway	0.1150 [0.0959, 0.1340]	0.0097	0.1350 [0.1110, 0.1580]	0.0122	0.0197 [−0.0019, 0.0413]	0.0110	0.854
Poland	0.0596 [0.0433, 0.0076]	0.0083	0.0759 [0.0518, 0.1000]	0.0123	0.0163 [−0.0060, 0.0385]	0.0114	0.786
Singapore	0.0340 [0.0183, 0.0497]	0.0080	0.0410 [0.0208, 0.0613]	0.0103	0.0071 [−0.0114, 0.0255]	0.0094	0.829
Slovenia	0.0376 [0.0257, 0.0495]	0.0061	0.0640 [0.0448, 0.0832]	0.0098	0.0264 [0.0078, 0.0451]	0.0095	0.587
Sweden	0.0400 [0.0215, 0.0584]	0.0094	0.0311 [0.0070, 0.0553]	0.0123	−0.0088 [−0.0301, 0.0124]	0.0016	1.280
United Kingdom	0.0690 [0.0568, 0.0816]	0.0063	0.0730 [0.0577, 0.0885]	0.0079	0.0039 [−0.0101, 0.0179]	0.0072	0.945
United States	0.0810 [0.0613, 0.1010]	0.0102	0.1080 [0.0826, 0.1340]	0.0132	0.0271 [0.0033, 0.0510]	0.0122	0.750

Note. 95% CI = 95% bias-corrected confidence intervals. A 95% CI that include zero indicates statistically nonsignificant results and is highlighted in bold.

Table 5.

Indirect Effects of Teacher Support on Reading Performance.

Countries/Regions	M₁(JOY)		M₂ (META)		M₃ (JOY-META)
Countries/Regions	a₁b₁ [95% CI]	SE	a₂b₂ [95% CI]	SE	a₃b₃ [95% CI]	SE
Australia	0.0175 [0.0146, 0.0203]	0.0015	0.0170 [0.0060, 0.0279]	0.0056	0.0205 [0.0175, 0.0235]	0.0015
Belgium	0.0062 [0.0041, 0.0083]	0.0011	0.0159 [0.0053, 0.0266]	0.0055	0.0092 [0.0066, 0.0117]	0.0013
Canada	0.0175 [0.0152, 0.0199]	0.0012	0.0306 [0.0226, 0.0387]	0.0041	0.0180 [0.0158, 0.0203]	0.0011
China (B-S-J-Z)	0.0297 [0.0256, 0.0338]	0.0021	−0.0080 [−0.0169, 0.0005]	0.0044	0.0142 [0.0116, 0.0168]	0.0013
Chinese Taipei	0.0124 [0.0089, 0.0160]	0.0018	0.0110 [−0.0020, 0.0234]	0.0065	0.0113 [0.0082, 0.0144]	0.0016
Denmark	0.0122 [0.0089, 0.0155]	0.0017	0.0813 [0.0658, 0.0967]	0.0079	0.0176 [0.0141, 0.0212]	0.0018
Estonia	0.0173 [0.0126, 0.0220]	0.0024	0.0232 [0.0058, 0.0406]	0.0089	0.0242 [0.0179, 0.0275]	0.0025
Finland	0.0150 [0.0111, 0.0189]	0.0020	0.0278 [0.0146, 0.0410]	0.0067	0.0138 [0.0107, 0.0169]	0.0015
France	0.0159 [0.0121, 0.0197]	0.0019	0.0284 [0.0178, 0.0391]	0.0054	0.0125 [0.0097, 0.0154]	0.0015
Germany	0.0111 [0.0077, 0.0146]	0.0017	0.0228 [0.0080, 0.0376]	0.0076	0.0150 [0.0115, 0.0186]	0.0018
Hong Kong	0.0150 [0.0111, 0.0189]	0.0020	0.0278 [0.0164, 0.0410]	0.0067	0.0138 [0.0107, 0.0169]	0.0016
Ireland	0.0239 [0.0184, 0.0293]	0.0028	0.0502 [0.0325, 0.0680]	0.0028	0.0267 [0.0212, 0.0321]	0.0028
Japan	0.0132 [0.0088, 0.0176]	0.0023	0.0160 [0.0025, 0.0286]	0.0067	0.0079 [0.0052, 0.0107]	0.0014
Macao	0.0273 [0.0208, 0.0347]	0.0035	0.0385 [0.0228, 0.0542]	0.0080	0.0120 [0.0085, 0.0156]	0.0018
New Zealand	0.0193 [0.0148, 0.0238]	0.0023	0.0257 [0.0092, 0.0422]	0.0084	0.0235 [0.0187, 0.0283]	0.0025
Norway	0.0084 [0.0043, 0.0125]	0.0021	0.0756 [0.0577, 0.0936]	0.0092	0.0308 [0.0254, 0.0363]	0.0028
Poland	0.0160 [0.0114, 0.0205]	0.0023	0.0293 [0.0147, 0.0438]	0.0074	0.0144 [0.0106, 0.0182]	0.0020
Singapore	0.0139 [0.0103, 0.0175]	0.0018	0.0066 [−0.0081, 0.0212]	0.0075	0.0136 [0.0103, 0.0168]	0.0017
Slovenia	0.0063 [0.0036, 0.0090]	0.0014	0.0229 [0.0115, 0.0342]	0.0058	0.0084[0.0062, 0.0107]	0.0012
Sweden	0.0065 [0.0034, 0.0095]	0.0016	0.0153 [−0.0022, 0.0328]	0.0089	0.0182 [0.0134, 0.0229]	0.0024
United Kingdom	0.0183 [0.0154, 0.0213]	0.0015	0.0310 [0.0200, 0.0427]	0.0058	0.0195 [0.0166, 0.0224]	0.0015
United States	0.0157 [0.0108, 0.0206]	0.0025	0.0440 [0.0249, 0.0622]	0.0095	0.0220 [0.0168, 0.0272]	0.0027

