Language exposure and English digital reading achievement in East Asia: Examining the role of students’ socioeconomic status

Abstract

In recent years, the importance of language exposure and socioeconomic status (SES) in shaping learners’ language learning outcomes has gained significant attention in language educational studies. However, studies that directly investigated the correlation between language exposure from parents and peers, SES, and English digital reading achievement remained limited. This present study examined how language exposure from parents and peers interacted with SES in predicting English digital reading achievement among the three East Asian economies using the PISA 2018 data with 7,703 student participants from Macao, Hong Kong, and Singapore. Results from the Pearson correlation and hierarchical linear regression revealed: (1) that there is a negative relationship between language exposure from parents and English digital reading achievement; (2) that language exposure from best friends and siblings showed a positive influence on English digital reading achievement; and (3) that the effects of language exposure from schoolmates differed among the three economies. Although SES exerted a significant and positive impact on English digital reading achievement among the three East Asian economies, SES did not moderate the relationship between language exposure and English digital reading achievement. Implications and limitations are discussed.

Keywords

English digital reading achievement hierarchical linear regressions language exposure PISA socioeconomic status

I Introduction

There is a growing trend that explores the effects of language exposure on language outcomes (Huang et al., 2018; Persici et al., 2022). Similarly, socioeconomic status (SES), which refers to individuals’ SES related to learning opportunities and resources (Destin et al., 2019), was found to be a crucial predictor of language learning achievements (Sun et al., 2015). According to the socio-ecology theory proposed by Bronfenbrenner (1979), language exposure and SES, all belong to microsystem factors. It seems that these two microsystem factors, language exposure and SES both contributed to students’ language learning achievements based on previous studies. For example, bilingual students’ exposure from their mothers showed a significant impact on their vocabulary knowledge and reading skills (Persici et al., 2022). On the other hand, SES was found to link with preschool students’ English vocabulary development (Farangi & Mehrpour, 2022).

However, the existing studies on language exposure and language achievements provided contradictory findings (e.g. Sun et al., 2018; Vaahtoranta et al., 2021), which cast the doubt on the impact of language exposure on English acquisition. For example, Persici et al. (2022) found that maternal language proficiency significantly related to Italian learners’ vocabulary and reading speed, while results of Vaahtoranta et al. (2021) indicated that parental language exposure did not affect students’ language skills, especially their vocabulary mastery. Meanwhile, the majority of studies concentrated on maternal or parental language exposure without considering a comprehensive view of language exposure both from parents and peers. Furthermore, studies on how language exposure interplays with SES in explaining digital reading achievement in English seem relatively new that still can be learned.

To fill in the research gaps, this study examined the interplay between language exposure from parents, siblings, best friends, and schoolmates, on the one hand, and SES, on the other hand, in predicting English digital reading achievement. Employing Pearson correlation and hierarchical linear regression analyses using PISA 2018 data (OECD, 2018), the research clarified the relationships among various sources of language exposure, SES, and English digital reading achievement. This study enriches the understanding of language acquisition and provides valuable insights into how SES and diverse language exposure contribute to English digital reading.

II Literature review

1 Socio-ecological theory

The socio-ecological theory, proposed by Bronfenbrenner (1979), posits that students’ learning process is influenced by various factors across multiple systems, including individual, microsystem, and mesosystem levels (Allen et al., 2021; Bronfenbrenner, 1979). According to this theory, microsystem factors refer to the immediate environment surrounding learners, encompassing both classroom and family contexts. These factors include elements such as language exposure from parents, language exposure from peers, and SES.

Existing studies have demonstrated the positive impact of language exposure and SES on students’ reading achievement (Bernardo, 2022). Recent research conducted by Luo, King, et al. (2024), utilizing machine learning methods, identified SES as the most significant predictor of English digital reading achievement among East Asian students. Additionally, both parental and peer language exposure were found to rank among the top contributing factors, based on the PISA 2018 dataset (Luo, King, et al., 2024). Given the crucial roles of SES and language exposure in influencing students’ digital reading outcomes, this research, guided by socio-ecological theory, focuses on the interplay of these essential microsystem factors. Specifically, it examined how SES, along with parental and peer language exposure, interacted to shape secondary learners’ English digital reading achievement. By investigating these relationships, the study hoped to provide a comprehensive understanding of how these variables collectively contribute to reading success in a digital context.

2 Language exposure

Based on Vaahtoranta et al. (2021), students learn a language always in a naturalistic context instead of via direct learning and teaching instruction. Language exposure includes literacy-related activities (Burgess et al., 2002), the quantity of words exposed to (Hart & Risley, 1995), the quality of speech exposed to (Huttenlocher et al., 2010), the frequency of target languages exposure, the number of people communicated with in target languages, and the linguistic proximity of language users’ first languages to second or foreign languages (Gathercole, 2016). Based on the 2018 Student Main Survey in PISA 2018 (OECD, 2018), language exposure in this current research refers to the language that students speak with their parents, siblings, and peers.

Prior literature has focused on the role of language exposure and how language exposure shape students’ language development (Persici et al., 2022). It might be attributed to the reason that native speakers were more likely to produce richer child-directed speech with more diverse types of child-directed speech (Hoff, 2020). On the contrary, language exposure at home from parents or siblings with low proficiency in that language was related to low performance in vocabulary (Lauro et al., 2020) and literacy skills (Puglisi et al., 2017). Gottardo et al. (2017) found that those Chinese–English bilinguals in Canada with less exposure to English tended to use similar word-reading processes. It was because of the essential role of vocabulary in second and foreign language acquisition (Teng & Cui, 2024) that language learners need to learn vocabulary to obtain language knowledge embedded in meaningful sentences and texts (Hsu et al., 2024). Bilingual students’ vocabulary knowledge and reading skills were found to be significantly associated with their mothers’ language skills in the same language (Persici et al., 2022).

