Cultural Differences in Narratives and Reading Achievement in African American Children

Cultural Differences in Narratives and Reading Achievement in African American Children Stories or narratives children produce reflect their knowledge of an event converted into a decontextualized communicative form. Children create these narratives by combining cognitive and linguistic skills (Hudson & Shapiro, 1991). First explored as an analytical tool by Labov in 1972, narratives have also been defined as cultural tools that provide information about past event structured into a culturally accepted format (Ochs, 1997). Traditionally, a well formed narrative includes an abstract, orientation, complicating action, and coda (Labov, 1972). It is generally acknowledged that narrative competence is essential for children’s language development and a predictor of their literacy accomplishment (Dickinson & Tabors, 2001; Griffin, Hemphill, Camp & Wolf, 2004). As a consequence, the development of children’s narrative skills and its relationship with emerging literacy skills has garnered substantial attention from educators and researchers. One reason why researchers believe that children’s narrative skills predict their literacy development is that both skills require decontextualization (Snow, 1983). For instance, while telling a story, children must transition from contextualized language use (e.g., where the communicative partner is familiar with the information being conveyed), to decontextualized language use (e.g., where the communicative partner is often unfamiliar with the information being conveyed). This transition requires children to use decontextualized way of communication to provide information about the story that is unshared between the teller and listener of the story (Curenton & Justice, 2004; Pellegrini, 1985). Since decontextualized language requires children to engage in higher level thinking in order to express abstract ideas and events in narrative format, it is seen by researchers as being more complex than contextualized language (Gardner-Neblett, Pugnello, & Iruka, 2012). It is for this reason that researchers believe that children who frequently use decontextualized language in their oral narratives will also demonstrate advantages in other areas of literacy, such as reading comprehension. Decontextualized communication requires the teller to use certain linguistic features, known as literate language features (hereafter LLFs). LLFs reduce ambiguity and increase explicitness of the information being conveyed (Curenton & Justice, 2004). These LLFs include the use of conjunctions, adverbs, elaborated noun phrases (ENPs), and mental and linguistic verbs. All of these linguistic devices can be used to enrich the message of the narrative. For instance, children can use ENPs to convey details about the people and objects in the narrative (e.g. “it was a big tree”) or provide mental verbs to communicate the character’s intent in the story (e.g. “he thought he had done it right”). Since the use of LLFs necessitates the understanding that the listener of the story will benefit from these explicit details, researchers have assumed that if children’s narratives includes these features, then they are likely using sophisticated thinking and language skills in their storytelling (Pellegrini, 1985). Although the development of narrative skills in Caucasian children has been studied extensively (Hudson & Nelson, 1983; Peterson & McCabe, 1994; Hudson & Shapiro, 1991), much less is known about the narrative development of Hispanic and African American children (Burns, de Villiers, Pearson, & Champion, 2012). In addition to the lack of research, of the studies that have looked at studied these populations, many of them suffer from methodological limitations (Cain, Eaton, Baker-Ward, & Yen, 2005). As a result, many of these studies yielded findings indicating that African American (hereafter, AA) children possessed less developed narrative skills relative to their Caucasian peers (Burns et al, 2012; Cain et al, 2005). More recent research has shown that these previous studies failed to capture both the range and sophistication of AA children’s narrative skills (Curenton & Justice, 2004; Price, Roberts, & Jackson, 2006), reflective of the rich oral narrative tradition within the AA community (Gardner-Neblett et al., 2012). In addition, we know very little about the development of LLFs in AA children’s narratives (Curenton & Justice, 2004). Past research has shown that children’s narrative skills are influenced by home environmental factors, like exposure to book reading and parent-child narrative interactions as well as school activities, like ‘show and tell’ (Uchikoshi, 2005). Considering that AA and Caucasian families vary in number of different way in terms of