formative assessments of reading or writing

Article

‘Dime De Que Se Trató/ Tell me what it was about’: Exploring emergent bilinguals’ linguistic resources in reading assessments with parent participation

Silvia Noguerón-Liu University of Colorado-Boulder, USA

Courtney Hokulaniokekai Shimek University of Georgia, USA

Chelsey Bahlmann Bollinger James Madison University, USA

Abstract

The purpose of this study was to explore the ways emergent bilingual first-graders draw

on multiple linguistic resources during reading assessments and the participation of their

Spanish-dominant parents in those assessments, as children engaged in English and

Spanish retelling tasks. Informed by a translanguaging lens, sociopsycholinguistic and

holistic approaches to reading and critical approaches to family literacy, the analysis

centres on assessment sessions with two mother–child dyads whose children attended

school in a relatively new migration setting. Primary data were drawn from four reading

assessment sessions and audio-recordings over a 7-month period with each child, indi-

vidual interviews and home visits with mothers, and field notes from research team

members. The analysis examined linguistic patterns related to second-language approxi-

mations and code-switching in miscues and oral retellings. The analysis also includes

coding of strategies and resources children used in their English and Spanish retelling of

the same text, using their home language to retell the texts to their mothers. Findings

illustrate that while children’s miscues may be shaped by their developing control of

syntactic structures and new vocabulary, they draw from multiple language resources in

Journal of Early Childhood Literacy

0(0) 1–23

! The Author(s) 2018

Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/1468798418770708

journals.sagepub.com/home/ecl

Corresponding author:

Silvia Noguerón-Liu, School of Education, 249 UCB, Boulder, CO 80309, USA.

Email: [email protected]

English retellings, conveying their complex understandings of texts. We also found that

the children negotiated translating and retelling for their parents in different ways,

shaped by their family literacy practices. These involved co-construction of stories, a

focus on accuracy and the paraphrasing and embellishing of stories and dialogue. Insights

from this study highlight the complexity of pooled language resources in young chil-

dren’s repertoires. Findings also document the situated nature of oral retelling at home,

when parents engage children in the sharing and translating of English books in ways that

align with existing roles, practices and goals. Implications for equitable literacy assess-

ment in new migration contexts are discussed.

Keywords

Bilingual children, assessment, Spanish, parental involvement, multilingualism, Latinx

children, home–school practices, family literacy, biliteracy

In this article, we discuss the ways first-grade emergent bilinguals draw on multiple linguistic resources during oral reading assessment tasks, and the ways their Spanish-dominant parents participate in the assessment process. Much of the US-based scholarship addressing bilingual assessment has drawn from research in contexts with large immigrant communities, as well as schools that incorporate home languages in some way (e.g. August et al., 2006; Freeman, 2001; Hopewell, 2011; Kabuto, 2016a). However, notions of linguistic and cultural diversity can vary when including factors such as the arrival of new immigrants to less traditional locations, mixed responses from local residents, or differentiated access to public resources (Vertovec, 2007). This super-diversity phenomenon has been identified in receiving nations like England, with the arrival of multiple-origin, spatially scattered immigrants since the mid-1990s (Vertovec, 2007) and in the United States, where Latinx1

immigrants have settled in new destinations such as the Midwest and Southern regions of the country (Hamann and Harklau, 2015). Our study took place in a small city in the US Southeast, where Mexican and Central American families have settled in the past 20 years. We focus on literacy assessment involving parents in an after-school programme, in a context where school-based assess- ment takes place mostly in English—the language children are still acquiring.

We situate the assessment of emergent bilinguals in larger equity conversa- tions. Jiménez (2004) argues that equitable literacy assessments should address their first languages and literacies, home and out-of-school practices and students’ perspectives. However, assessment tools vary in their depend- ability since emergent bilinguals have diverse proficiencies across languages

2 Journal of Early Childhood Literacy 0(0)

and modes (e.g. L2 literacy, L1 writing), the tests’ linguistic demands require certain proficiency levels, and test-raters may have different qualifications to score students’ responses (Solano-Flores, 2008). Even when reading is assessed in students’ first languages, high-stake cut-off scores may be based on assumptions about languages as isolated units (Hopewell and Escamilla, 2014). This study responds to these issues by adopting a holistic, translangua- ging lens to examine the reading performance of bilingual young children.