3 Language exposure and language academic achievement

Previous studies have investigated the role of language exposure and language academic outcomes in vocabulary, reading speed, and phonological awareness (Persici et al., 2022; Vaahtoranta et al., 2021). Some prior studies found that students with more exposure to the target languages were more likely to achieve better language outcomes (Leona et al., 2021). This has been explained as the exposure to target languages (Unsworth, 2016), and the length of exposure provided students more opportunities to develop the languages (Vaahtoranta et al., 2021).

However, the existing literature has revealed that more language exposure does not always lead to students’ better language proficiency and the associations between language exposure and language performance were always direct or linear (Vaahtoranta et al., 2021). For example, recently, Persici et al.’s (2022) research among 140 primary school students in Italy found that mothers’ language proficiency significantly predicted students’ vocabulary and reading speed but had nothing to do with the accuracy of words and reading. Interestingly, it is shown that a high frequency of language exposure from the target language at home combined with low levels of maternal target language proficiency was negatively related to students’ reading (Persici et al., 2022). The high degrees of parental language exposure with lower language exposure quality resulted in worse language reception, vocabulary expression (Buac et al., 2014), and literacy skills (Puglisi et al., 2017). The reason might be attributed to students’ language achievements influenced by the quantity and quality of language exposure (Hoff, 2018, 2020).

On the other hand, some previous studies found that language exposure from parents did not affect students’ language skills (Sun et al., 2018; Vaahtoranta et al., 2021). The results of Vaahtoranta et al. (2021) among 316 preschool students in Germany indicated that language exposure at home could not significantly predict preschool students’ language vocabulary mastery. A possible explanation could be that participants received sufficient exposure to German due to the similar hours of exposure at kindergartens, so their additional language exposure at home did not further cultivate their language development (Vaahtoranta et al., 2021).

Taken together, the contradictory findings of these existing studies, which employed diverse methods, raised doubts about the influence of language exposure on language performance. Furthermore, a limitation of previous language exposure mainly focused on parents’ exposure and did not provide a holistic view of the influence of language exposure from parents but also from peers on reading achievement. Another gap among these studies is that the data were mostly collected within one country, and the sample size was not large. Given these gaps, investigating the relationship between language exposure from parents and peers, and digital reading achievement in the subject domain of English with large samples to generalize the findings seems crucial.

4 The moderation effect of SES

In the existing literature, SES was operationalized as the combination of parental income, educational levels, and occupations (Gottfried, 1985; Mueller & Parcel, 1981). SES is an essential factor that shapes students’ language development and outcomes (Huang et al., 2018). For example, students with disadvantaged SES backgrounds were more likely to learn English at a slower pace and have more reading difficulties (Huang et al., 2018). According to Hamid (2011) and Liu et al. (2016), those language learners from better SES families tended to have access to more language input and learning resources, including English books, media, and courses outside of school, which would benefit their English proficiency.

Regarding the relationship between SES and language outcomes, results shown in previous studies were mixed. For example, the results of Hamid (2011) indicated a positive link between SES on English as foreign language learners’ English proficiency outcomes among secondary students in rural Bangladesh, finding that rural students from families with lower incomes and educational levels of parents tended to exhibit low levels of English academic achievement. It may be because students from lower SES backgrounds may face challenges such as limited exposure to the English language outside of school (Liu et al., 2016), less parental involvement in their education (Scaff et al., 2022), and fewer opportunities for enrichment activities, which collectively hinder their ability to achieve proficiency in English. Similarly, using a quasi-experimental design among 60 Iranian preschool students, Farangi and Mehrpour (2022) found that SES was significantly related to students’ vocabulary development. In addition, although Butler (2013) found a significant relationship between SES and English speaking and pronunciation in elementary and middle school students in China, the effect of SES on listening, reading, and writing was not significant. On the contrary, in Sun et al.’s (2015) research among young English learners in China, results revealed that SES did not significantly predict students’ vocabulary and grammar performance. English learners’ vocabulary and grammar outcomes were influenced by language exposure at school and from media (Sun et al., 2015).

Since the effect of SES on students’ development was indirect and always moderated and mediated by other factors (Ding et al., 2024), there is a growing body of research investigating the socioeconomic factors with several psychological or language-related variables in L1 and L2 learning (Liu et al., 2016). In recent years, researchers have increasingly paid attention to the link between language exposure and SES. For example, the results of Butler (2013) revealed a significant relationship between SES and parental home literacy exposure. Using a story-telling task and a questionnaire among 97 tenth- and eleventh-grade student participants from Taiwan, Huang et al. (2018) found that language exposure was related to SES, and the relationship between SES and second language outcomes was mediated by language exposure. More recently, Scaff et al. (2022) conducted a study among 91 two- to three-year-old children in France through a paired forced-choice task to explore the correlation between SES, language exposure, and word comprehension. Results demonstrated that both SES and maternal language exposure showed a direct influence on vocabulary comprehension (Scaff et al., 2022), but the link between SES and language exposure has not been further discussed.

5 English digital reading in East Asia

The PISA test began assessing learners’ digital reading skills in 2009, with a greater emphasis on this area noted in PISA 2018 (OECD, 2019a). The digital reading tests include a variety of resources – such as textbook articles, emails, and blogs – to evaluate how effectively students engage with texts for specific purposes (OECD, 2022). To succeed in the digital reading test, students need to read fluently, accurately, and automatically to grasp the main ideas (OECD, 2022). Additionally, they are expected to possess skills in locating information (OECD, 2022), reflecting on content and form, critically assessing the quality and reliability of information, and addressing contrasting viewpoints (OECD, 2019a).

PISA participants take the reading test in multiple languages (OECD, 2019a), which means their digital reading proficiency is evaluated in various linguistic contexts across different economies (Luo, Lin, et al., 2024). For instance, some students in East Asian regions use English as a second language while completing the PISA test (OECD, 2019b).