early socialization of literacy related practices (Heath, 1994), it intuitive to assume that AA and Caucasian children would also differ in their use and development of LLFs. Given the cultural value placed on narrative skill in the AA community, it is important to consider the possibility that AA children’s oral language skills are more accurately reflected in their oral narratives than their performances on standardized language measures (Curenton & Justice, 2004). For example, past research has shown that the achievement gap between AA and Caucasian children, favoring Caucasian children, evident in development the preschool years persists and widens as children continue to progress through school (Apel & Thomas-Tate, 2009; Burchinal et al., 2011; Champion, Rosa-Lugo, Rivers, & McCabe, 2010). This trend is particularly evident on standardized tests of vocabulary (Craig, Connor, & Washington, 2003). The most significant impact of this disparity in vocabulary is on the literacy outcomes of AA children. Researchers have posited a wide variety of theories to account for this persistent disparity between the oral language skills of AA and Caucasian children. The most commonly posited reason for AA children’s difficulties and underachievement is the lack of economic and social resources (e.g. Burchinal et al., 2011; Fargas & Beron, 2004). While there is no question that the impact of having less advanced vocabulary and literacy development is most problematic for AA children from low-income homes, research has shown that the gap is also evident in AA children from middle-income homes relative to their Caucasian peers (Ogbu & Davis, 2003). Thus, the relationship between culture/race and income must also be carefully considered when attempting to determine the root causes of this oral language achievement gap. In addition to considering access to resources, it is also imperative for researchers to consider how children’s oral language skills are assessed. For example, most of the research investigating the achievement gap has relied on standardized oral language measures. However, more recently researchers have started to advocate for the using narratives as a more culturally sensitive and accurate way to assess AA children’s oral language development (Cain et al, 2005; Curenton & Justice, 2004). Based on these recommendations and previous findings, the goal of the current study was to examine the relationship between oral language skills, measure through narratives, and reading development in AA and Caucasian children. In particular, we were interested in determining whether differences were evident in the use of decontentxtualized language, specifically LLFs, in AA and Caucasian children from the same classroom. The study also explored relations amongst children’s use of LLFs and their scores on a standardized vocabulary measure and reading assessment. Based on previous research documenting cultural differences in oral narrative traditions, we predicted that differences would be evident between AA and Caucasian children’s use of LLFs in their narratives. Method Participants As part of a larger longitudinal study, 29 children (11 males, 18 females) were recruited for this study. AA (n = 19) and Caucasian (n = 10) 3rd and 4th graders (Mage = 9.04, SDage = 0.69) from Southeastern elementary school were tested. This particular school services a diverse racial economic community with approximately 65% of students qualifying for free or reduced lunch.) Procedure and Measures The children were tested individually, in school, by a trained research assistant. The following measures were used to assess: Receptive vocabulary. The PPVT-IV is a standardized test of receptive vocabulary in English (Dunn & Dunn, 2007) and was administered using the standard procedure. The children were told, “I will show you some pictures and say a word each time I show you the pictures. Can you show me the picture of the word I am saying?” If the child hesitated in responding the experimenter repeated, “Can you show me the picture for (word)?” Thus the children were required to select from the pictures the one that matched the word spoken by the tester. Narrative skill. To elicit narratives, a three-picture narrative measure was adapted from Kalia (2009). Children were shown three pictures that conformed to a story structure. These pictures were taken from Mercer Meyer’s wordless picture books, Frog Stories. Children’s narratives were audio-recorded and transcribed. They were coded for the frequency of LLFs (adapted from Curenton & Justice, 2004) after two coders independently achieved reliability on 25% of the narratives. Reliability achieved ranged from 93% to 100% agreement between the two coders. The rest of the narratives produced by the children were coded by one coder. See Table 1 