This study brings together critical perspectives to parental engagement and resource-based approaches to family literacy (Delgado-Gaitán, 2012; Hurtig and Dyrness, 2011) by including parents in the analysis of assessment prac- tices in elementary schools. We find these lenses to be relevant in contexts of new migration in the United States, where schools face challenges addressing rapidly changing demographics (Hamann and Harklau, 2015). Our analysis offers implications for collaboration between parents, teachers and research- ers. The following questions guide our enquiry:

. What patterns are there in students’ linguistic repertoires in their oral reading and retelling of texts in English?

. What strategies and resources do students use in their retelling of a text in their first and second languages?

. How does their mothers’ participation in Spanish retelling shape their performance?

Theoretical framework

We adopt a translanguaging lens to examine children’s performance in decoding and comprehension. We consider how the children draw from all their lin- guistic resources, even when the tasks are predominantly in their second language. We integrate this lens with sociopsycholinguistic approaches to reading as a process, building on previous research on the reading processes of young bilinguals. Lastly, we draw from critical and participatory approaches to family literacy in order to reposition parents’ knowledge and participation in assessment as a joint enquiry with researchers.

Translanguaging and assessment

The concept of translanguaging captures the language practices of speakers beyond the separation of two languages, conceptualizing them within one linguistic repertoire (Garcı́a and Wei, 2014). Garcı́a (2009) initially

Noguerón-Liu et al. 3

emphasized the multiplicity of resources and practices that individuals negoti- ate on a continuum in order to navigate their everyday contexts, defining translanguaging as ‘multiple discursive practices in which bilinguals engage in order to make sense of their bilingual worlds’ (p. 45, author’s emphasis). This concept moves from framing language as a system of structures to that of languaging and language practices, focusing on processes of interaction, meaning-making and the agency of speakers (Garcı́a and Wei, 2014). Families and communities are central in Garcı́a’s (2009) initial definition, considering how family mem- bers have different linguistic abilities, and translanguaging is an inclusive discursive practice when they all come together.

Garcı́a and Wei (2014) visualize assessment from a translanguaging stance as allowing students to showcase their knowledge in flexible ways, choosing preferred multilingual and/or multimodal practices that best fit their response. In our analysis, we adopt the translanguaging assessment framework proposed by Garcı́a et al. (2017). Principles include: (a) considering various angles and perspectives (students, teachers and families); (b) considering students’ performance independently and with moderate assistance, as well as using resources such as text features or dictionaries; (c) grounding assess- ment on authentic tasks; and (d) distinguishing whether students can use features across their full language repertoire (with freedom to select across languages) or language-specific features (e.g. Spanish oracy, English literacy). Even though this model is intended for classroom-based assessment, we adopted these guidelines to expand the assessment process using a benchmark tool.

Sociopsycholinguistic approaches to reading and emergent bilinguals

Our analysis is also grounded on sociopsycholinguistic and holistic views of reading and assessment (Goodman et al., 2005), since we analysed miscues and strategies in oral reading and retelling data. Reading is conceptualized as a meaning-making, strategic process, as children systematically use their semantic, syntactic, graphophonic and pragmatic cueing systems, bringing background, linguistic and strategic knowledge to the task of transacting with text (Goodman et al., 2005). Goodman et al. emphasize that the lin- guistic knowledge of bilingual readers must be considered in the analysis of semantic and syntactic cueing systems because their second language is still developing. They recommend asking readers to use their first language to retell, as they may have comprehended more than what they can articulate in English. We adopt Goodman’s model together with a translanguaging

4 Journal of Early Childhood Literacy 0(0)

lens to examine whether students are relying on general linguistic features (across all their language resources) or language-specific features (in English or Spanish) in oral reading and retelling.