As an international language in multilingual environments, English serves as an effective communication tool (Matsuda, 2017). In East Asia, particularly in places like Hong Kong and Singapore, English is an official language used in government and education (Zhang et al., 2021). However, there are significant disparities in English education across East Asian regions, including variations in educational policies, students’ proficiency levels (Bolton & Bacon-Shone, 2020), and English-medium instruction (Chiu & Walker, 2007). These differences may lead to varied outcomes in secondary students’ English reading abilities (Zhang et al., 2021). Given the importance of English and the diverse learning contexts in East Asia (Luo, King, et al., 2024; Saraceni, 2020), it is worth exploring students’ digital reading performance in English through the PISA 2018 dataset to investigate how the interplay between language exposure and SES may influence students’ English digital reading achievement.

III The current study

Although there is also some empirical evidence that demonstrated the relationship between language exposure, SES, and second or foreign language outcomes, research that directly examined the correlation between language exposure from parents and peers, SES, and English digital reading achievement is still limited to the best of our knowledge. As the existing literature was mainly conducted within a region with limited participants, to generalize the findings this present study also aimed to address the research gaps by exploring how language exposure interplayed with SES in predicting students’ English digital reading achievement in three East Asian economies through the use of the worldwide large-scale dataset, PISA 2018 data. This study sought to answer the following research questions:

• Research question 1: What is the relationship between language exposure from mother, SES, and English digital reading achievement?

• Research question 2: What is the relationship between language exposure from father, SES, and English digital reading achievement?

• Research question 3: What is the relationship between language exposure from siblings, SES, and English digital reading achievement?

• Research question 4: What is the relationship between language exposure from best friend, SES, and English digital reading achievement?

• Research question 5: What is the relationship between language exposure from schoolmates, SES, and English digital reading achievement?

IV Method

1 Participants

This study utilized data from the PISA 2018 (http://www.oecd.org/pisa/data). In Macao, Hong Kong, and Singapore, some students completed the reading assessment in English via computers, allowing their data to represent PISA English digital reading performance. From the PISA 2018 dataset, 7,703 participants who took the test in English were selected, comprising students from Macao (n = 679), Hong Kong (n = 348), and Singapore (n = 6,676). The sample from Hong Kong and Macao was small, with students mostly from international schools.

The selection of participants from Macao, Hong Kong, and Singapore was mainly due to two primary reasons. First, the PISA 2018 dataset indicated that students from these regions achieved high average digital reading scores of 549, 525, and 524, ranking second, third, and fourth among 76 economies and regions (OECD, 2019b). This high performance is likely attributed to Confucian cultural values, which emphasize education as a pathway for personal development (Luo, King, et al., 2024). Second, multiculturalism is a prominent and shared characteristic of the education systems in Singapore, Macao, and Hong Kong (Ustun et al., 2022), and these regions include a significant number of immigrants from diverse populations (Hirschmann, 2024).

2 Measures

a Dependent variable

To assess participants’ English digital reading performance from the three East Asian economies, this study utilized the plausible value labeled PV1READ from PISA 2018. According to OECD (2019b), participants completed a range of text-based items aimed at evaluating their reading skills and abilities, including tasks related to locating, evaluating, and reflecting on information.

b Independent variable

Two independent variables of language exposure (labeled ST023Q01TA, ST023Q02TA, ST023Q03TA, ST023Q04TA, ST023Q05TA) and socioeconomic status (labeled ESCS) from the PISA 2018 dataset were selected to represent students’ language exposure and SES. In the PISA 2018 (OECD, 2019a), language exposure refers to the amount and variety of language experiences students encounter, including their exposure to home language, siblings, best friends, and classmates. SES encompasses factors such as family income, parental education, and occupation, which can affect students’ access to resources and opportunities for learning. The selection of these two variables may provide valuable insights into how language skills and academic performance may vary across secondary learners from Macao, Hong Kong, and Singapore and shed light on the complex interplay between linguistic environments and economic backgrounds in shaping educational outcomes.

3 Data analysis

a Missing data

In this research, Markov Chain Monte Carlo (MCMC) multiple imputations in SPSS (Ni & Leonard, 2005) was used to estimate the missing data since missing data exists in PISA datasets. MCMC algorithms have satisfactory convergence properties of unbiasedness and high efficiency in dealing with complicated missingness in large-scale databases.

b Pearson correlation

In this present study, Pearson correlation was conducted among SES, language exposure factors, and students’ digital reading achievement in English with a sample of 7,703 students from Macao, Hong Kong, and Singapore using PISA 2018 data, aiming at investigating the correlation between SES, language exposure, and digital reading achievement in English.

c Hierarchical linear regression

In this current research, to further explore the relation between language exposure factors and English digital reading achievement and the moderation role of SES, the hierarchical linear regression analysis was performed: language exposure from mother (LE-M), language exposure from father (LE-F), language exposure from siblings (LE-S), language exposure from best friend (LE-B), and language exposure from schoolmates (LE-SM) were applied as independent variables, SES as the moderator, while Y represented the dependent variable (English digital reading achievement). The linear regression model equations were illustrated as followed. Notably, model 1 is the baseline model for comparison with other models.

• Model 1: Y = β₁ (SES);

• Model 2: Y = β₁ (SES) + β₂ (LE-M) + β₃ (LE-F) +β₄ (LE-S) + β₅ (LE-B) + β₆ (LE-SM);

• Model 3: Y = β₁ (SES) + β₂ (LE-M) + β₃ (LE-F) +β₄ (LE-S) + β₅ (LE-B) + β₆ (LE-SM) + β₇ (SES*LE-M) + β₈ (SES*LE-F) +β₉ (SES*LE-S) + β₁₀ (SES*LE-B) + β₁₁ (SES*LE-SM).