for details. Reading comprehension. The North Carolina EOG Reading Comprehension Test was used to assess knowledge and reading skill sets students acquired by the end of the school year. Teachers and school administrators administer the test over a 3-day period. Children’s scores can range from 330 to 354, with varying achievement levels (e.g., Level I – Level IV). Results Descriptive statistics for the participants’ receptive vocabulary, narratives skills (i.e., LLFs), and reading scores are available in Table 1. Our prediction was that AA and EA children would differ in their use of LLFs in their narratives. In order to test our prediction that AA and Caucasian children would differ in their use of LLFs, we conducted a series of one-way between subjects ANOVAs. The results indicated that AA and Caucasian children’s narratives differed in the use of conjunctions F(1, 28) = 4.38, p < .05, with AA children using significantly more conjunctions (e.g., “And the frog, turtle, and the boy riding on the boat” ) than their Caucasian classmates . Also, although not statistically significant, a marginal difference between the groups emerged in their use of Simple Elaborated Noun Phrases (SENPs) F(1, 28) = 3.07, p = .09, with AA children using more SENPs than Caucasian children. No other significant differences emerged, all ps > .10. See Table 1. We also examined whether the two groups of children differed in their receptive vocabulary and literacy achievement using one-way between subjects ANOVAs. The results indicated that the AA and EA children did not differ on their PPVT and EOG reading achievement scores, all ps > .10. Table 1 Means and standard deviations for vocabulary, narrative, and literacy achievement for African American and Caucasian children Measure Mean (SD) Mean (SD) African American Caucasian (n = 19) (n = 10) PPVT 100.63 (16.99) 106.67 (10.63) EOGs: Reading 337.64 (23.67) 348.70 (6.46) 7.32 (2.45)  5.20 (4.07)  Select Examples Scores SENPs: Simple Consists of a single modifier and a elaborated noun noun. phrases “There’s a boy.” CENPs: Complex 1.47 (1.12) 2.40 (2.01) Consist of two or more modifiers and a elaborated noun noun. phrases “He got pushed down in the water.” Adverbs 2.63 (1.16) 3.10 (1.96) Modifiers to increase the explicitness of action and event descriptions. “He was mad at something.” Linguistic verbs .11 (.32) .10 (.31) Verbs referring to acts of speaking. “He instructed the dog to catch a stick.” Mental verbs .31 (.74) .10 (.42) Verbs referring to acts of thinking. “He decided to chase the frog.” Conjunctions 6.52 (3.22)* 4.10 (2.37)* Words used to organize information and clarify relationships. “And the frog and the turtle are watching” *p < .05; p = .09 The relationships between LLFs, PPVT scores, and reading achievement in AA children were examined using bivariate correlations (see Table 2). Thee analyses indicate that AA children’s receptive vocabulary was related to their EOG reading scores and their use of complex elaborated noun phrases (CENPs). For AA children, use of CENPs was associated with use of adverbs. Moreover, AA children who had larger vocabularies were more likely to use more complex language in their narratives, as evidenced by the significant correlation between children’s PPVT scores and their use of CENPs. Finally, AA children’s use of linguistic verbs (e.g., “He called his mom”) in their narratives was negatively associated with SENPs. More specifically, children who used more SENPs (i.e., simpler sentences) were less likely to provide explicit details about characters’ speech in their stories. Table 2 Correlations between African American children’s vocabulary, literate language features, and literacy achievement PPVT EOG SENPs CENPs Adverbs LV MV Conjunctions PPVT EOG SENPs - .48* - -.17 -.34 - .55* .41 -.26 .31 .32 .24 .10 .16 -.48* .15 .15 -.06 -.13 -.20 .39 CENPs - - - - .52* .16 .21 -.10 Adverbs - - - - - -.04 .07 -.15 Linguistic .32 -.00 verbs (LV) Mental .16 verbs (MV) *p < .05; SENPs = Simple elaborated noun phrases; CENPs = Complex elaborated noun phrases Discussion The present study was an exploratory examination of the use of literate language features (LLFs) in the narratives of AA and Caucasian children. In consideration of past research demonstrating that AA children experience a consistent disadvantage on standardized vocabulary tests relative to their Caucasian peers, the current study sought to determine whether assessing narrative skills was a more culturally-sensitive and accurate way to assess AA children’s oral language development (Curenton & Justice, 2004). We also examined the associations between AA children’s use of LLFs and their performance on a standardized vocabulary test (i.e. PPVT) and their end of grade standardized reading scores. This investigation of the differences in