Note. 95% CI = 95% bias-corrected confidence intervals. A 95% CI that include zero indicates statistically nonsignificant results and is highlighted in bold.

Total Effects and Direct Effects of Teacher Support on Reading Performance

In Table 4, the total effects are significantly positive for all datasets. Regarding the direct effect, it appeared to be nonsignificant (c′ in bold) in most territories. The direct effect is often used as statistical evidence to measure the strength of mediation (Zhao et al., 2010), which will be discussed later.

Indirect Effects of Reading Engagement

The total indirect effect (ab) examines the mediating role of reading engagement. As Table 4 suggests, the unstandardized regression coefficients range from 0.031 to 0.115, and the effect size is reported to measure the total mediated effects of teacher support on reading performance through reading engagement. Since a suppression effect, when the direct effects c’ and mediated effects ab have opposite signs (MacKinnon et al., 2000), were detected in this study, there are some territories whose effect sizes are above 1 (see Belgium, Canada, France, Macao, Germany, Sweden). The effect sizes of the other datasets range from 58.7% to 95.9%. All indicators of effect size suggest that a large proportion of effect from teacher support on reading performance is mediated through two intermediary variables, indicating that reading engagement is a significant mediator in all territories.

The specific indirect effect represents the ability of the intermediary variable to mediate the effect controlling for all other mediators (Preacher & Hayes, 2008). Table 5 reports three mediating effects (a₁b₁, a₂b₂, a₃b₃) of two meditators, that is, JOY and META. According to Table 5, JOY is a significant and positive mediator for all datasets. The unstandardized regression coefficients range from 0.0062 to 0.0297, among which China (B-S-J-Z) exhibited the largest impact. With respect to META, 18 out of 22 territories reported significantly positive mediating effect of META between teacher support and reading performance (a₂b₂), while only a few territories, that is, China (B-S-J-Z), Chinese Taipei, Singapore, and Sweden, reported no significant mediating effect. Additionally, the relationship between teacher support and reading performance are serially mediated from JOY to META (a₃b₃ significant) in all territories. By comparing three mediating effects, META is the upmost mediator in the largest number of territories (16 out of 22), eliminating China (B-S-J-Z), Chinese Taipei, Singapore with JOY as their predominant mediator, and Australia, Estonia, Sweden reporting the most significant mediation effect through JOY and then META.

Discussion

Teacher Support, Reading Engagement, Reading Performance

Our results indicate consistency on the total effects of teacher support on reading performance across different territories. According to Cohen (1988), the effect size over 0.50 is considered large. In our study, the effect sizes of total indirect effects strongly indicates that mediation of reading engagement did exist and provide compelling evidence, from a total sample of 171,429 participants, that reading engagement mediates the relation between teachers support and reading outcomes.

Notably, the strengths of mediation differ across our sampled territories. Methodologists have proposed interpretations of the implications of various medication types. Zhao et al. (2010) asserted that full mediation indicates the validity of the hypothesized meditation framework, while partial mediation points to the likelihood of an omitted mediator. It is claimed that a significant direct effect usually results from the omission of a mediator from the model (Shrout & Bolger, 2002). Correspondingly, more than half of the territories (c′ in bold, see Table 4) exhibited a full mediation pattern, which extensively validates our mediation hypotheses as consistent with the hypothesized model. The remaining territories should be considered by future studies to explore additional mediators for a comprehensive model.