V Results

Table 1 demonstrates the descriptive statistics, including mean and standard deviations for the variables: language exposure factors, SES, and English digital reading achievement among Macao, Hong Kong, and Singapore. The theoretical ranges of language exposure factors ranged between 1 to 4. The mean values of the language exposure factors among the whole population ranged from 1.95 to 2.65 with the standard deviations from 0.65 to 0.94. The means of SES and English reading achievement of the whole population was 0.15 and 541, respectively. The standard deviations of SES and English reading achievement among the whole sample were 0.92 and 111, respectively.

Table 1.

Descriptive statistics.

Variables	Whole population (n = 7,703)		Macao (n = 679)		Hong Kong (n = 348)		Singapore (n = 6,676)
Variables	Mean	SD	Mean	SD	Mean	SD	Mean	SD
LE-mother	1.95	0.89	1.28	0.73	1.51	0.82	2.04	0.87
LE-father	2.05	0.93	1.36	0.83	1.58	0.90	2.15	0.91
LE-siblings	2.52	0.94	1.76	1.11	2.39	1.19	2.61	0.87
LE-friend	2.50	0.77	1.64	0.89	2.17	0.95	2.61	0.68
LE-schoolmates	2.65	0.65	1.74	0.91	2.36	0.91	2.75	0.52
SES	0.15	0.92	0.04	0.83	0.42	1.03	0.15	0.92
Reading	541.00	111.00	477.00	95.00	522.00	115.00	549.00	110.00

Notes. LE-mother = language exposure from mother; LE-father = language exposure from father; LE-siblings = language exposure from siblings; LE-friend = language exposure from best friends; LE-schoolmates = language exposure from schoolmates; SES = socioeconomic status; EDRA = English digital reading achievement.

The results of Pearson correlations among language exposure from mother, language exposure from father, language exposure from siblings, language exposure from best friends, language exposure from schoolmates, SES, and English digital reading achievement among the whole sample, Macao, Hong Kong, and Singapore are illustrated in Table 2 to Table 5, respectively.

Table 2.

Results of Pearson correlation of the whole sample.

	1	2	3	4	5	6	7
LE-mother	1
LE-father	0.73**	1
LE-siblings	0.51**	0.52**	1
LE-friend	0.47**	0.46**	0.49**	1
LE-schoolmates	0.39**	0.38**	0.40**	0.71**	1
SES	0.29**	0.26**	0.21**	0.23**	0.18**	1
EDRA	0.22**	0.22**	0.27**	0.33**	0.28**	0.38**	1

Notes. *p < 0.05; **p < 0.01. LE-mother = language exposure from mother; LE-father = language exposure from father; LE-siblings = language exposure from siblings; LE-friend = language exposure from best friends; LE-schoolmates = language exposure from schoolmates; SES = socioeconomic status; EDRA = English digital reading achievement.

Table 3.

Results of Pearson correlation in Macao.

	1	2	3	4	5	6	7
LE-mother	1
LE-father	0.78**	1
LE-siblings	0.50**	0.48**	1
LE-friend	0.62**	0.56**	0.54**	1
LE-schoolmates	0.58**	0.56**	0.49**	0.85**	1
SES	0.03	0.04	0.08*	0.14**	0.13**	1
EDRA	0.04	0.02	0.17**	0.22**	0.20**	0.34**	1

Table 4.

Results of Pearson correlation in Hong Kong.

	1	2	3	4	5	6	7
LE-mother	1
LE-father	0.81**	1
LE-siblings	0.44**	0.44**	1
LE-friend	0.53**	0.55**	0.48**	1
LE-schoolmates	0.47**	0.50**	0.38**	0.80**	1
SES	0.15**	0.08	0.21**	0.08	−0.05	1
EDRA	0.06	0.04	0.17**	0.09	−0.07	0.50**	1

Table 5.

Results of Pearson correlation in Singapore.

	1	2	3	4	5	6	7
LE-mother	1
LE-father	0.71**	1
LE-siblings	0.49**	0.50**	1
LE-friend	0.40**	0.38**	0.43**	1
LE-schoolmates	0.28**	0.27**	0.29**	0.60**	1
SES	0.33**	0.30**	0.23**	0.27**	0.23**	1
EDRA	0.20**	0.20**	0.25**	0.31**	0.27**	0.38**	1

In the whole sample and Singapore, language exposure from mother (whole sample: r = 0.22, p < 0.01; Singapore: r = 0.20, p < 0.01), language exposure from father (whole sample: r = 0.22, p < 0.01; Singapore: r = 0.20, p < 0.01), language exposure from siblings (whole sample: r = 0.27, p < 0.01; Singapore: r = 0.25, p < 0.01), language exposure from best friends (whole sample: r = 0.33, p < 0.01; Singapore: r = 0.31, p < 0.01), and language exposure from schoolmates (whole sample: r = 0.28, p < 0.01; Singapore: r = 0.27, p < 0.01), were all positively related to English digital reading achievement in the whole sample.

As for the results in Macao, except for the positive but insignificant association between language exposure from mother (r = 0.04), language exposure from father (r = 0.02), and English digital reading achievement, language exposure from siblings (r = 0.17, p < 0.01), language exposure from best friends (r = 0.22, p < 0.01), and language exposure from schoolmates (r = 0.20, p < 0.01) were significantly and positively associated with English digital reading achievement.

Regarding the results between language exposure and English digital reading achievement in Hong Kong, only language exposure from siblings significantly and positively related to English digital reading achievement (r = 0.17, p < 0.01).

Among the correlations between SES and English digital reading achievement were positive and significant among the three East Asian economies: whole sample (r = 0.38, p < 0.01), Macao (r = 0.34, p < 0.01), Hong Kong (r = 0.50, p < 0.01), and Singapore (r = 0.38, p < 0.01).

