Standards of academic performance place a high demand on students’ English language. To help Spanish-speaking preschoolers who are developing English as a second language meet these demands, researchers recommend strengthening their first language to facilitate development of their second language. Head Start teachers and research assistants delivered 12 Spanish and 12 English language lessons to eight preschoolers in small groups. Lessons targeted storytelling and vocabulary and occurred 4 days a week for 20 min. A multiple-baseline experimental design across groups was used to examine the effect of the Spanish–English narrative intervention on children’s retelling skills and a pretest posttest design without a control group documented children’s acquisition of the target words. Results indicated that children made gains in English retelling while maintaining their already high Spanish retelling skills. Improvements in vocabulary were observed in English but not in Spanish.
narratives, vocabulary, dual language, storytelling, bilingual development, language intervention
Hispanic/Latino children represent 38% of the enrollment in Head Start preschool programs across the United States, and about 85% of those children are from Spanish-speaking families (Office of Head Start, 2016). Dual language learners (DLLs), in this case Spanish-speaking preschoolers learning English as a second language, are at high risk of later academic and reading difficulties in English (National Center for Education Statistics, 2015). Over 80% of fourthgrade Hispanic children and 92% of fourth-grade English language learners read below a proficient level (National Center for Education Statistics, 2015). This group of students is disproportionately placed into special education in the United States (Denton, West, & Walston, 2003). Although the majority of them do not have language-related disorders, their reading performance in English is not meeting expectations.
For children’s whose first language is Spanish, there is growing evidence to suggest dual language instructional approaches can lead to greater academic achievement and proficiency in their second language (August & Shanahan, 2006; Rolstad, Mahoney, & Glass, 2005; Slavin & Cheung, 2005). Positive outcomes result from comprehensive early childhood bilingual education programs (Barnett, Yarosz, Thomas, Jung, & Blanco, 2007; Durán, Roseth, Hoffman, & Robertshaw, 2013) and dual language supplemental interventions for children with disabilities (Restrepo, Morgan, & Thompson, 2013) and children with risk factors (Lugo-Neris, Jackson, & Goldstein, 2010; Méndez, Crais, Castro, & Kainz, 2015). A number of researchers suggest that high-quality early childhood practices for DLLs, whether comprehensive or supplemental, should incorporate the children’s home language, explicitly teach meaningful vocabulary words, and foster formal or academic language necessary to succeed in school (Castro, Espinosa, & Paez, 2011; García & Miller, 2008).
Research has clearly indicated that oral language in early childhood is significantly related to later reading and academic success (National Early Literacy Panel, 2008; National Reading Panel, 2000). In particular, reading comprehension relies heavily on oral vocabulary (Cain & Oakhill, 2011; National Reading Panel, 2000; Perfetti & Hart, 2001; Verhoeven & Van Leeuwe, 2008) and narrative ability (e.g., Griffin, Hemphill, Camp, & Wolf, 2004; Mehta, Foorman, Branum-Martin, & Taylor, 2005), which are suggested targets for DLLs (Castro et al., 2011). Although the majority of research linking vocabulary and narrative language to reading comprehension is correlational, Clark, Snowling, Truelove, and Hulme (2010) found a causal relation between their vocabulary and narrative oral language intervention and reading comprehension of third graders. Earlier interventions could prevent reading comprehension problems, and language interventions that incorporate preschoolers’ home language could have a powerful effect on later reading achievement of DLLs. Therefore, the purpose of this study was to examine the effect of a Spanish–English narrative intervention with embedded vocabulary instruction on the vocabulary and narrative skills of DLLs.
Evidence suggests that storybook-based vocabulary instruction in both English and Spanish yields vocabulary growth in both languages, or at the very least helps maintain the vocabulary already acquired. Restrepo et al. (2013) found that a dual language vocabulary intervention facilitated English receptive and expressive vocabulary acquisition to the same extent as the English-only intervention, and that Spanish vocabulary was also supported. Similarly, Méndez et al. (2015) found superior effects for the use of preschoolers’ home language (Spanish) and English over the use of the participants’ second language (English) only.