We build on studies describing comprehension assessment with bilingual children. Younger readers (PK-2), are still developing automaticity and decod- ing skills, which may shape their comprehension of text (Stahl and Garcı́a, 2017); however, emergent bilinguals’ comprehension may also be hampered by gaps in background knowledge, fluency, vocabulary knowledge or less familiarity with mainstream discourse patterns (August et al., 2006). Another issue is a pervasive monolingual view in the design and cut-off points in high-stake tools. Hopewell and Escamilla (2014) analysed scores of 268 emergent bilingual third-graders on Developmental Reading Scales (DRA-2) and their Spanish equivalent and found that students’ performances were more likely to be categorized as ‘at risk’ when only their cut-off scores in English were considered. Yet, when authors used a bilingual holistic frame- work instead, the number of at-risk identifications was reduced. Hence, sev- eral factors complicate assessment with young bilinguals: their simultaneous development of second language and automaticity skills, possible gaps in background knowledge, vocabulary and discourse patterns, and a monolin- gual framework behind testing benchmarks.

This study is also informed by past research analysing oral reading and retelling tasks with bilingual readers. This work shows that when assessed in their two languages, they integrate cueing systems and comprehension strate- gies across them, improving their performances (Freeman, 2001; Hopewell, 2011; Kabuto, 2016a). Our study expands on this work by focusing on emer- gent bilinguals who do not receive explicit Spanish instruction at school, and who may be assessed primarily in their second language. Like Croce (2014), we examine patterns in children’s second language approximations in retellings, looking for patterns and evidence of meaning-making, while also considering how their culturally based knowledge shapes text interpretation.

Critical and participatory approaches to family literacy

Lastly, this study was informed by approaches that leverage non-dominant parents’ expertise from a critical perspective, centred on advocacy, cultural critique and dialogue towards more equitable home–school relationships (Delgado-Gaitán, 2012; Hurtig and Dyrness, 2011). Drawing on critical theory, family engagement requires a different methodological approach, as researchers and parents come together to engage in joint e7nquiry and

Noguerón-Liu et al. 5

action to examine school-based and community practices. Literacy practices are central in past work with immigrant families because of the academic knowledge demands to communicate with schools, and the differences between reading practices at home and in classrooms (Delgado-Gaitán, 2012). This study strives to bridge the power asymmetry bilingual families may face by including parents as experts in their children’s literacy practices, rather than positioning parents as learners who lack ‘best practice’.

We expand on studies that have focused primarily on home literacy prac- tices and beliefs (e.g. Baird et al., 2015; Gillanders and Jiménez, 2004; Reyes, 2006) which and documented parents’ strategies for raising biliterate children. Our work aims to facilitate opportunities for parents to participate in and examine school-based practices, especially considering assessment implications for their children’s academic standing. For instance, Kabuto’s work (2016b) involved parents of students receiving special education services in retrospective miscue analysis, showing the ways disability labels are constructed in relation to families’ histories. Likewise, Corella Morales and Lee (2015) explain how code- switching in oral assessments was similar to observed household codeswitching practices. Since exposure to Spanish is taking place primarily at home in our context, we argue that recruiting parents as partners in assessment can inform expansive views on reading and retelling.

Methodology

We adopt a critical approach to family involvement (Hurtig and Dyrness, 2011) where by researchers and participants engage in joint enquiry, co- constructing spaces for dialogue about equitable home–school partnerships. Silvia facilitated 24 family sessions using bilingual social studies’ texts during the 7-month programme ‘Nuestra historia y herencia’ (Our history and heritage), implemented at a neighbourhood library (2015–2016). To document reading strategy patterns, each child was individually assessed four times with their parents present.