The hierarchical linear regressions have been performed through SPSS, in which the language exposure factors predicted students’ English digital reading achievement with SES as the moderator. The results of hierarchical linear regressions in Macao, Hong Kong, and Singapore are shown in Table 6.

Table 6.

Results of hierarchical linear regressions of the whole sample.

Models	Factors		Whole sample
			r²		β
Model 1	SES		0.14		0.38***
Model 2	SES		0.22		0.313***
	LE-mother				−0.03*
	LE-father				−0.00
	LE-siblings				0.12***
	LE-friend				0.16***
	LE-schoolmates				0.08***
Model 3	SES		0.219		0.19***
	LE-mother				−0.04*
	LE-father				−0.01
	LE-siblings				0.12***
	LE-friend				0.16***
	LE-schoolmates				0.08***
	SES*LE-mother				0.05
	SES*LE-father				0.01
	SES*LE-siblings				0.02
	SES*LE-friend				0.01
	SES*LE-schoolmates				0.04
Models	Macao		Hong Kong		Singapore
	r²	β	r²	β	r²	β
Model 1	0.12	0.34***	0.25	0.50***	0.14	0.38***
Model 2	0.174	0.31***	0.28	0.46***	0.20	0.31***
		−0.10		−0.07		−0.04*
		−0.11		0.01		0.00
		0.12**		0.08		0.11***
		0.18*		0.25**		0.14***
		0.06		−0.25**		0.09***
Model 3	0.208	−0.08	0.287	0.50**	0.21	0.15*
		−0.08		−0.12		−0.04**
		−0.10		0.06		−0.00
		0.12**		0.06		0.12***
		0.15*		0.26**		0.14***
		0.07		−0.24**		0.09***
		0.11		0.24		0.02
		−0.07		−0.21		0.02
		0.15		0.07		0.01
		0.06		−0.14		0.05
		0.19		0.02		0.06

Notes. *p < 0.05; **p < 0.01; ***p < 0.001. LE−mother = language exposure from mother; LE−father = language exposure from father; LE−siblings = language exposure from siblings; LE-friend = language exposure from best friends; LE-schoolmates = language exposure from schoolmates; SES = socioeconomic status; EDRA = English digital reading achievement.

The results of hierarchical linear regressions revealed a negative effect of language exposure from mother on English digital reading achievement (see Model 2): whole sample (β = –0.03, p < 0.05), Macao (β = −0.01, p > 0.05), Hong Kong (β = –0.07, p > 0.05), and Singapore (β = –0.04, p < 0.05). The effects of language exposure from father on English digital reading achievement among Macao, Hong Kong, and Singapore were all insignificant. The language exposure from siblings (whole sample: β = 0.12, p < 0.001); Macao (β = 0.12, p < 0.01; Hong Kong (β = 0.08, p > 0.05; Singapore (β = 0.11, p < 0.001), and language exposure from best friends (whole sample: β = 0.16, p < 0.001); Macao (β = 0.18, p < 0.05; Hong Kong (β = 0.25, p < 0.01; Singapore (β = 0.14, p < 0.001) showed a positive influence on English digital reading achievement among Macao, Hong Kong, and Singapore. The impacts of language exposure from schoolmates on English digital reading achievement were diverse among Macao, Hong Kong, and Singapore.

The relationships between SES and English digital reading achievement among Macao, Hong Kong, and Singapore were significant and positive (see Model 1): the whole sample (β = 0.38, p < 0.001), Macao (β = 0.34, p < 0.001), Hong Kong (β = 0.50, p < 0.001), and Singapore (β = 0.38, p < 0.001).

The indirect effects of the language exposure factors on English digital reading achievement through the moderation of SES were all insignificant among Macao, Hong Kong, and Singapore (see Model 3).

VI Discussion

This current research aimed to examine the relationship between language exposure, SES, and English digital reading achievement and the interaction between language exposure and SES in affecting English digital reading performance using the large-scale dataset, PISA 2018, among Macao, Hong Kong, and Singapore.

1 The relationship between language exposure and English digital reading achievement

Interestingly, language exposure from mother was found to have significant and negative effects on students’ English digital reading achievement among Macao, Hong Kong, and Singapore. The negative role of language exposure from mother is consistent with previous studies showing the negative correlation between maternal language exposure and language outcomes (Buac et al., 2014; Persici et al., 2022). Those students exposed to lower-quality maternal exposure were inclined to have worse reading performance. One possible explanation is that more exposure from mothers with low proficiency in English might be detrimental as low maternal proficiency in the target language was related to poorer vocabulary knowledge (Hammer et al., 2012), reading skills (Persici et al., 2022), and linguistic complexity (Vaahtoranta et al., 2021). It seems that being exposed to language exposure with a lower level of proficiency could not support students’ language acquisition (Hoff, 2018). This finding provides strong scientific support for the negative role of maternal language exposure in English reading and is aligned with Huttenlocher et al. (2010) and Persici et al. (2022), which emphasized the quality of maternal language exposure in students’ language acquisition. Furthermore, in some Asian contexts mothers often take on the responsibility of teaching vocabulary and reading skills, driven by high expectations (Liu & Chung, 2022). This achievement-oriented family atmosphere may inadvertently lead to anxiety among language learners in English learning, ultimately decreasing their success in English digital reading.

Regarding the association between language exposure from peers – including siblings, best friends, and schoolmates – the results were somewhat diverse and mixed among Macao, Hong Kong, and Singapore. The diverse findings among Macao, Hong Kong, and Singapore may be because the length of language exposure from siblings, best friends, and schoolmates may be different among the student participants in Macao, Hong Kong, and Singapore. Previous studies demonstrated that the length of exposure was an indicator of language achievement (Blom & Bosma, 2016; De Carli et al., 2015) and students’ learning efficiency in acquisition (Thompson & von Gillern, 2020). In terms of the findings from language exposure from peers on English digital reading achievement, future work is suggested to explore how language exposure from peers interacts with language outcomes.