Emotional Engagement as a Mediator

The present study identifies that emotional engagement is a persistent predictor in relation to teachers’ facilitation of reading outcomes. Moreover, our study provides new findings by extending the current Chinese context in literature with the cross-national generalizability of 22 countries/regions. Teacher support can help students exert control over their study, maintain affective relationship with teachers and achieve desired competence. Thus, students who perceive teacher support for their basic needs of autonomy, relatedness and competence become intrinsic motivated to improve their academic performance (Deci & Ryan, 2000). In terms of reading, supported students are easier to gain positive reading experiences or enjoyment, consequently, improve the reading amount and breadth, strategy use and achieve higher reading achievement (Cheema, 2018; Guthrie & Wigfield, 2000).

Intriguingly, the model coefficient for China (B-S-J-Z) was largest among all contexts, and significantly larger than the other path coefficients (a₂b₂, a₃b₃). There are two possible explanations for this exception. First, recent studies have rejected the stereotypy of Chinese students as passive learners. Instead, they exhibited high level of reading enjoyment than OECD average (Gu & Lau, 2023; Ma et al., 2021; Wang & Hu, 2022). As the Confucian heritage culture place high value on education, Chinese learners maintain positive attitude toward learning, hence are more devoted and dedicated in academics (Watkins & Biggs, 2001). Second, the new reforms in the curriculum, pedagogical methods have collectively shaped the Chinese educational system with “the Chinese characteristics for a new era.” The current curriculum in mainland China attempts to abandon rote learning and shift to a humanistic and competence-based approach (Ministry of Education of the People’s Republic of China, 2022). The educational objective is underpinned by equipping the students with core competences, rather than merely knowledge, that prepare them to fulfill their potential in personal and occupational life. The primary goal for teachers is to engage students in an active, cooperative, and reflective learning environment (Tam, 2015). Therefore, Chinese teachers are guided to use innovative teaching methods to motivate and engage adolescents, which dispelled the traditional impression of the Chinese classroom as authoritarian and mechanical skill-training (Guo et al., 2018; Ma et al., 2021; Qian & Lau, 2022). The comprehensive influences from the learners, cultural backgrounds and educational systems contribute to this “Chinese exception.” Additionally, the study results also reveal that the new reform works efficiently in encouraging Chinese teachers to use diverse pedagogies to engage adolescents in light of their outperformance in the PISA 2018 reading test.

Cognitive Engagement as a Mediator

The findings confirm the mediation effect of cognitive engagement among most of the territories, supporting hypothesis 2. This is congruent with the studies that verified the parallel correlations in science literacy (Yıldırım, 2012), and one of the first study to examine these correlations in the reading domain. Among all intermediary variables, cognitive engagement proved to be the most influential in promoting reading performance in the majority of territories. It is plausible that when teachers provide explicit strategy instruction, students have a better grasp of reading strategies, and those cognitively engaged readers are more proficient in decoding a text, understanding words and constructing the meaning of a text (Miyamoto et al., 2019). Hence, teacher support can precisely predict reading comprehension by facilitating the cognitive engagement.

However, META did not serve as a sole mediator in some territories. Further examination of their path coefficients showed that TS is nonsignificantly correlated with META (a₂). According to the 2018 PISA framework, readers become independent of the teacher after these strategies have been acquired (OECD, 2019). Some European researchers also found that teacher autonomy support failed to predict students’ metacognitive strategy use in a longitudinal study by arguing that metacognition was a more stable personality trait than intrinsic motivation, thus was harder to be influenced by teacher support (Schuitema et al., 2016). In the Swedish context, the deterioration of reading literacy and its negative correlation with ICT accessibility according to the PISA results initiated the government’s concern about the excessive digitalization in education and a call to revert to traditional pedagogical methods in reading instruction (Díaz et al., 2024). Research showed that Swedish teachers resist technology integration (Syahrial et al., 2022) and lack the preparation with technology (Schmitz et al., 2023), which might result in inadequate and inefficient teacher support for students’ cognitive engagement. The evidence above could possibly explain the result why there were nonsignificant associations between TS and META in these territories.