Silvia was the project PI and facilitator of programme sessions. She is a bilingual/biliterate faculty member who shared a cultural/linguistic back- ground with families. Courtney and Chelsey conducted the reading assess- ments with focal students. They were former early childhood educators, of European-American descent, and doctoral students at the time of the research. Both are English-dominant, with some receptive skills and understanding of Spanish. Silvia had served as a translator in parent–teacher conferences in the

6 Journal of Early Childhood Literacy 0(0)

school that focal children attended. The inclusion of parents in assessment ses- sions was intended to open a dialogue about the reading process with a similar team of participants: English-dominant former teachers (Courtney and Chelsey), a bilingual cultural broker (Silvia) and an emergent bilingual reader.

Participants’ backgrounds

A total of nine families participated in the study. All parents reported being Spanish-dominant and arriving as adults to the United States. Most children attended the same elementary school (with a 68% Latinx enrolment) and received ESL services within the classroom through a push-in model, where they remained in their mainstream room and received content instruction alongside targeted second language instruction from an ESL teacher (State Department of Education ESL guidelines, 2013: 17). The school had a strong commitment to supporting Latinx students and families, encouraging the use of Spanish at home and making culturally relevant materials available. We focus on two mother–child dyads: Dionisia and her son Luis, and Teresa and her daughter Alicia (all names are pseudonyms), who lived in close proximity to each other. Their cases illustrate the development of readers in early stages of balancing decoding strategies with English acquisition in first grade; however, there were differences in their academic performances and experiences. We compare these cases to illustrate the particularities of prac- tices in each household, as well as the challenges each mother faced commu- nicating with schools. Silvia had known both families for two years prior to the study from their participation in other projects.

Alicia (age 6) had a soft voice and was initially shy with strangers but warmed up when reading with others. She started learning English at a neighbourhood tutoring programme at age 3. Teresa, her mother, had lived in the United States for almost 10 years, coming from a small city in Guerrero, Mexico in her mid-30s. As a child, Teresa worked to support her family, so she did not attend elementary school. Hence, she was not able to read in Spanish; she could recognize a few words but could not decode texts on her own. She relied on her spouse, librarians and relatives to decode forms and applications. Teresa explained she had honest conversations with Alicia (her only child) about the importance of school and reading, describing her own regrets about not going to school. She took Alicia to the library every day and enrolled her in after-school and summer programmes. She was known in the neighbour- hood to be an excellent cook and gardener.

Noguerón-Liu et al. 7

Luis (age 7) was curious about stories, video games and meeting new people, never shy to share opinions. He began Head Start at around age 3 but had difficulties in his speech development. Dionisia, his mother, explained that initially he received special education services in schools where he was one of a few Latinx students (most were African American), before his first year at the neighbourhood elementary school. He was retained in kindergar- ten, a decision Dionisia supported, claiming he could not speak well in ‘any language’. Silvia noticed Luis occasionally distorted some initial sounds or merged words together; he also rushed his speech. His behaviour was also a concern to his mother; he was frequently suspended for failing to follow directions and fighting with peers. Dionisia had lived in the United States for almost 16 years. She completed fifth grade in a small town in the coastal state of Veracruz, Mexico. Like Teresa, she was an involved library patron, and spent all her time caring for her son’s needs and multiple appointment referrals. She claimed she had to advocate for her child and ‘defend his rights’ at every institution, as she felt he was not always treated fairly.

Data collection

Each child was individually assessed four times at the library using the Fountas and Pinnell Benchmark Assessment System (BAS) (Fountas and Pinnell, 2011). This test was chosen because it was similar to the reading benchmark used in the children’s school. The child reads a levelled text orally, while the administrator records the child’s miscues, calculates the rate of words read correctly and uses a rubric to score fluency. This is followed by a comprehension conversation, with unassisted retelling (e.g. ‘What was this story about?’), and questions on key details, inferential thinking and text features. Students were assessed with fiction and non-fiction texts.

