The effect of otitis media and daycare quality on mother/child bookreading and language use at 48 months of age

Abstract

The purpose of this study was to understand the effect of quality of childcare and experience with otitis media (OM) early in life on mother/child bookreading and language use at 48 months of age. Forty-one children were followed from daycare entry in the first year of life until 4 years of age. Weekly health screenings from daycare entry until 4 years of age were performed by a nurse and physician to assess OM. Half the children were classified as chronic OM and half as nonchronic OM. Half the children were in low-quality care and half in high-quality care, but there were no SES confounds with quality of care. Children who had chronic OM and were in low-quality care spoke half as many words and different words during bookreading, answered fewer abstract questions correctly, and were rated more poorly on language use by their teachers than the other three groups of children.

Introduction

Although there are many factors in the lives of young children that have been linked to later child outcomes, some recent studies have examined the possible short-and long-term effects of early experience with otitis media (OM) on children's cognitive, social, and especially language skills development (Vernon-Feagans, 1999). This surge in studies on this disease has been precipitated by the fact that OM is now the most frequently diagnosed disease in early childhood, affecting almost all children under the age of 3 (Rosenfeld & Bluestone, 1999). Estimates suggest that from one-third to one-half of children suffer from chronic bouts of this disease in the first few years of life (Vernon-Feagans, 1999). In recent studies of daycare-attending children, up to one-half of the children suffered from chronic OM over the first 3 years of life Roberts et al., 1995, Vernon-Feagans et al., 1997, Zeisel et al., 1995. Children's hearing can be affected during the bouts of this disease with a mild to moderate conductive hearing loss that can last for many months. Research has focused on the possible effects of this disease because the intermittent hearing loss occurs during the critical early years of life when important skills in the areas of language, sociability, and cognition are being acquired.

The initial work on the effects of OM examined the direct effects of this disease on language and language-related skills both in early childhood and at school age, with the supposition that the hearing loss during episodes of OM was the real causal variable. This has been called the Global Language Model (Feagans & Proctor, 1994). Results from several recent studies suggest a relationship between bouts of OM in early life and performance on expressive language tests Friel-Patti & Finitzo, 1990, Rach et al., 1991, Teele et al., 1984. Chronic OM may also affect children's early expression of communicative intents. Yont, Snow, and Vernon-Feagans (2000) found that 12-month-old infants with chronic OM used fewer types of nonverbal strategies to help parents interpret their otherwise unintelligible vocalizations than children without OM. In contrast, there are also other studies that have shown no effects on language, even in the early years of life Gravel & Wallace, 1992, Wright et al., 1988. Confounds with SES and other sample characteristics, as well as the measurement of OM, may have contributed to these conflicting results. In addition, most of the studies were designed to examine the possible main effects of OM on children's development rather than examining moderators that could be important in understanding the possible effects of this disease.

Some recent studies of the effects of OM have used the methodological and theoretical work in the social sciences to develop a different model of the effects of this disease Rutter, 1987, Sameroff, 1983. This kind of model stresses that the accumulation of risks in early childhood may be more important than looking at one particular risk factor such as OM. In addition, this kind of cumulative risk/interaction model would also argue that although hearing loss is still the original causal variable in understanding the effects of OM, hearing loss may lead to secondary causal variables that lead to developmental problems. In this more interaction-oriented model, it is argued that children who suffer from chronic OM with its accompanying hearing loss may have learned, during these long bouts of OM, to “tune out” or ignore lengthy extended language discourse in favor of attention to more salient aspects of the environment. This strategy of ignoring language may become a way of engaging the world, leading to less attention to language in situations requiring sustained attention to comprehend information and less effective strategies in communicating with others. Children with persistent OM in early childhood have been found to have trouble sustaining attention to language in the first few years of life and at school age Feagans et al., 1987, Feagans et al., 1994, Roberts et al., 1995, Roberts et al., 1989. This lack of attention early in life may lead to problems comprehending extended discourse in conversation, storytelling, or topic elaborations that require sustained attention (Vernon-Feagans, 1999). This line of reasoning shifts the emphasis to examining the secondary causal variables like attention and also stresses examining developmental outcomes like bookreading and conversational skills. A second important aspect of this kind of interactional model not only stresses the concept of mediating variables in order to specify more clearly the possible mechanisms of causality but also stresses moderators or the conditions under which the developmental effects are produced (Baron & Kenny, 1986). In this kind of model, it would be postulated that children with persistent OM could be buffered against the negative effects in protective environments where there is greater one-to-one interaction with adults, little or no background noise, and where adults are sensitive to the needs of the child with reduced hearing. These environmental effects are the conditions (moderators) under which OM may have little effect on development. Conversely, risky environments, where the child has little one-to-one interaction with adults, persistent background noise, and caregivers who are not sensitive to the child's hearing loss, may actually exacerbate the negative effects of OM. Thus, these negative conditions (moderators) specify the conditions under which OM may have greater effects on development.

The most recent research findings on OM seem to support this kind of model and may help to explain some of the previous studies where null findings were found (most of those studies were interested in finding the main effects of OM while not considering possible moderating variables). For instance, Roberts et al., 1995, Roberts et al., 1998 found that in a low-income African American sample of children, chronic OM might alter the pattern of interaction between mother and child, such that less optimal interactions led to poorer language outcomes early in life. In another series of studies, it has been found that children with chronic OM in low-quality child care are most at risk for a host of developmental problems, including poorer attention to language in daycare during bookreading, poorer expressive language as measured by the Sequenced Inventory of Communication Development (SICD), and a greater likelihood of being more introverted in interaction with a friend later in childhood Creps & Vernon-Feagans, 1999, Feagans et al., 1994, Vernon-Feagans et al., 1997. On the other hand, children with chronic OM in high-quality care were found to be buffered against many of these negative effects. Thus, there may be important mediators and moderators in the home and childcare that are important to understand when examining the effects of OM.