Oral, narrative-based language interventions have recently emerged in the research literature as an effective means of addressing academic language skills of diverse learners (Brown, Garzarek, & Donegan, 2014; Gillam & Gillam, 2016; Spencer, Petersen, Slocum, & Allen, 2015; Weddle, Spencer, Kajian, & Petersen, 2016), but there are no studies investigating the effects of a dual language narrative intervention. Likewise, there are no narrative intervention studies that targeted the acquisition of specific vocabulary words. Most early educators are familiar with storybook interventions which are a commonly used method of promoting vocabulary and comprehension (Roberts, 2008; Van Kleeck, 2008). However, narrative interventions may have some powerful advantages over storybook reading. First, in narrative interventions the stories can be engineered to be the exact structure, length, and complexity needed to foster academic language of children, regardless of their age and language development. Target vocabulary words and contextual support can be intentionally embedded in the narratives used during intervention (Lee, Roberts, & Coffey, 2017). Second, some storybook reading interventions encourage children to retell parts or entire stories, but it typically requires several repeated readings before young children are able to expressively retell the story. In narrative interventions, children have several opportunities to retell or tell stories and thereby receive many opportunities to practice using complex academic language related to the stories (Petersen, 2011).
There is good reason to believe that narrative intervention could be a viable dual language approach for promoting kindergarten readiness of young DLLs. Narratives are replete with complex academic language such as adverbs, adjectives, conjunctions, causal and temporal ties, and subordinate and relative clauses. Most narratives have a basic underlying structure referred to as a cognitive schema which is observed in the quality and number of story grammar elements (e.g., setting, initiating event, attempt, consequence) included in a narrative (Stein & Glenn, 1979). There are many similarities among story grammar elements across languages, and there is evidence suggesting children can transfer story elements and complex syntax across languages (Pearson, 2002). A dual language intervention can take advantage of the interrelatedness of narrative structure (or schema) and other shared features of Spanish and English (Fiestas & Peña, 2004) languages to hasten the acquisition of academically related oral language (Cummins, 1984; MacSwan & Rolstad, 2005; Petersen, Thompsen, Guiberson, & Spencer, 2015). If stories used in narrative intervention are carefully constructed, they may be able to facilitate the acquisition of vocabulary in addition to promoting narrative structure and complex syntax (Lee et al., 2017).
The findings from several research studies examining the efficacy and feasibility of an English-only narrative intervention program (Story Champs; Spencer & Petersen, 2012b) informed the design of a dual language version of the narrative intervention. In a number of group and singlecase experimental design studies, Story Champs has improved preschool children’s narrative retell skills, as well as their ability to generate personal stories and answer questions about stories (Spencer, Kajian, et al., 2013; Spencer, Petersen, & Adams, 2015; Spencer, Petersen, Slocum, Allen, 2015; Spencer & Slocum, 2010; Weddle et al., 2016). Many of the preschool participants were Spanish-speaking English learners attending Head Start preschools. While Story Champs has a positive track record with Spanishspeaking English learners, Spanish has not been deployed during Story Champs intervention, and vocabulary instruction has not been systematically embedded in the narratives used for intervention. Based on the evidence that strengthening children’s first language can facilitate the development of their second language (Baker, 2000; Coltrane, 2003; Cummins, 2000; Gibbons, 2002) and the early childhood practices for DLLs should include explicit vocabulary instruction, we developed a dual language (Spanish/ English) narrative intervention with embedded vocabulary instruction. In the current study, we investigated the extent to which the dual language intervention improved preschool DLLs’ narrative retell skills and vocabulary acquisition. The main objective of the dual language intervention was to improve the children’s English language skills (i.e., narrative retells and vocabulary) while maintaining or improving their Spanish language skills. The following research questions were addressed:
Research Question 1: To what extent does a dual language narrative intervention improve preschoolers’ English narrative retell skills?
Research Question 2: To what extent does a dual language narrative intervention with embedded vocabulary instruction improve preschoolers’ knowledge of targeted Spanish and English words?
Teacher participants. The director of a Head Start program in the southwestern United States recommended three Head Start teachers serving primarily Spanish-speaking children to participate in this study. The three teachers accepted the researchers’ invitation to participate. According to a demographic survey completed by the teachers, all teachers taught in half-day classes, four days a week. In Class 1, the lead teacher had taught preschool for 26 years, had an associate’s degree, and spoke fluent conversational Spanish, but could not read Spanish well. In Class 2, the lead teacher had taught preschool for 17 years and spoke fluent conversational Spanish, but could not read Spanish well. At the time of this study, she was working on her bachelor’s degree. In Class 3, the lead teacher was a native Spanish speaker from Mexico, was biliterate in English and Spanish, had been teaching 4 years, and had completed a few college courses. All three teachers were observed using the Classroom Assessment Scoring System (CLASS; Pianta, La Paro, & Hamre, 2008) in the same time frame as this study. Their scores are displayed in Table 1.