Using first languages in retelling is permitted in the BAS guidelines, but there are no specific directions. Hence, in addition to the standard procedures, we incorporated an invitation to retell in Spanish to parents in the third and fourth sessions. To ensure inter-rater reliability, Courtney and Chelsey con- ducted each assessment together, annotating miscues as the child read aloud. To include parents in the process, Silvia sat by each mother and explained what was happening, translating as needed. After the assessment, parents, children and all authors engaged in a debriefing conversation. Data collected include assessment forms and audio-recording of each session, four individual interviews and home visits with each mother, and short reflective field notes from each author, post-assessment.

8 Journal of Early Childhood Literacy 0(0)

Data analysis

Our analytic process began by creating a bilingual reader’s profile for each child. In both Goodman et al. (2005) and Garcı́a et al. (2017), student profiles are thorough explorations of children’s language and literacy skills to identify their strengths and areas of growth for instructional purposes. Reader’s profiles combine statistical information from miscue analysis, pat- terns in reading strategies, and retellings from a single text (Goodman et al., 2005). Garcı́a et al. (2017) designed student profile forms to explore bilin- gual practices at home, exposure to both languages, countries where students have lived, and assessment data in both languages. We integrated these profile guidelines to address the translanguaging assessment framework described above (Garcı́a et al., 2017), including (a) various angles and perspectives through interviews with mothers and our interaction with them during assessment; (b) students’ performance independently and with moderate assistance by examining their individual performance reading the BAS texts, and then their retelling with mothers’ involvement; (c) assessments on authentic tasks by incorporating the family literacy practice of retelling in Spanish; and (d) analysis of miscues and retelling on features across students’ full language repertoire (with freedom to select across languages) and language-specific features (e.g. Spanish oracy, English lit- eracy). Our profiles included summary charts for performance across all texts, miscue analysis, strategies in retelling and excerpts from mothers’ inter- views. In this article, we include only excerpts from such profiles that help answer our research questions.

In the findings below, we mostly focused on one assessment event when children read in English first, and then retold the story of their choice in Spanish to their mother. We followed the standard BAS procedures to calculate accuracy rates, fluency and comprehension scores. In four sessions, Luis read a total of six books (levels A–D) and Alicia read eight books (levels C–H). We assigned quantitative scores for comprehension conversations following the BAS guidelines, which ranged from satisfactory (5/6) to excellent (6/6) for Luis, and from limited (4/6) to satisfactory (5/6) for Alicia. However, given critiques of the limits of comprehension quantitative scoring in this BAS version (Stahl and Garcia, 2017), we only report our qualitative analysis here. The third edition of BAS has updated its comprehension scoring rubrics (Heinemann, 2017).

For the analysis of oral reading patterns (RQ1), we followed the in-depth procedures in Goodman et al. (2005). This involves tabulating all miscues and coding for syntactic acceptability (is it grammatically acceptable in the child’s dialect?), semantic acceptability (is the meaning of the miscue acceptable within

Noguerón-Liu et al. 9

the text or at the sentence level?), high graphic similarity (does the miscue look like the text word?). We paid attention to patterns in syntactic acceptability, given children’s second language development. If miscues were syntactically and semantically acceptable, we analysed whether they changed the meaning of the text.

To document children’s use of linguistic resources in oral reading (RQ1), we categorized miscues with these codes: (a) grammatical approximations (e.g. ‘they tooked the mail’), (b) miscues of tier II or tier III vocabulary words (academic language and discipline-specific concepts), (c) non-words (e.g., ‘sile’ for ‘some’) and (d) miscues of function words (e.g. prepositions). Second, we used these same categories for oral retellings, to corroborate whether children were gaining control of grammatical constructions that may have influenced miscues. We added codes for (a) non-standard English varieties (e.g. ‘they be good’) and (b) code-switching to Spanish at the inter- or intra-sentential level (e.g. ‘pero, what is it?’). We also flagged instances for further enquiry on Spanish transfer in syntax or vocabulary (e.g. ‘for you can see’/ ‘para que veas’).