The diverse findings also guide us to consider the design of the items that measured language exposure in the PISA 2018 student main survey. Based on PISA 2018 student questionnaire, the items of language exposure failed to capture the quality of language exposure from parents and peers, the variety of language exposure, and the quantity of language exposure, which have been proven as crucial indicators of language outcomes (Persici et al., 2022). Without taking the quality of language exposure and the variety of language exposure into consideration, the items that assessed students’ language exposure might lose an important aspect to capture the exact impact of language exposure on students’ English digital reading achievement. Thus, it is necessary to include items regarding the quality and quantity of language exposure and the variety of language exposure in the PISA questionnaire to allow more effective discussions on language exposure and language achievements.

2 The role of SES

The results of this study demonstrated that SES significantly and positively predicted students’ English digital reading achievement among Macao, Hong Kong, and Singapore. This finding corroborates with previous studies suggesting the positive effect of SES on language learning outcomes (Farangi & Mehrpour, 2022; Huang et al., 2018; Scaff et al., 2022). This may be because parents with a higher education level were more likely to own better English knowledge and skills (Sun et al., 2015), show more positive attitudes toward English learning (Ma et al., 2024), and invest resources for students’ English learning due to economic reasons (Liu et al., 2016). This study adds to the existing literature by providing scientific support in the area of digital reading among secondary school students via the sample from Macao, Hong Kong, and Singapore, which also contributes to the heated discussion regarding the effect of SES on language learning, especially English learning.

However, SES did not moderate the relationship between language exposure and English digital reading achievement among Macao, Hong Kong, and Singapore. This finding is partially aligned with the result of Prevoo et al. (2015), revealing that SES was not a moderator in language development. It should be noted that the language development in Prevoo et al. (2015) focused on vocabulary interdependence, which is different from digital reading.

The fact that there was no moderating effect of SES on the relationship between language exposure and English digital reading achievement may be attributed to several reasons. First, as existing literature revealed that the quality and frequency of language exposure – whether from parents, peers, or educational resources – play a more critical role in developing reading skills (Bernardo, 2022), this suggests that those learners with rich language interactions may perform well, and the influence of socioeconomic background would be decreased. In addition, in Macao, Hong Kong, and Singapore where educational attainment is highly valued (Sun et al., 2018; Ustun et al., 2022), there may be strong community and school support for language development, which might mitigate the disparities typically associated with SES. Furthermore, the multicultural and multilingual environments present in regions like Macao, Hong Kong, and Singapore might enhance peer interactions that provide ample language exposure across various SES groups (Ustun et al., 2022). While SES remained an important factor in educational outcomes, its effect on moderating the relationship between language exposure and reading achievement may be diminished in these three specific regions.

VII Implications

Since research on the influence of language exposure from parents and peers on English reading, particularly with SES as a moderating factor, is limited in the field of language acquisition, this study addressed a critical research gap. By utilizing Pearson correlation and hierarchical linear regression, this research clarified the mechanisms through which different sources of language exposure interact with SES to influence English digital reading performance.

The theoretical implications of these findings are noteworthy. First, the negative association between parental language exposure and English digital reading achievement suggests that not all forms of language input are beneficial to language learners, which may challenge conventional assumptions about parental involvement. This highlights the need for a more nuanced understanding of how parental language proficiency and engagement can impact student outcomes. In addition, the positive impact of language exposure from best friends and siblings emphasizes the role of peer interactions in language acquisition, suggesting that informal learning environments are crucial in shaping language outcomes. Although SES positively affected English digital reading achievement, its lack of moderating influence on the link between language exposure and English digital reading performance implies that other variables, such as cultural attitudes toward language learning and educational practices, may play a more significant role. In a word, this research enriches existing literature on language acquisition by integrating the complex interplay of parental and peer influences and SES, providing a more holistic picture of understanding students’ English digital reading achievement.

As for the practical implications, the findings supported the importance of the quality of parental language exposure and provided suggestions on measures of language exposure. This research also confirmed that SES was a significant predictor of English digital reading achievement. It encourages parents to put more effort into offering students more English learning resources related to reading. In addition, educators could give students more high-quality reading input, for example, creating meaningful opportunities (Huang et al., 2018) and ensuring equal access to digital reading for all English learners.

VIII Limitations and directions for future research

Several limitations in this study need to be noted and discussed. First, language exposure was measured through the language that students spoke with parents and peers based on the PISA 2018 questionnaire. As language exposure includes multiple aspects, such as language-related activities (Burgess et al., 2002), the quantity of vocabulary input (Hart & Risley, 1995), and the quality of language exposure (Huttenlocher et al., 2010), future studies could consider using measurements with diverse dimensions and various research designs to offer more clarity for the mechanism on language exposure.

Second, SES was only measured through home possessions, highest parental occupation, and education according to the PISA questionnaire. It should also be noted that the participants from Macao and Hong Kong mainly came from international schools, suggesting that the samples from these regions contained more affluent families than disadvantaged ones. This may influence the moderate role of SES in the relationship between language exposure and digital reading achievement. To gain a deeper understanding of the influence of SES on English digital reading, future studies are encouraged to investigate other potential mediating variables, such as attitudes, beliefs, and expectations from parents (Butler, 2013), as well as students’ learning motivation and self-efficacy (Huang et al., 2018). Third, this research only focused on digital reading in English in the language education field. We suggest future work pay attention to digital reading in other languages as well as other language skills and knowledge, such as pronunciation, vocabulary knowledge, and listening comprehension.

