RCT of a Reading Aloud Intervention in Brazil: Do Impacts Differ Depending on Parent Literacy?

Access: Institutional Access

Publication year

2020

study description

Secondary analysis of a single-blind cluster randomized controlled trial (RCT).

core topic(s)

Shared Reading

Population Characteristics

International , Poverty/Low-Income , Toddler/Preschool , Urban

Exposures, Outcomes, Other

Child Behaviors and Skills , Home Language/Literacy/Learning Environment , Language and Literacy Development , Libraries and Public Resources , Parent-Child Relationships/Interactions , Play , Programs and Interventions (other) , Universidade do Bebê

metric(s)

Adult-Child Interactive Reading Inventory (ACIRI)
StimQ Cognitive Home Environment Questionnaire (StimQ)
Peabody Picture Vocabulary Test (PPVT)

objectives

This study investigated: (1) whether parent literacy is associated with parent–child reading aloud interactions, the overall cognitive home environment, and child language and cognitive outcomes; and (2) whether impacts of a toddler/preschool reading aloud program differ depending on parent literacy among low-income families in northern Brazil.

exposure

Universidade do Bebê (UBB).

outcomes evaluated

Shared reading, cognitive home environment, and child language/cognitive outcomes.

setting

Low-income neighborhoods in Boa Vista, a medium-sized city in northern Brazil with a high poverty rate.

methods

We performed a secondary analysis of a single-blind cluster randomized controlled trial (RCT) of a parenting intervention (Universidade do Bebê [UBB]) implemented in educational childcare centers in Boa Vista, Brazil. UBB consisted of: (1) a lending library in which children’s books were borrowed weekly; and (2) monthly parent workshops focused on reading aloud. Control families received usual care without the reading aloud program...Parents and children were evaluated at baseline/enrollment and approximately 6–9 months later. Measures: (1) Predictor/moderator: parent literacy; (2) Parent outcomes: parent–child interaction during shared storybook reading; cognitive home environment; (3) Child outcomes: expressive and receptive vocabulary; IQ; working memory and short-term phonological memory. Analyses accounted for baseline performance, sociodemographics, and clustering within centers and sites.

sample size

n=506 (dyads, total); n=279 (dyads, intervention); n=287 (dyads, control)

measures

Measure of Parent Literacy: The Cloze test, validated for Brazilian Portuguese, was used to evaluate parent reading comprehension at follow-up only.

 

Measure of Parent-Child Interactive Reading: The Adult/Child Interactive Reading Inventory (ACIRI) consisting of observation of parent-child interactions behaviors real-time coded during shared storybook reading.

 

Measure of Cognitive Home Environment: StimQ, a structured interview with the caregiver, was used to assess parent–child interactions in play, shared reading, and daily routines. Parent interview included three subscales:

• 1. READ subscale assessed frequency and quality of reading interactions.
  2. Parental Verbal Responsivity (PVR) assessed caregiver-child verbal interactions.
  3. Parent Involvement in Developmental Advance (PIDA) assessed caregiver teaching and play activities.

Measure of Child Outcomes:

• • Child Expressive Vocabulary: Teste Infantil de Nomeação, a Brazilian adaptation of the Boston Naming Test (BNT).
  • Child Receptive Vocabulary: Teste de Vocabulário por Imagens Peabody (TVIP), a Brazilian adaptation of the Peabody Picture Vocabulary Test.
  • Child IQ: Two subscales, Categories and Situations, from the Brazilian version of the Snijders-Oomen nonverbal intelligence test (SON-R).
  • Child Working Memory: Teste Infantil de Memória (TIMT; now called TIME-R), a non-verbal working memory task developed by Morais and Macedo (2011)
  • Child Phonological Short-Term Memory: Teste de Repetição de Palavras e Pseudopalavras (TRPP) is a Brazilian validated phonological short-term memory task.

Measure of Sociodemographic and Depressive Covariates: parent interviews including child variables, family characteristics, and parental depressive symptoms assessed by the Edinburgh Postnatal Depression Scale.

results

Five hundred and six mother-infant dyads (279 intervention; 287 control) were enrolled at mean (SD) child age 37.4 (6.5) months...Parent literacy was positively associated with parent–child verbal/reading interactions and child developmental outcomes, supporting our first hypothesis, that low parent literacy would be associated with lower quantity and quality of parent–child reading and lower scores on assessments of child development. UBB had positive impacts on parent–child interaction, cognitive stimulation, IQ, and receptive vocabulary for both high and low literacy parents, contrary to our second hypothesis, that low parent literacy would be associated with reductions in impacts of the reading aloud intervention.

conclusions

The comparable impacts of a reading aloud program on parent–child verbal/reading interactions and child cognitive outcomes regardless of parent literacy level suggest that preventive interventions in LMICs should consider promotion of shared reading even for low-literacy families.

limitations

A limitation of this study is that it was performed in a single city, Boa Vista, in one country, Brazil. Additional study is needed to determine whether findings are generalizable to other LMICs or HICs. A second limitation is the availability of instruments to evaluate young children and parent–child interactions during book reading in Brazil, although we observed impacts on key measures developed (TIME-R) or validated (Boston Naming test, Peabody, and SON-R) in Brazil. Given that the Cloze test is not a standardized measure, we performed sensitivity analyses with different cut points and found similar results. Further, the Cloze test provides a general estimate of reading comprehension, but it does not assess specific components such as word reading. Therefore, we do not know whether similar results would be obtained among parents with more specific limitations with word reading. A third limitation is that randomization took place only for families already attending childcare centers, and it is possible that childcare center attendance could have amplified impacts. Such impacts might be especially important for children whose parents have lower literacy levels, given that educational childcare can provide additional opportunities for children to experience book sharing with adults. Although the current analyses do not address the potential role of the childcare centers in supporting child outcomes, they suggest that interventions that support parent–child book sharing can enhance child development for children already attending educational childcare regardless of parental literacy level. Finally, a limitation of this analysis is that it did not address whether there were sustained impacts beyond completion of one year of the program. Parental literacy level might moderate the intervention's impacts in the longer term, even if no effect was found on immediate impacts. While our findings of early impact are nonetheless relevant, ongoing support for parenting and high-quality education may be necessary to sustain impacts beyond program completion and across children’s schooling (Pages, Lukes, Bailey, & Duncan, 2019). Furthermore, a higher level of support may be needed to sustain impacts among children whose parents have lower literacy levels.