Serial Mediator of Emotional Engagement and Cognitive Engagement

Our results suggests that adolescents who perceive support from their teachers feel more emotionally engaged and thus spontaneously use more complex reading strategies to improve reading performance. More importantly, our study fills the gap by depicting the internal mechanism and the simultaneous function of subcomponents of reading engagement that relates to teacher support and reading performance. The mechanism behind can be explained by the following assumptions. The teacher-supported students are more emotionally engaged. Such readers might easily get involved in deep interactions with the text (e.g., immerse in the plot, empathize the characters, select the relevant information), and therefore adopt sophisticated and extensive reading strategies, such as underline the important sentences, relating personal experiences to the text or assess the information source, to help interpret the text. Additionally, reading enthusiasts read and use metacognitive strategies more frequently and become more competent in reading strategy use. Even though teacher support fails to impact reading performance through metacognition, teachers can facilitate students’ reading achievement by stimulating their interest in reading, whereas students driven by intrinsic motivation will activate their metacognition when processing reading texts. This hypothesis is supported with consistent evidence from all sampled countries/regions, which generalizes the research findings and provides global researchers with cross-national evidence. As Table 5 illustrates, the serial mediating effect maintains the most significant in Estonia, compared with the other mediators. From Estonian Lifelong Learning Strategy 2020 to its follow-up of the Education Strategy 2021–2035, the key educational goals were set out to achieve the personalization and diversification of learning (Ministry of Education and Research of Estonia, 2021). Our results corroborated the implementation of educational policy decisions in cultivating Estonian population’s lifelong learning competences, such as learner autonomy, growth mindset etc.

Conclusion

This study contributes to the educational field by illustrating multiple mediation pathways among adolescents’ perceptions of teacher support, emotional engagement, cognitive engagement and reading performance, which exacts the sub-datasets of 22 countries/regions from a large-scale educational assessment. First, it reinforces the mediating role of emotional engagement in relations with teacher support and reading performance in all contexts. Second, teacher support exerts positive, significant indirect effect on reading performance through cognitive engagement in most places. The findings verify the theoretical assumptions of reading engagement theory that instructional practices can take effects through reading engagement to facilitate reading performance. Empirically, the findings provide solid and reliable evidence with a relatively wide range of sample, as few studies have extensively investigated this issue from as many datasets with diverse countries/regions at once. Thus, this study offers a comprehensive depiction by employing PISA data, which can trigger general interest and allow cross-national/regional comparisons for global educators, scholars, and policymakers. Furthermore, this study further clarifies the serial mediating effect of emotional engagement and then cognitive engagement, highlighting how two mediators work together to predict reading performance. Theoretically, it also adds new findings to the reading engagement model by revealing the internal mechanism of how teacher support predicts the reading performance via reading engagement, that is, the simultaneous function of emotional and cognitive engagement in relations with teacher support and reading performance should be taken into account. The research findings address the synergistic effects of situational factor (teacher support) and individual attribute (emotional and cognitive constructs), and suggest teachers to utilize motivational enhancing practices and cognitive activation strategies, which are of vital importance to promote reading growth. However, there might be equivalent mediators that remain to be further explored for contexts in which reading engagement serves as a partial mediator.

Implications

The mediation hypotheses of present study have great practical implications. On the national level, the government should devote to promoting the competence-oriented educational reforms and multi-dimensional evaluation. The policy initiatives should proactively support schools and teachers in establishing an inclusive environment that student’s perceptions of instruction quality and sense of fulfillment are highly valued. On the school level, teachers’ belief of multiple intelligences and sustainable development should be encouraged to allow teachers being more emotionally and cognitively supportive. As the findings indicated that students’ perception of teacher support is indirectly correlated with reading performance through emotional engagement and cognitive engagement. Therefore, it is essential that teachers attach great importance to improving reading enjoyment and metacognitive strategies. Moreover, the study showed that the effect of teacher support on emotional engagement will serially take effect on reading performance through cognitive engagement. Therefore, it is especially meaningful for adolescents to receive more joyful reading experience from teachers. For learners whose metacognition has weak or no associations with teacher support, continuous enhancement of reading enjoyment is still advisable, as it will rouse an intensive drive that promotes reading strategy use to achieve higher performance at length. To promote reading engagement, it is considered as a general rule for teachers to create an enjoyment-inducing classroom climate and provide academic scaffolding with sufficient reading strategy instructions. The process feedback from teachers can also help students realize their academic potential (Hu & Zhang, 2024).

Limitations

There are some limitations to be acknowledged. First, it is a cross-sectional study collecting data from self-report questionnaires, so causal relations could not be established. Future studies are advised to consider in-depth observations or longitudinal designs to validate the students’ perception of teacher behaviors and detect the dynamics of individual development, which may provide more conclusive evidence on the tested relations. Comparable studies can be conducted to reveal the potential reasons behind the differences across countries/regions from the perspectives of the instructional, individual and cross-cultural differences. Second, the studied variables are multiple-dimensional constructs that could be interpreted differently across studies, there might be equivalent mediators that remain to be further explored. It is beneficial to identify other individual factors such as control-value appraisals, to expand the current findings on reading instructions and learning in further studies.