IX Conclusions

This research adds to the ongoing discourse on language exposure on language outcomes by investigating the correlation between language exposure from parents and peers, SES, and English digital reading achievement in three East Asian economies. This study confirmed the negative effects of parental language exposure on students’ English digital reading performance, while the association between language exposure from siblings, best friends, schoolmates, and English digital reading outcomes differed. SES did not moderate the relationship between language exposure and English digital reading achievement among the three East Asian economies.

Since language exposure is complex to be measured and would also be affected by a wide range of factors, researchers need to pay attention to the selection of instruments for language exposure and explore how language exposure interplays with other variables in explaining students’ English reading. In addition, the different association between language exposure from peers and English digital reading achievement among the three East Asian economies shown in the results of this research highlights the importance of future work to use various research designs and methodologies.

Footnotes

Author Note

This article is based on Shuqiong Luo’s PhD thesis entitled ‘Factors affecting English reading in Macau, Hong Kong, and Singapore: Combining machine learning methods and hierarchical linear regressions using PISA 2018 data’ submitted to the Faculty of Education, University of Macau.

Data availability

The dataset analysed during the current research is available in the PISA 2018 repository: .

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Shuqiong Luo

References

Allen

K.A.

Fortune

K.C.

Arslan

(2021). Testing the social-ecological factors of school belonging in native-born, first-generation, and second-generation Australian students: A comparison study. Social Psychology of Education, 24, 835–856.

Bernardo

A.B.I.

(2022). Growth mindset and reading proficiency of ESL learners: Examining the role of students’ socioeconomic status using PISA 2018 Philippine data. European Journal of Psychology of Education, 38, 675–693.

Blom

Bosma

(2016). The sooner the better? An investigation into the role of age of onset and its relation with transfer and exposure in bilingual Frisian–Dutch children. Journal of Child Language 43, 581–607.

Bolton

Bacon-Shone

(2020). The statistics of English across Asia. In: Bolton

Botha

Kirkpatrick

(Eds.), The handbook of Asian Englishes (pp. 49–80). Wiley Blackwell.

Bronfenbrenner

(1979). The ecology of human development: Experiments by nature and design. Harvard University Press.

Buac

Gross

Kaushanskaya

(2014). The role of primary caregiver vocabulary knowledge in the development of bilingual children’s vocabulary skills. Journal of Speech, Language, and Hearing Research, 57, 1804–1816.

Burgess

S.R.

Hecht

S.A.

Lonigan

C.J.

(2002). Relations of the home literacy environment (HLE) to the development of reading-related abilities: A one-year longitudinal study. Reading Research Quarterly, 37, 408–426.

Butler

(2013). Parental factors and early English education as a foreign language: A case study in Mainland China. Research Papers in Education, 29, 410–37.

Chiu

M.M.

Walker

(2007). Leadership for social justice in Hong Kong schools: addressing mechanisms of inequality. Journal of Educational Administration, 45, 724–39.

10.

De Carli

Dessi

Mariani

, et al. (2015). Language use affects proficiency in Italian–Spanish bilinguals irrespective of age of second language acquisition. Bilingualism: Language and Cognition, 18, 324–339.

11.

Destin

Hanselman

Buontempo

Tipton

Yeager

D.S.

(2019). Do student mindsets differ by socioeconomic status and explain disparities in academic achievement in the United States? AERA Open, 5(3), 2332858419857706.

12.

Ding

Zhang

Shi

(2024). The mediating role of executive function between socioeconomic status and academic achievement: A meta-analytic structural equation model. Learning and Individual Differences, 110, 102418.

13.

Farangi

M.R.

Mehrpour

(2022). Iranian preschoolers’ vocabulary development: Background television and socio-economic status. Journal of Early Childhood Literacy, 24, 422–44.

14.

Gathercole

V.C.M.

(2016). Factors moderating proficiency in bilingual speakers. In E.

Nicoladis

Montanari

(Eds.), Bilingualism across the lifespan: Factors moderating language proficiency (pp. 123–136). De Gruyter Mouton.

15.

Gottardo

Koh

P.W.

Chen

Jia

(2017). Models of English and Chinese word reading for adolescent Chinese–English bilinguals. Reading and Writing, 30, 1377–1406.

16.

Gottfried

A.E.

(1985). Academic intrinsic motivation in elementary and junior high school students. Journal of Educational Psychology, 77, 631–45.

17.

Hamid

M.O.

(2011). Socio-economic characteristics and English language achievement in rural Bangladesh. Bangladesh eJournal of Sociology, 8, 31–50.

18.

Hammer

C.S.

Komaroff

Rodriguez

B.L.

Lopez

L.M.

Scarpino

S.E.

Goldstein

(2012). Predicting Spanish–English Bilingual Children’s Language Abilities. Journal of Speech, Language, and Hearing Research, 55, 1251–1264.

19.

Hart

Risley

T.R.

(1995). Meaningful differences in the everyday experience of young American children. P.H. Brookes.

20.

Hirschmann

(2024). Number of immigrants in Singapore 2005–2019. Statista. Available at: https://www.statista.com/statistics/698035/singapore-number-ofimmigrants/#:~:text=heimmigrantpopulationofSingapore,ofthecountryin2018

21.

Hoff

(2018). Bilingual development in children of immigrant families. Child Development Perspectives, 12, 80–86.

22.

Hoff

(2020). Lessons from the study of input effects on bilingual development. International Journal of Bilingualism, 24, 82–88.

23.

Hsu

T.C.

Chang

Jen

T.H.

(2024). Artificial intelligence image recognition using self-regulation learning strategies: effects on vocabulary acquisition, learning anxiety, and learning behaviours of English language learners. Interactive Learning Environments, 32, 3060–78.

24.

Huang

Chang

S.Y.H.

Niu

Zhi

(2018). Examining the effects of socio-economic status and language input on adolescent English learners’ speech production outcomes. System, 73, 27–36.

25.