To explore strategies and resources in bilingual retellings (RQ2), we iden- tified target strategies in the scripted questions (e.g. inference) and compared performances in English and Spanish. We searched for potential comprehen- sion ‘blocks’ where the child’s understanding of the text may have been hampered by (a) potentially unknown vocabulary; (b) meaning-change mis- cues; or (c) background knowledge. Identifying these blocks allowed us to track strategies children used to negotiate meaning. We then coded resources children used during the retelling, including illustrations, captions and unprompted re-reading of the text. Lastly, we looked at mother’s roles and participation in the retelling process and debriefing (RQ3). We coded thema- tically their construction of children’s performance (e.g. evaluation state- ments, comparison with past performances), as well as their reporting of related home practices.

Collaborative coding process

We engaged in collaborative coding of most data, relying on dialogue and joint reflection on discrepancies as sources of further enquiry. We acknowl- edge the diverse interpretive stances of each coder/researcher, informed by our positionality in assessment events (administrator/observer), our relation- ships with participants, and linguistic and professional repertoires. As Saldaña (2013) explains, each coder experiences the relation with the data differently,

10 Journal of Early Childhood Literacy 0(0)

and hence, a process of dialogical intersubjectivity and group consensus aligns better with our analytical process.

Yet, the initial purpose of having two assessment administrators was to obtain two records of each oral assessment. We held meetings after each assessment round in which Silvia compared Courtney’s and Chelsey’s miscues and scores and listed discrepancies. We individually listened to the audio- recordings to revisit discrepancies and come to an agreement; all reported scores are results of this process. Silvia compiled Courtney’s and Chelsey’s initial miscue analysis coding, and then conducted the in-depth analysis of meaning-changing miscues and second-language acquisition categories. All authors coded the retelling conversations individually, using codes devel- oped by Silvia, who then compiled all instances identified by the team. Silvia coded all Spanish data (retelling with mothers and interviews).

Findings

In the following sections, we present patterns in students’ oral reading and retelling, and our analysis of resources across linguistic repertoires (RQ1). We then compare students’ performance in their retellings of a text in English and Spanish (RQ2). Finally, we include their mothers’ roles and comments in response to data analysed across sections (RQ3).

Repertoires in oral reading and retelling

Tables 1 and 2 present results from three focal texts for each child; these texts illustrate differences in their use of cueing systems over time and provide context for specific data excerpts analysed below. In the case of repeated errors

Table 1. Luis’ BAS results in focal texts.

Date

Title, level

and genre

Accuracy

%

Syntactic

acceptability

(%)

Semantic

acceptability

(%)

Graphic

similarity

(%)

No meaning

change (%)

Miscues

analysed

July 2015 At the Park

(A, non-fiction)

25 83.3 100 0 100 6

June 2016 Socks

(C, non-fiction)

72.2 23.5 5.9 35.3 5.9 17

June 2016 Our Teacher,

Mr. Brown

(D, non-fiction)

80.5 68 24 32 12 25

Noguerón-Liu et al. 11

(e.g. a substitution of ‘they’ for ‘the’ four times), the miscue was only coded once. Thus, the number of miscues analysed and the number of errors in accuracy rates are different, because BAS guidelines indicate repeated miscues must be counted as an error every time (unless they are names or proper nouns). This scoring guideline is also found in Clay (2000: 17).

Luis’ miscues and oral retelling patterns. In Level A, Luis’ errors were mostly inser- tions beyond the number of words on the page. For instance, in At the Park, he read ‘I can slide at the playground’ instead of ‘I can slide’. These insertions were mostly syntactic and semantically acceptable, but with little visual atten- tion to print. By levels C and D, Luis’ attempts to solve words (e.g. wăk/wake, we/with) had graphic similarity, but were not syntactically or semantically acceptable, and they affected meaning. His last performance showed some growth in syntactic acceptability, but the semantic acceptability of miscues was low, especially with substitutions using nonsense words (e.g. turk/tea- cher). Some miscues seemed related to his developing control of syntax – for example ‘they can to play’, or ‘putted’ instead of ‘put’.