Supplemental Material

sj-docx-1-sgo-10.1177_21582440251338334 – Supplemental material for Teacher Support for Students’ Reading Performance

Supplemental material, sj-docx-1-sgo-10.1177_21582440251338334 for Teacher Support for Students’ Reading Performance by Lan Lyu and Jie Hu in SAGE Open

Footnotes

ORCID iDs

Lan Lyu

Jie Hu

Ethical Considerations

This study was a secondary analysis of existing data with no identifying information from an open-access database. Hence, no ethical approval from the Institutional Ethics Committee was required.

Author Contributions

Conceptualization: Lan Lyu, Jie Hu; Methodology: Lan Lyu; Data formal analysis: Lan Lyu; Writing—original draft preparation: Lan Lyu; Writing—review & editing: Lan Lyu, Jie Hu; Funding Acquisition: Lan Lyu; Supervision: Jie Hu.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Annual Research Project [grant number: 2023RWSK03] and the Provincial Education Science Research Project [grand number: 2025SCG287].

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data that support the findings of this study are openly available in PISA 2018 Database online ().

Supplemental Material

Supplemental material for this article is available online.

References

Appleton

J. J.

Christenson

S. L.

Kim

Reschly

A. L.

(2006). Measuring cognitive and psychological engagement: Validation of the student engagement instrument. Journal of School Psychology, 44(5), 427–445. https://doi.org/10.1016/j.jsp.2006.04.002

Bureau

J. S.

Gareau

Guay

Mageau

G. A.

(2021). Investigating how autonomy-supportive teaching moderates the relation between student honesty and premeditated cheating. British Journal of Educational Psychology, 92(1), 175–193. https://doi.org/10.1111/bjep.12444

Cheema

J. R.

(2018). Adolescents’ enjoyment of reading as a predictor of reading achievement: New evidence from a cross-country survey. Journal of Research in Reading, 41(S149), S162. https://doi.org/10.1111/1467-9817.12257

Chen

(2020). ICT-related behavioral factors mediate the relationship between adolescents’ ICT interest and their ICT self-efficacy: Evidence from 30 countries. Computers & Education, 159, 104004. https://doi.org/10.1016/j.compedu.2020.104004

Cohen

(Ed.). (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Erlbaum.

Deci

E. L.

Ryan

R. M.

(2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/s15327965pli1104_01

Díaz

Nussbaum

Greiff

Santana

(2024). The role of technology in reading literacy: Is Sweden going back or moving forward by returning to paper-based reading? Computers & Education, 213, 105014. https://doi.org/10.1016/j.compedu.2024.105014

Förtsch

Werner

Dorfner

von Kotzebue

Neuhaus

B. J.

(2016). Effects of cognitive activation in biology lessons on students’ situational interest and achievement. Research in Science Education, 47(3), 559–578. https://doi.org/10.1007/s11165-016-9517-y

González

Paoloni

P. V.

(2015). Perceived autonomy-support, expectancy, value, metacognitive strategies and performance in chemistry: A structural equation model in undergraduates. Chemistry Education Research and Practice, 16(3), 640–653. https://doi.org/10.1039/c5rp00058k

10.

Guay

Stupnisky

Boivin

Japel

Dionne

(2019). Teachers’ relatedness with students as a predictor of students’ intrinsic motivation, self-concept, and reading achievement. Early Childhood Research Quarterly, 48, 215–225. https://doi.org/10.1016/j.ecresq.2019.03.005

11.

Guo

Zhang

(2018). A multilevel analysis of the effects of disciplinary climate strength on student reading performance. Asia Pacific Education Review, 19(1), 1–15. https://doi.org/10.1007/s12564-018-9516-y

12.

Guthrie

J. T.

Klauda

S. L.

(2014). Effects of classroom practices on reading comprehension, engagement, and motivations for adolescents. Reading Research Quarterly, 49(4), 387–416. https://doi.org/10.1002/rrq.81

13.

Guthrie

J. T.

Wigfield

(2000). Engagement and motivation in reading. In Kamil

M. L.

Mosenthal

P. B.

Person

P. D.

Barr

(Eds), Handbook of Reading Research (Vol. 3, pp. 403–424). Routledge. https://doi.org/10.4324/9781410605023

14.