In the present study, the quality of the daycare environment was examined as a possible moderator of the effects of OM on discourse/language skills. A number of studies have shown that children in high-quality care have better language and social skills than children in low-quality care McCartney, 1984, McCartney & Clarke-Stewart, 1999, McCartney et al., 1985, NICHD Early Child Care Research Network, 2000, Vandell & Powers, 1983, Vernon-Feagans et al., 1999. Children in high-quality childcare have been rated by their teachers as more social (Vandell & Powers, 1983) and verbally facile. Some of these social effects have lasted into the elementary school years (Vandell, Henderson, & Wilson, 1988). It has also been reported that children in low-quality care at 24 and 36 months of age have poorer vocabulary and syntax than children in high-quality care, even when SES is not a factor differentiating daycares Hurley & Vernon-Feagans, 1998, Vernon-Feagans et al., 1997.

One important skill that children acquire over the preschool years that might be sensitive to variation in quality of care and OM is children's bookreading skill with an adult. This skill generally requires a number of language skills that are dependent on the children's ability to sustain attention to language over a rather long period and the ability to engage in language interaction with an adult. There is a large body of research that links young children's experience and skill with bookreading to vocabulary, preliteracy, and reading skills Dickinson & Smith, 1994, Heath, 1983, Pellegrini et al., 1995, Valdez-Menchaca & Whitehurst, 1992, Wells, 1985, Whitehurst et al., 1988. These studies have reported that joint bookreading between mother/caregiver and child contain aspects of both oral and written language that may be helpful to children during the transition to school and to the kind of decontextualized language used in books Snow & Ninio, 1986, Snow et al., 1982, Snow et al., 1991, Teale, 1984. The kinds of devices that parents/caregivers use in bookreading that have been hypothesized to facilitate development are the use of scaffolding strategies that build on the shared knowledge of the child and caregiver to extend the child's knowledge, the indirect teaching of vocabulary through conversational strategies, and the use of the back-and-forth question/answer format that often extends the meaning of the story and is so characteristic of teaching strategies in elementary school.

The frequency of adult bookreading during the preschool years in both middle-class and low-income families was found to be significantly related to a number of reading achievement and literacy-related activities in a recent meta-analysis of early bookreading studies (Bus, van IJzendoorn, & Pellegrini, 1995). They examined 16 studies on bookreading and language growth, 16 studies on bookreading and emergent literacy, and 9 studies on bookreading and reading achievement. The meta-analysis of these three relationships resulted in medium to large effect sizes ranging from d=0.55 to d=0.67. Although these authors conclude that there is a strong relationship between the frequency of preschool bookreading and later language/literacy, they realize that there are still questions remaining about the conditions under which bookreading is important for children. It was also the case that this meta-analysis did not address the quality of parent/child bookreading or the individual differences among children in relationship to bookreading skills.

Some recent studies have tried to examine the quality of strategies used by adults during bookreading as well as individual differences in children's response to bookreading as possible predictors of later language and literacy. Dickinson and Smith (1994) described a number of different bookreading styles used by preschool teachers and found that a “performance-oriented” style was related to receptive vocabulary growth. This style represented reading the whole story uninterrupted, accompanied by focused discussion of the story by the teacher before or after the story was read. Other studies have focused on the advantage of using a “dialogue” bookreading style, which emphasizes the back-and-forth question asking/answering style characteristic of early schooling practices. For instance, a series of studies Whitehurst et al., 1994, Whitehurst et al., 1994, Whitehurst et al., 1988 found that a “dialogue” style that includes the adult using open-ended questions appeared to increase children's vocabulary and literacy skills. In a follow-up to these earlier studies, 4-year-old children in New Zealand (Reese & Cox, 1999) were placed in one of three different reading styles for 6 weeks. Results suggested some complex interactions among child characteristics and reading style but generally supported more positive vocabulary growth for children who used a “dialogue” approach to bookreading that resulted in a more interactive style during bookreading using a question/answer format.

Senechal (1997) found that, for 3- and 4-year-old children, listening to a book a number of times helped children acquire vocabulary, especially receptive vocabulary. On the other hand, questioning the children during bookreading with “who” and “what” questions that were open-ended had a powerful effect on expressive language acquisition. Feagans and Fendt (1991) and Vernon-Feagans (1996) found that low-income children, who had attended a daycare intervention that emphasized language use, answered more abstract questions correctly after a storytelling task in kindergarten in both a paraphrase task and in a bookreading task with their teacher. They also found that certain wrong answers children gave about the story were related to teachers' ratings of language use and to reading achievement. “Irrelevant answers” (e.g., answering a “why” or “when” question with a location word) made by children in answer to abstract questions about the story were more difficult for teachers to respond to effectively and these answers were also negatively related to a host of school performance measures, including reading achievement.

The present study explores whether daycare quality is a moderator of the effects of OM on a mother/child bookreading task in which the mother asks easy and difficult questions of her child during a bookreading task in the home. Data from this task provided language use data during bookreading conversations for both mother and child. Mother and teacher ratings of the child's language use in the home and daycare were also obtained.