Child participants. At the time that parent permission was obtained, parents completed a short demographic survey in their preferred language. Sixteen children from the three classes whose parents identified Spanish as one of the primary languages spoken at home were screened for inclusion in the study. Brief language samples were collected in English and in Spanish using the retell section of the preschool Narrative Language Measures: Listening (NLM: Listening) subtest of the CUBED assessment (Spencer & Petersen, 2012a). A score of eight on the English NLM: Listening retell was used as a cut score for inclusion of research participants because previous research indicated it was a developmental standard for English-speaking preschoolers (Spencer, Kajian, et al., 2013; Spencer, Petersen, & Adams, 2015). Only children who scored 8 or below on the English NLM: Listening retell section were selected for the study. Ten children met the inclusion criteria. One child was eliminated from the study because of frequent absences, and one was removed because he was too reticent to speak and we could not confirm that Spanish was his dominant language.
Of the eight research participants, all except one were identified as Hispanic/Latino; one child was identified as White. Spanish was the primary language in seven of the children’s homes, and Spanish and English were the primary languages in the other child’s home. All children came from low-income households. None of the children qualified for or received special education services. The Clinical Evaluation of Language Fundamentals–Preschool–2 Spanish (CELF-P-2; Wiig, Secord, & Semel, 2009) was used to further describe participants’ Spanish language abilities prior to the study. In addition, Spanish language samples were gathered using the wordless picture book Frog Where Are You? by Mercer Mayer (1969). To collect the language samples, an examiner read a short story that went with the illustrations about a boy who loses his pet frog, and children retell the story. Children’s stories were recorded, transcribed, and analyzed using the Systematic Analysis of Language Transcripts (SALT) software (Miller & Iglesias, 2008). Total Number of Words (TNW) is a measure of language productivity, Mean Length of Utterance (MLU) and the subordination index are estimates of complex language, and Number of Different Words (NDW) is an estimate of a child’s breadth of vocabulary. More information about the CELF-P-2 and the language samples is available by emailing the first author. See Table 2 for additional information; children’s names have been replaced with pseudonyms.
It is unknown whether the Spanish lessons facilitated or accelerated children’s gains in English, but other evidence seems to support an additive effect (Méndez et al., 2015; Restrepo et al., 2013). It is possible that the native language component of intervention facilitated growth in English, but given the limited scope and exploratory nature of the current study, we were unable to examine that aspect of the intervention. Nonetheless, the advantage of dual language development is well documented in the literature (August & Shanahan, 2006; Rolstad et al., 2005; Slavin & Cheung, 2005). Without it, many families lose their ability to communicate, and children either do not continue to develop their native language skills for academic purposes or demonstrate protracted development (Castilla-Earls et al., 2016; Morgan, Restrepo, & Auza, 2013). In the context of Englishonly education, these effects are expected to continue and worsen beyond preschool. In the current study with a supplemental program, children’s Spanish narrative development seemed to be up to par to developmental expectations. However, without continued Spanish language promotion, they are at risk of losing proficiency in Spanish.
Despite a general experimental effect of the dual language narrative intervention on children’s English retelling skills, there were considerable individual differences. Some children showed quick responses and for others, the effect took longer to emerge. Evidenced by the magnitude and immediacy of change, five of the children showed strong effects while three participants showed slower and more moderate improvements. Eventually, all children retold English stories with NLM: Listening scores above an eight. This score represents a minimally complete story and is considered developmentally appropriate for preschool children. Four of the participants produced retells between scores of 15 and 18, which are exceptional for preschool children. A story earning 15 points includes a basic episode (problem, action, consequence, or ending), some enhanced components such as the setting and feelings, and usually a few language complexity elements such as because, when, and after. Young children who can reliably produce stories with this level of complexity are better prepared for academic instruction in elementary school. In fact, the higher scoring children in this study produced narratives with greater complexity than what is currently expected in kindergarten and first-grade curriculum standards (e.g., Common Core State Standards; National Governors Association Center for Best Practices & Council of Chief State School Officers, 2010).
Children’s response to intervention did not correspond to their performance on the Spanish norm-referenced language tests administered at the beginning of the study. Some children with lower Spanish language skills on the CELF-P-2 made substantial gains in English (e.g., Adan and Jaden) and some children with average Spanish language skills showed moderate gains related to the intervention (e.g., Iris and Sherry). Based on the current intervention and arrangement, we were unable to predict children’s English response to instruction based on their Spanish language skills. This limited relationship between vocabulary words during the children’ school day helped children learn the words in English. Teachers naturally used the target words in English, which provided additional practice opportunities. Because English was the primary language of instruction throughout the day, the children did not have the opportunity to listen to or use the Spanish vocabulary words to the same extent as the English words. These results contrast with those of Restrepo et al. (2013), in which gains in Spanish were documented. However, there were multiple methodological differences between the studies: Restrepo et al. (2013) targeted mostly nouns, there was explicit vocabulary instruction 4 days a week instead of 2 days a week, and they provided review lessons. It is possible that direct and indirect instruction in Spanish, as opposed to direct instruction only, may speed up the acquisition process. Marulis and Neuman (2010) reported that effective vocabulary instruction leveraged instructional time across the day, thereby increasing the number of exposures and opportunities for practice.