Huttenlocher

Waterfall

Vasilyeva

Vevea

Hedges

L.V.

(2010). Sources of variability in children’s language growth. Cognitive Psychology, 61, 343–365.

26.

Lauro

Core

Hoff

(2020). Explaining individual differences in trajectories of simultaneous bilingual development: Contributions of child and environmental factors. Child Development, 91(6), 2063–2082.

27.

Leona

N.L.

van Koert

M.J.

van der Molen

M.W.

Rispens

J.E.

Tijms

Snellings

(2021). Explaining individual differences in young English language learners’ vocabulary knowledge: The role of Extramural English Exposure and motivation. System, 96, 102402.

28.

Liu

Chung

K.K.H.

(2022). Effects of fathers’ and mothers’ expectations and home literacy involvement on their children’s cognitive: Linguistic skills, vocabulary, and word reading. Early Childhood Research Quarterly, 60, 1–12.

29.

Liu

Chung

K.K.

McBride

(2016). The role of SES in Chinese (L1) and English (L2) word reading in Chinese-speaking kindergarteners. Journal of Research in Reading, 39, 268–291.

30.

Luo

King

R.B.

Wang

Leung

S.O.

(2024). English digital reading achievement for East Asian students: Identifying the key predictors using a machine learning approach. Asia Pacific Journal of Education. Advance online publication. https://doi.org/10.1080/02188791.2024.2398120

31.

Luo

Lin

King

R.B.

Leung

S.O.

(2024). The influence of ICT competencies on English digital reading achievement. International Journal of Educational Development, 110, Article 103139.

32.

Xiao

Liu

(2024). Motivational beliefs of urban and rural students in English as a foreign language learning: The case of China. Journal of Multilingual and Multicultural Development, 45, 1524–1537.

33.

Matsuda

, (2017). Preparing teachers to teach English as an international language. Multilingual Matters.

34.

Mueller

C.W.

Parcel

T.L.

(1981). Measures of socioeconomic status: Alternatives and recommendations. Child Development, 52, 13–30.

35.

Leonard

J.D.

(2005). Markov chain Monte Carlo multiple imputation using Bayesian networks for incomplete intelligent transportation systems data. Transportation Research Record, 1935(1), 57–67.

36.

OECD. (2018). Student questionnaire for PISA 2018: Main survey version. OECD. Available at https://www.oecd.org/pisa/data/2018database/CY7_201710_QST_MS_STQ_NoNotes_final.pdf (accessed January 2025).

37.

OECD. (2019a). PISA 2018 assessment and analytical framework. OECD. Available at: https://www.oecd-ilibrary.org/education/pisa-2018-assessment-and-analytical-framework_b25efab8-en (accessed January 2025).

38.

OECD. (2019b). PISA 2018 results: What students know and can do. Volume I. OECD. Available at: https://doi.org/10.1787/5f07c754-en (accessed January 2025).

39.

OECD. (2022). PISA 2022 ICT framework. OECD. Available at: https://www.oecd-ilibrary.org/education/pisa-2022-assessment-and-analytical-framework_dfe0bf9c-en (accessed January 2025).

40.

Persici

Majorano

Bastianello

Hoff

(2022). Vocabulary and reading speed in the majority language are affected by maternal language proficiency and language exposure at home: a study of language minority bilingual children in Italy. International Journal of Bilingual Education and Bilingualism, 25, 3729–44.

41.

Prevoo

M.J.L.

Malda

Emmen

R.A.G.

Yeniad

Mesman

(2015). Language Use and SES as Potential Moderators. Language learning, 65, 449–69.

42.

Puglisi

M.L.

Hulme

Hamilton

L.G.

Snowling

M.J.

(2017). The home literacy environment is a correlate, but perhaps not a cause, of variations in children’s language and literacy development. Scientific Studies of Reading, 21, 498–514.

43.

Saraceni

(2020). Globalization and Asian Englishes. In: K. Bolton, W. Botha, & A. Kirkpatrick k . (Eds.), The handbook of Asian Englishes. Wiley Blackwell, 707–724.

44.

Scaff

Fibla

Cristia

(2022). Factors structuring lexical development in toddlers: The effects of parental education, language exposure, and age. Journal of Child Language, 50, 757–77.

45.

Sun

Steinkrauss

Tendeiro

De Bot

(2015). Individual differences in very young children’s English acquisition in China: Internal and external factors. Bilingualism: Language and Cognition, 19, 550–66.

46.

Sun

Yin

Amsah

N.F.B.B.

O’Brien

B.A.

(2018). Differential effects of internal and external factors in early bilingual vocabulary learning: The case of Singapore. Applied Psycholinguistics, 39, 383–411.

47.

Teng

M.F.

Cui

(2024). Comparing incidental learning of single words and collocations from different captioning conditions: The role of vocabulary knowledge and working memory. Journal of Computer Assisted Learning, 40, 973–89.

48.

Thompson

C.G.

von Gillern

(2020). Video-game based instruction for vocabulary acquisition with English language learners: A Bayesian meta-analysis. Educational Research Review, 30, 100332.

49.

Unsworth

(2016). Early child L2 acquisition: Age or input effects? Neither, or both? Journal of Child Language, 43, 608–34.

50.

Ustun

Cansiz

Ozdemir

Cansiz

(2022). Student and school-level factors to predict science literacy for two top-performing countries in PISA 2015: Finland and Singapore. International Journal of Science Education, 44, 579–603.

51.

Vaahtoranta

Suggate

Lenhart

Lenhard

(2021). Language exposure and phonological short-term memory as predictors of majority language vocabulary and phonological awareness in dual language learning. Bilingualism: Language and Cognition 24, 319–32.

52.

Zhang

Lau

(2021). Home environment and development of English as a second/foreign language for young children in Asian contexts: a systematic review and meta-analysis. Early Education and Development, 34, 1–32.