Guthrie

J. T.

Wigfield

You

(2012). Instructional contexts for engagement and achievement in reading. In Christenson

S. L.

Reschly

A. L.

Wylie

(Eds.), Handbook of research on student engagement (pp. 601–634). Springer. https://doi.org/10.1007/978-1-4614-2018-7

15.

Y. X.

Lau

K. L.

(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

16.

Hardy

M. A.

(1993). Regression with dummy variables. Sage. https://doi.org/10.4135/9781412985628

17.

E. S. C.

Lau

(2018). Reading engagement and reading literacy performance: Effective policy and practices at home and in school. Journal of Research in Reading, 41(4), 657–679. https://doi.org/10.1111/1467-9817.12246

18.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118

19.

(2024). The influence of age at first regular digital device access on digital reading performance: the mediating effect of cognitive flexibility. Humanities and Social Sciences Communications, 11(1), 799. https://doi.org/10.1057/s41599-024-03292-y

20.

Wang

(2022a). Influence of students’ perceptions of instruction quality on their digital reading performance in 29 OECD countries: A multilevel analysis. Computers & Education, 189, 104591. https://doi.org/10.1016/j.compedu.2022.104591

21.

(2022b). Impact of extracurricular synchronous and asynchronous computer-mediated communication between students and teachers on digital reading performance: Evidence from 53 countries/regions. Education and Information Technologies, 28(2), 1559–1586. https://doi.org/10.1007/s10639-022-11223-0

22.

Zhang

(2024). Growth mindset mediates perceptions of teachers’ and parents’ process feedback in digital reading performance: Evidence from 32 OECD countries. Learning and Instruction, 90, 101874. https://doi.org/10.1016/j.learninstruc.2024.101874

23.

Jang

Seo

Y. S.

Brutt-Griffler

(2022). Building academic resilience in literacy: Digital reading practices and motivational and cognitive engagement. Reading Research Quarterly, 58(1), 160–176. https://doi.org/10.1002/rrq.486

24.

Jiang

A. L.

Zhang

L. J.

(2021). University teachers’ teaching style and their students’ agentic engagement in EFL learning in China: A self-determination theory and achievement goal theory integrated perspective. Frontiers in Psychology, 12, 704269. https://doi.org/10.3389/fpsyg.2021.704269

25.

Khine

M. S.

Fraser

B. J.

Afari

Liu

(2023). Language learning environments and reading achievement among students in China: Evidence from PISA 2018 data. Learning Environments Research, 26(1), 31–50. https://doi.org/10.1007/s10984-021-09404-8

26.

Kikas

Silinskas

Soodla

(2015). The effects of children’s reading skills and interest on teacher perceptions of children’s skills and individualized support. International Journal of Behavioral Development, 39(5), 402–412. https://doi.org/10.1177/0165025415573641

27.

Kitsantas

Cleary

T. J.

Whitehead

Cheema

(2021). Relations among classroom context, student motivation, and mathematics literacy: A social cognitive perspective. Metacognition and Learning, 16(2), 255–273. https://doi.org/10.1007/s11409-020-09249-1

28.

Krull

J. L.

MacKinnon

D. P.

(2001). Multilevel modeling of individual and group level mediated effects. Multivariate Behavioral Research, 36(2), 249–277. https://doi.org/10.1207/s15327906mbr3602_06

29.

Lau

K. L.

Chen

X. B.

(2013). Perception of reading instruction and self-regulated learning: A comparison between Chinese students in Hong Kong and Beijing. Instructional Science, 41(6), 1083–1101. https://doi.org/10.1007/s11251-013-9265-6

30.

Law

Y. K.

(2011). The role of teachers’ cognitive support in motivating young Hong Kong Chinese children to read and enhancing reading comprehension. Teaching and Teacher Education, 27(1), 73–84. https://doi.org/10.1016/j.tate.2010.07.004

31.

MacKinnon

D. P.

Krull

J. L.

Lockwood

C. M.

(2000). Equivalence of the mediation, confounding and suppression effect. Prevention Science, 1(4), 173–181. https://doi.org/10.1023/a:1026595011371

32.

Mak

S. K.

Cheung

K. C.

Soh

Sit

P. S.

Ieong

M. K.

(2016). An examination of student- and across-level mediation mechanisms accounting for gender differences in reading performance: A multilevel analysis of reading engagement. Educational Psychologist, 37(10), 1206–1221. https://doi.org/10.1080/01443410.2016.1242712

33.

Luo

Xiao

(2021). Perceived teacher support, self-concept, enjoyment and achievement in reading: A multilevel mediation model based on PISA 2018. Learning and Individual Differences, 85, 101947. https://doi.org/10.1016/j.lindif.2020.101947

34.