the use of LLFs by AA and Caucasian children demonstrated that the AA children used significantly more conjunctions and marginally more simple elaborated noun phrases (SENPs) in their narratives (see Table 1). Since it well known that parent-child interaction influences the development of narrative skills (e.g. Dickinson & Tabors, 2001) and previous research has shown that AA and Caucasian families have distinct practices for socializing their children in literacy related activities (e.g. Heath, 1994), we had predicted that the two groups of children would differ in their use of LLFs in narratives. Our findings supported our hypothesis and bolster the proposal that AA and Caucasian children produce different kinds of narratives due to socio-cultural influences (Bliss, Covington, & McCabe, 1999). However, our findings are also in contrast to the work of Curenton and Justice (2004), who found no differences in the use of LLFs in AA and Caucasian preschoolers’ narratives. It is possible that our findings differ because the children in our study were older and had already had several years of formalized education. It is also important to note that our sample emerged from one classroom and is, therefore, more homogenous. In contrast, Curention and Justice drew their sample from various classroom contexts ranging from Headstart program to private preschools. In addition to between group comparisons, we also examined the interrelations between LLFs, receptive vocabulary, and reading achievement within the AA group. The findings revealed, as expected, that AA children’s performance on the standardized test of vocabulary (i.e. PPVT) was positively associated with their performance on the standardized assessment of their reading. Furthermore, AA children’s use of CENPs (e.g. “he was a green, happy frog”) was also positively related to their performance on the PPVT. This finding provides complementary support for other findings demonstrating that oral narrative skills are associated with their vocabulary development in Caucasian children (e.g. Dickinson & Tabors, 2001). In addition, the current results also extend the literature by illustrating this relationship in school-aged AA 3rd and 4th grade children. However, in contrast to previous results with Caucasian children, we did not find an association between AA children’s narrative skills and literacy development (e.g. Griffin et. al, 2004). This finding is consistent with previous research demonstrating that the relationship between AA children’s oral language and their literacy development is complex and that AA children’s oral language skills do not always predict their literacy achievement (Connor & Craig, 2006). Interestingly, we did not find any associations between the LLFs that AA children used more in their narratives (i.e. conjunctions and SENPs), than Caucasian children, and any of the standardized measures. This provides support for the contention that the ways in which AA children use their oral language skills do not always map onto standardized measures of children’s oral language or literacy development (Connor & Craig, 2006; Craig et. al, 2003). Although the current findings suggest that the relationships between receptive vocabulary, narrative skills, and literacy development in AA children may be more complex than in Caucasian children, the small sample size warrants caution when generalizing these findings. As such, further examination of these interrelations would add to our understanding of narrative development in AA children. Recent research has also provided compelling evidence to suggest that the racial and socioeconomic disparities in the acquisition of oral vocabulary knowledge account for a significant portion of the gap in reading achievement between AA and Caucasian children (Champion et al., 2010). However, since we still know very little about the development of oral language skills in AA children, despite economic status (Craig et al., 2003), it is hard to pinpoint the specific mechanisms driving this striking disparity in early oral vocabulary knowledge let alone know how to fix it. The current study has attempted to address the gap in the literature by examining the relationship between AA children’s use of LLFs in their narratives and their reading achievement. However, much more work still need to be done. Many more crosssectional and longitudinal examinations are needed to better document the differing patterns of associations between AA and Caucasian school-aged children’s narrative and literacy skills in order to determine the nature and origin of the factors drive these differences. Hopefully, this research will allow us to someday soon not only narrow, but completely close the persistent achievement that influences the educational outcomes and life chances for AA children.