In classrooms where the majority of the children are English-only speakers, it may be challenging for teachers to encourage the use of Spanish vocabulary words throughout the day. Even the teacher of Class 3 whose first and most proficient language was Spanish spoke primarily English to the children throughout the school day. Therefore, the limited gains in Spanish could be the result of less time dedicated to Spanish instruction and general conversation in the classroom. Perhaps an alternative would be to promote Spanish skills at home by providing parents/caregivers with a list of words or materials to help them emphasize the words at home.
Although the dual language narrative intervention shows promise for improving oral language skills, we are mindful of several limitations. The first limitation is related to our inability to conduct a follow-up probe examining the extent to which the children maintained their retell improvements after a period of no instruction. We did not collect intervention fidelity data during baseline to help differentiate the conditions. Had we documented the extent to which teachers taught vocabulary and narratives during baseline, we could have more confidence in the intervention effect. The CLASS results offer some information about how well teachers modeled language in their classrooms. On a 1 to 7 scale, where 7 is high, the teachers scored relatively low (3, 2.7, and 4), suggesting that teachers did not demonstrate strong language modeling in their classroom.
Another limitation is related to our approach to measuring vocabulary acquisition. Typically, researchers select words consisting mostly of nouns and easily pictured verbs so that picture vocabulary assessments will be sensitive to intervention effects and developmentally appropriate for young children (Hoffman, Teale, & Paciga, 2013; LugoNeris et al., 2010; Restrepo et al., 2013). In our dual language intervention lessons, we targeted adjectives and verbs that are more abstract and challenging to picture, such as brave, rough, tremble, and dangerous. Because we did not restrict our vocabulary targets based on the ease of assessment, it is likely that our receptive picture vocabulary assessment did not adequately measure the children’s true vocabulary growth. The portrayal of dynamic vocabulary using static pictures could have also reduced children’s ability to appropriately respond to the questions and therefore, demonstrate their vocabulary knowledge.
There are many potentially fruitful avenues for future research stemming from the limitations and findings of this study. Considering this was an early phase study in the iterative development of a new curriculum, it lacked a handful of rigorous methodological features that can be improved in follow-up studies. For example, future research should examine the maintenance of effects following the withdrawal of the intervention, use a control group with randomization to examine vocabulary acquisition, and ensure at least five data points occur within each condition for single-case design studies. Research that examines the effect of English-only versus dual language interventions on English language outcomes would be a good next step in this line of research, especially given we were unable to isolate the effect of Spanish intervention component in the current study. It would be worthwhile to enhance or modify the intervention so that it is potent enough to improve Spanish vocabulary acquisition as well. In this study, parents reported that they completed many of the take-home activities with their children and they enjoyed the activities. This suggests boosting the dose and quality of Spanish language promotion via a stronger take-home component is worth exploring.
A number of improvements regarding measurement are also noteworthy. Given the limitations of the receptive vocabulary measures used in the current study, future research should include improved measures to ensure valid assessment of vocabulary growth. In general, vocabulary measurement in preschool is an area that warrants additional research because there is great value in teaching more challenging words than what can be depicted in an illustration and there are many weaknesses related to the receptive picture vocabulary methods commonly used (Hoffman et al., 2013). Likewise, it would be advantageous to have a distal measure of language to determine whether the intervention has a lasting and robust impact. For the current study, which was the first study of the dual language narrative intervention in an abbreviated format (8 weeks), it was not likely that distal outcomes would have been impacted. If subsequent studies feature a longer and more fully developed intervention, the inclusion of distal language measures is reasonable and necessary.
This study reflects an initial attempt to examine the effect of a Spanish–English narrative intervention with embedded vocabulary instruction on children’s Spanish narrative retelling skills and acquisition of English and Spanish vocabulary words. Using multiple-baseline design conventions, we established a causal relation between the dual language intervention and the English retell outcomes; however, only a portion of the children showed clear level and trend changes. Retell improvements were delayed for three children. Although only a few children learned some of the Spanish vocabulary words, all but two children learned many of the English vocabulary words. Findings suggest the dual language intervention has promise for promoting English language while maintaining children’s first language. Future research is needed to improve the rigor of the evidence and to enhance the potency of the intervention.
The authors express gratitude to Northern Arizona Council of Governments Head Start and Sally House, Lupita Soto, and Rosanne Gonzales who made valuable contributions to the study.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305A140093 for US$1,481,960. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.