Ministry of Education and Research of Estonia. (2021). Estonian education strategy 2021–2035. https://digital-skills-jobs.europa.eu/en/actions/national-initiatives/national-strategies/estonia-education-strategy-2021-2035

35.

Ministry of Education of the People’s Republic of China. (2022). Chinese curriculum standards for compulsory education. Beijing Normal University Publisher. http://www.moe.gov.cn/srcsite/A26/s8001/202204/t20220420_619921.html

36.

Miyamoto

Pfost

Artelt

(2019). The relationship between intrinsic motivation and reading comprehension: Mediating effects of reading amount and metacognitive knowledge of strategy use. Scientific Studies of Reading, 23(6), 445–460. https://doi.org/10.1080/10888438.2019.1602836

37.

Mullis

I. V. S.

Martin, M.O., Kennedy

A. M.

Foy

(2007). PIRLS 2006 international report: IEA’s progress in international reading literacy study in primary schools in 40 countries. Chestnut Hill. https://timssandpirls.bc.edu/pirls2006/intl_rpt.html

38.

OECD. (2011). PISA 2009 results: Students online digital technologies and performance (Vol. 3). OECD Publishing.

39.

OECD. (2019). PISA 2018 assessment and analytical framework. PISA | OECD iLibrary.

40.

Patrick

Ryan

A. M.

Kaplan

(2007). Early adolescents’ perceptions of the classroom social environment, motivational beliefs, and engagement. Journal of Education & Psychology, 99(1), 83–98. https://doi.org/10.1037/0022-0663.99.1.83

41.

Pekrun

(2006). The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review, 18(4), 315–341. https://doi.org/10.1007/s10648-006-9029-9

42.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3), 879–891. https://doi.org/10.3758/brm.40.3.879

43.

Qian

Lau

(2022). The effects of achievement goals and perceived reading instruction on Chinese student reading performance: Evidence from PISA 2018. Journal of Research in Reading, 45(1), 137–156. https://doi.org/10.1111/1467-9817.12388

44.

R Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing. http://web.mit.edu/r_v3.4.1/fullrefman.pdf

45.

Reeve

Cheon

S. H.

(2021). Autonomy-supportive teaching: Its malleability, benefits, and potential to improve educational practice. Educational Psychologist, 56(1), 54–77. https://doi.org/10.1080/00461520.2020.1862657

46.

Rosseel

(2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02

47.

Schiefele

(1999). Interest and learning from text. Scientific Studies of Reading, 3(3), 257–279. https://doi.org/10.1207/s1532799xssr0303_4

48.

Schiefele

Schaffner

Möller

Wigfield

(2012). Dimensions of reading motivation and their relation to reading behavior and competence. Reading Research Quarterly, 47(4), 427–463. https://doi.org/10.1002/rrq.030

49.

Schmitz

M. L.

Antonietti

Consoli

Cattaneo

Gonon

Petko

(2023). Transformational leadership for technology integration in schools: Empowering teachers to use technology in a more demanding way. Computers & Education, 204, 104880. https://doi.org/10.1016/j.compedu.2023.104880

50.

Schuitema

Peetsma

van der Veen

(2016). Longitudinal relations between perceived autonomy and social support from teachers and students’ self-regulated learning and achievement. Learning and Individual Differences, 49, 32–45. https://doi.org/10.1016/j.lindif.2016.05.006

51.

Shrout

P. E.

Bolger

(2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7(4), 422–445. https://doi.org/10.1037/1082-989x.7.4.422

52.

Silinskas

Pakarinen

Niemi

Lerkkanen

M. K.

Poikkeus

A. M.

Nurmi

J. E.

(2016). The effectiveness of increased support in reading and its relationship to teachers’ affect and children’s motivation. Learning and Individual Differences, 45, 53–64. https://doi.org/10.1016/j.lindif.2015.11.025

53.

Syahrial

Kurniawan

D. A.

Asrial

Sabil

Maryani

Rini

E. F. S.

(2022). Professional teachers: Study of ICT capabilities and research competencies in urban and rural? Cypriot Journal of Educational Sciences, 17(7), 2247–2261. https://doi.org/10.18844/cjes.v17i7.7590

54.

Taboada

Tonks

S. M.

Wigfield

Guthrie

J. T.

(2008). Effects of motivational and cognitive variables on reading comprehension. Reading and Writing, 22(1), 85–106. https://doi.org/10.1007/s11145-008-9133-y

55.

Tam

A. C. F.

(2015). The role of a professional learning community in teacher change: A perspective from beliefs and practices. Teachers and Teaching, 21(1), 22–43. https://doi.org/10.1080/13540602.2014.928122

56.

