As Taiwan’s birth rate plunges, the number of children with an immigrant family background in schools continues to rise. Simultaneously, evidence is accumulating for various obstacles encountered by these families and children, including measurable financial and educational disadvantages, as well as immeasurable challenges caused by language and cultural differences. The literature has shown that children of immigrant families largely fall behind their nonimmigrant-family peers in several domains of development.
Inconsistent findings have been documented in recent years. Chen (2010) as well as Ho and colleagues (2011) have reported no significant difference in academic performance between nonimmigrant- and immigrant-family students. Focusing on children whose mothers were from South-East Asian countries, who generally are of lower socioeconomic status (SES) compared with their Taiwanese-native counterparts, Wu and Tsai (2014) also found no significant difference in school performance between nonimmigrant- and immigrant-family children. Finally, emerging research pointed to the possibility that immigrant-family children can even develop stronger creativity than their nonimmigrant-family counterparts (Chang et al., 2014).
Observing the inconsistency, we proposed that the relationship between child development and family backgrounds— specifically, growing up in an immigrant or nonimmigrant family—may be contextually conditioned and therefore moderated and potentially relived. Acknowledging other possibilities in the literature (e.g., parental involvement; Lahaie, 2008), we focused on one novel, potential moderating factor in the immigration context: the familial tendency of parent–child shared reading.
Parent–child shared reading is believed to facilitate young children’s language and cognitive development. Studies have indicated that 12–16-month-olds’ language ability is correlated with whether they received shared reading with parents when they were aged 8 months. Research conducted in different cultures has also supported that shared reading strengthens older children’s language development (Chow & McBride-Chang, 2003; Farrant & Zubrick, 2011) and long-term learning efficacy (Blewitt et al., 2009).
Given the central role of language in social life, it is anticipated that the benefits of parent–child shared reading extend into the socioemotional domain of development. Supporting this idea, Baker (2013) found that preschoolers’ reading ability is associated with the degree to which their parents were involved in their reading and writing activities at the age of 24 months; the effect of such shared reading in the early days is generalized to children’s later mathematic ability as well as attentional and socioemotional regulation. Critically, the effects cannot be explained away by the possibility that parents’ socioeconomic resources contribute coincidentally to both their reading time with their children and their children’s development (Hutton et al., 2015). The effects even seem to be stronger among disadvantaged children—such as those of low SES (Shahaeian et al., 2018), which most immigrant families in Taiwan experience—compared with among advantaged ones.
To help close gaps between immigrant- and nonimmigrant children in Taiwan, the present research compared the effects of parent–child shared reading on key domains of child development between the two groups. Based on the literature, we argued that shared reading may reduce disparities between children from immigrant and nonimmigrant families. Specifically, our hypotheses were as follows:
Developmental gaps. Because of obstacles they face, children with an immigrant-family background exhibit lower development compared with their nonimmigrant-family counterparts. The lack of development was expected to appear in all four domains of cognitive, language, social, and emotional.
Shared-reading effects. However, these developmental disparities would be narrowed as the amount of parent–child shared reading in families increases. We expected this buffering effect on the cognitive, language, social, and emotional domains.
Parental involvement. Given the significance of parental involvement in the child-developmental literature, that parent– child shared reading can be a mere instantiation of parental involvement, as well as anonymous reviewers’ suggestions, we considered and controlled for the effects of general parental involvement in the current study to explore the unique contributions of shared reading to child development.
We used the open dataset of “Kids in Taiwan: National Longitudinal Study of Child Development and Care” in the Survey Research Data Archive of Academic Sinica, Taiwan, to test the hypotheses. The sample was nationally representative, and 2164 participants surveyed were 36-month-old Taiwanese children (Chang, 2019); their adult caregivers answered the survey. Furthermore, we focused on 1959 participants whose fathers and mothers were both native Taiwanese as the nonimmigrant- family group of the present study, and the remaining 145 participants whose fathers were Taiwanese and mothers were of a foreign-country origin as the immigrant-family group. We excluded those whose fathers were immigrants because there were merely 28 of them.
Behavioral Rating Inventory of Cognitive Development for 2–5 Year Olds (Wang et al., 2015). This Likert-type questionnaire, included in the source data and used to assess children’s cognitive development, consists of 18 items for children’s memory ability and six for their executive functioning.
Inventory of Language Development for 0–6-Year-Old Children (Liu et al., 2018). This Likert-type questionnaire, included in the source data and used to assess children’s language development, consists of three items for language comprehension, nine for language expression, and six more for emergent literacy.
Inventory of Social and Emotional Development for Children (Chang, 2019). The Likert-type questionnaire, which was developed in the Kids in Taiwan project and included in the source data, consists of 13 items for children’s social regulation and 15 for their emotion regulation. The two subscales were used to assess children’s social and emotional development, respectively.
Parent–child shared reading index. We averaged across three Likert-type-scaled questions in the source dataset to assess shared reading: (1) Number of books at home that are suitable for the child to read (including books purchased, given, and borrowed)? (2) On average, how many times a week do you read a book to the child? and (3) On average, how much time do you spend reading a book to the child each time?
Parental involvement index. Once for the mother and a second time for the father, we averaged across the frequencies (four-point Likert-type scale) of five activities that caregivers can do with the surveyed children in the source dataset to assess parental involvement: (1) taking care of the child’s basic needs such as food and clothing; (2) teaching the child routine rules; (3) helping the child with his/her learning activities; (4) caring about and responding positively to the child; and (5) playing with the child.
Following coding schemes detailed in the main text of the paper, the results indicated that immigrant-family children had significantly older, lower-educated, and less-involved fathers; younger, lower-educated, and less involved mothers; and lived in lower-income households compared with their nonimmigrant-family counterparts. We thus controlled for these potential confounders in the later analysis of shared reading.
In terms of development, we found that immigrant-family children had significantly weaker cognitive, language, social, and emotional abilities compared with their nonimmigrant-family counterparts. These findings replicated past research, supported our hypotheses, and justified the current investigation into the possible benefits of parent–child shared reading.
We dummy coded children from immigrant families as 1 and those from nonimmigrant families as 0 to study the moderating effects of shared reading on this code of family status. In a regression wherein a domain of child development was predicted by the family-status code, mean-centered shared reading, the interaction of the two, and the aforementioned controls, the results (Figure 1) revealed that parent–child shared reading significantly buffers the detrimental effects of young children’s immigrant-family status (versus nonimmigrant) in the cognitive, language, social, and emotional development domains. Upon closer examination, simple-main-effect analysis further indicated that nonimmigrant-family children only surpassed their immigrant-family counterparts in all domains when shared reading in the families was low (i.e., at +1 SD). However, when shared reading was abundant (i.e., at + 1 SD compared with the whole sample, or at + 0.18 SD if compared with other immigrant-family children only, as the value is their “within-group” + 1 SD), we observed no relative obstacles for immigrant- family children, and their cognitive and emotional abilities were trending in the direction of exceeding those of nonimmigrant- family children.
Using a nationally representative sample, we reported high-powered evidence that parent–child shared reading, such as reading bedtime stories together, reduces and sometimes neutralizes developmental disparities encountered by 3-year-old children from immigrant families compared with those from nonimmigrant families. This beneficial effect generalizes across the cognitive, language, social, and emotional domains. When shared reading is strong, there is even an indication that immigrant- family children may perform better than their nonimmigrant-family counterparts in cognitive and emotional domains. This suggestion, nonetheless, requires scrutiny in future research that considers the contextual limitations of our present work, such as that we only considered immigrant mothers but not fathers and that immigrant families are generally disadvantaged.