Tao

Meng

Gao

Yang

(2022). Perceived teacher support, student engagement, and academic achievement: A meta-analysis. Educational Psychologist, 42(4), 401–420. https://doi.org/10.1080/01443410.2022.2033168

57.

Torgo

(Ed.). (2020). Data mining with R: Learning with case studies, second edition (Chapman & Hall/CRC Data Mining and Knowledge Discovery) (2nd ed.). Chapman and Hall/CRC.

58.

Trigueros

Aguilar-Parra

J. M.

Lopez-Liria

Cangas

A. J.

González

J. J.

Álvarez

J. F.

(2020). The role of perception of support in the classroom on the students’ motivation and emotions: The impact on metacognition strategies and academic performance in Math and English classes. Frontiers in Psychology, 10, 2794. https://doi.org/10.3389/fpsyg.2019.02794

59.

van Kraayenoord

C. E.

Schneider

W. E

. (1999). Reading achievement, metacognition, reading self-concept and interest: A study of German students in grades 3 and 4. European Journal of Psychology of Education, 14(3), 305–324. https://doi.org/10.1007/bf03173117

60.

Walker

C. O.

Greene

B. A.

Mansell

R. A.

(2005). Identification with academics, intrinsic/extrinsic motivation, and self-efficacy as predictors of cognitive engagement. Learning and Individual Differences, 16(1), 1–12. https://doi.org/10.1016/j.lindif.2005.06.004

61.

Wang

J. H.

Guthrie

J. T.

(2004). Modeling the effects of intrinsic motivation, extrinsic motivation, amount of reading, and past reading achievement on text comprehension between U.S. and Chinese students. Reading Research Quarterly, 39(2), 162–186. https://doi.org/10.1598/rrq.39.2.2

62.

Wang

(2022). Perceived teacher autonomy support for adolescents’ reading achievement: The mediation roles of control-value appraisals and emotions. Frontier in Psychology, 13, 959461. https://doi.org/10.3389/fpsyg.2022.959461

63.

Watkins

D. A.

Biggs

J. B.

(Eds.) (2001). Teaaching the Chinese learner: Psychological and pedagogical perspectives. Comparative Education Research Centre.

64.

Wentzel

K. R.

(2016). Teacher-student relationships. In Wentzel

K. R.

Miele

D. B.

(Eds.), Handbook of motivation at school (pp. 211–230). Routledge. https://doi.org/10.4324/9781315773384

65.

Wigfield

Guthrie

J. T.

Perencevich

K. C.

Taboada

Klauda

S. L.

McRae

Barbosa

(2008). Role of reading engagement in mediating effects of reading comprehension instruction on reading outcomes. Psychology in the Schools, 45(5), 432–445. https://doi.org/10.1002/pits.20307

66.

Wigfield

Tonks

S. M.

(2004). The development of motivation for reading and how it is influenced by CORI. In Perencevich

K. C.

Guthrie

J. T.

(Eds.), Motivating reading comprehension: Concept-oriented reading instruction (pp. 249–272). Routledge. https://doi.org/10.4324/9781410610126

67.

Yıldırım

(2012). Teacher support, motivation, learning strategy use, and achievement: A multilevel mediation model. Journal of Experiential Education, 80(2), 150–172. https://doi.org/10.1080/00220973.2011.596855

68.

Young

M. R.

(2005). The motivational effects of the classroom environment in facilitating self-regulated learning. Journal of Marketing Education, 27, 25–40. https://doi.org/10.1177/0273475304273346

69.

Zhou

Yang

(2022). Mobile-assisted or paper-based? The influence of the reading medium on the reading comprehension of English as a foreign language. Computer Assisted Language Learning, 35(1–2), 217–245. https://doi.org/10.1080/09588221.2021.2012200

70.

Zhang

Jin

Fan

(2024). Factors influencing students’ reading literacy in Morocco: A multilevel analysis. Humanities and Social Sciences Communications, 11(1), 1056. https://doi.org/10.1057/s41599-024-03575-4

71.

Zhang

Goh

C. C. M.

Kunnan

A. J.

(2014). Analysis of test takers’ metacognitive and cognitive strategy use and EFL reading test performance: A multi-sample SEM approach. Language Assessment Quarterly, 11(1), 76–102. https://doi.org/10.1080/15434303.2013.853770

72.

Zhao

Lynch

J. G.

Chen

(2010). Reconsidering Baron and Kenny: Myths and truths about mediation analysis. Journal of Consumer Research, 37(2), 197–206. https://doi.org/10.1086/651257

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.07 MB