We noticed similar patterns in his oral retellings, in verb conjugations in the third person (‘he ride bike’, ‘he help them to write’) or mixing function words that sound similar (‘the fly’, instead of ‘they fly’). We also found syntactic structures common in non-standard English varieties, including variations or omissions of the verb to be (e.g. ‘she playing football’, ‘they be good’) and subject-verb agreement (‘you was dead’). Most remarkable was the frequent intrasentential code-switching with the conjunction ‘porque’ (because). He used ‘porque’ at the beginning or middle of a sentence or clause a total of 13 times:

I know why porque his dad loves him a lot. (At the Park)

Table 2. Alicia’s BAS results in focal texts.

Date

Title, level,

genre

Accuracy

%

Syntactic

acceptability

(%)

Semantic

acceptability

(%)

Graphic

similarity

(%)

No meaning

change (%)

Miscues

analysed

July 2015 Shopping

(C, non-fiction)

87.5 100 100 33.3 100 3

Jan. 2016 Bedtime

for Nick

(G, fiction)

92.1 71.4 78.6 57.1 57.1 14

Jan. 2016 Trucks

(H, non-fiction)

86.7 66.7 16.7 61.1 16.7 18

12 Journal of Early Childhood Literacy 0(0)

The pictures porque they are playing together. (Best Friends) Porque, he don’t go to my, he don’t go to the school. (Our Teacher, Mr. Brown)

These patterns show Luis’ diverse repertoire in retelling, transversing language boundaries. On the one hand, his accuracy rates (below 90%), low semantic acceptability and high number of meaning-changing miscues all indicated significant areas of need, which Dionisia noticed at home: ‘yo sé que lee por medio de los dibujos’ [I know he reads from the pictures] (interview, 25 July 2015). Yet, in his oral retellings, Luis could reconstruct a story communicat- ing across linguistic repertoires. For instance, excerpts from his responses in Our Teacher, Mr. Brown show his use of Spanish conjunctions (‘que’ in lieu of ‘that’ and ‘porque’/because), illustrating Luis’ efforts to describe actions in a story drawing from his first language and approximations of repetitive pat- terns he notices in his second language.

Courtney: What did you learn in this book? Luis: Que, all the kids like Mr. Brown? [. . .] The teacher. They help him, they help him with the story, they help him to write, they help him to play with them, to play ball with them, they really love him. And they. . . have to paint with him, they have to get the best teacher. He reading a book with them. [. . .] Courtney: Why do you think Carl likes school? Luis: Porque his third-grade teacher, they are nice, he is nice and he help them to write, knows to make good sentences. His teacher is really good. (Retelling, 6 June 2016)

Because texts at early levels are simple with high picture-to-text correspon- dence, Luis was able to include several actions in the story (derived from illustrations) and inferences (e.g. they really love him). However, he also added details that were not in the book, such as the grade level (he inferred students were in third grade) and the pattern ‘they help him to. . .’ not present in the text. As Table 1 indicates, his miscues had low graphic similarity and semantic acceptability, changing meaning. These patterns match Dionisia’s comments about Luis’ retelling ‘como si estuviera platicando’ [as if he was talking to you] (interview, 25 July 2015) and her concerns over his decoding not matching the words in the page.

Alicia’s miscues and oral retelling patterns. Alicia’s oral reading also showed some variation in the distribution of cueing systems, with fewer semantically

Noguerón-Liu et al. 13

acceptable miscues in non-fiction with higher vocabulary demands (such as Trucks, in Table 2). Some words told by the administrator in these non-fiction texts included Tier 2 and Tier 3 words, such as snowplow, enormous and pollen. Some miscues changed verb conjugation (e.g. brush/brushed) and function words (e.g. could/can’t), with syntactic acceptability ranging from 40% to 100% across all the texts she read. Her oral retellings also showed developing control of irregular past tense (buy-ed, took-ed) and pronouns (e.g. ‘them was painting’). She used miscues in retellings a few times, some- times changing meaning partially (e.g., ‘work the sentence’ instead of ‘write the sentence’). However, during the first assessment (‘Shopping’), Alicia con- sistently substituted the word ‘cart’ (as in shopping cart) for ‘car’ – a word found in the text eight times. Courtney was puzzled by this repeated miscue since it lowered her accuracy rate to the frustrational level of 87.5%. Not counting this miscue every time, accuracy would have been 96% (indepen- dent level). After her retelling, Courtney asked Alicia:

Courtney: Do you ever go grocery shopping with your mom? Alicia: [Nods] Courtney: What do you put your groceries in, when you go to the grocery store? Alicia: [no answer] Courtney: How do you say it in Spanish? [points to picture] Alicia: El carro. Courtney: ¿El carro? Alicia: [Nods]

By substituting cart with car, Alicia’s miscue was not only syntactically and seman- tically acceptable (with high graphic similarity), it was also semantically acceptable in her first language, since ‘carro’ made sense as the right word. Hence, Alicia was drawing on her first-language and background knowledge in this miscue, which did not alter the meaning of the text. Yet, a focus on accuracy would place this level as being too hard for her. In debriefing with Teresa, she corroborated the use of the word ‘carro’ in their grocery shopping trips, and her intention to reinforce different ways to call this item (in both languages) on future trips.

Luis’ and Alicia’s performances show the development of their language and reading skills over time, as they used different cueing systems and strategies to solve words, shaping the quantity and type of syntactically acceptable miscues. We noticed how even when some miscues may have been shaped by their developing control of syntax (e.g. different conjugation of verbs), they were

14 Journal of Early Childhood Literacy 0(0)

still able to produce retellings that were satisfactory, using the same miscues from their oral reading. While this is not regarded as a language test, multiple linguistic resources shape children’s performance. In Luis’ retellings, his constant use of the conjunction porque [because] signals his comprehension of cause and effect in the texts he read. His unassisted retelling (integrating second-language approxima- tions, repetition of sentences and first-language resources) includes various details depicted in illustrations and some of the print. Lastly, Alicia’s use of what seemed to her to be a cognate (cart for carro) shows the relevance of debriefing miscues that may be shaped by students’ pooled linguistic knowledge.

Strategies and resources across bilingual retellings

Teresa and Dionisia reported the children usually engaged in retelling and translations of school books they took home. Teresa described asking ‘le digo, ¿de qué se trató el libro?’ [I tell her, what was the book about?] (interview, 28 July 2015). Likewise, Dionisia was impressed by Luis’ explanations, especially with dual language books, where she could check herself that his retelling matched the text: ‘Todo lo que yo leı́ en español, él lo decı́a y lo explicaba’. [everything I read in Spanish, he said it, and he explained] (debriefing, 22 July 2015).

In the following texts, the relationship between a child and his/her pet is central to the plot and resolution. In Socks, a young girl is trying to wake up her cat, who is shown sleeping in different rooms and on furniture. In Bedtime for Nick, a young boy cannot go to sleep until his dog enters his room and sleeps on his dog bed. Neither Luis nor Alicia had pets.

Luis’ retelling of ‘Socks’. Even when his accuracy was 72% and several miscues were nonsense words, Luis could identify the conflict and resolution of the story: ‘He didn’t want to wake up. And the girl want to wake up porque, he, he wanted to play with him. And he didn’t wake up. And at the end, he finally wake up and eat.’ He was able to infer characters’ motives (he woke up ‘porque he was hungry’) and described the characters’ features based on illustrations (‘porque his eyes closed’). Lastly, he pointed to a speech bubble with the word ‘purr’: ‘He finally wakes up, and he say purr, porque he was eating. Just meow.’ ‘Meow’ (not in the text) was brought up again in the Spanish retelling.

When Luis retold the story to his mother, she asked a few questions and tried to co-construct the story. In lines 12–13, Dionisia focused on the clarity of Luis’ speech, asking him to pronounce the beginning sound in ‘despertarlo’ more clearly. Yet, in line 17, when Dionisia started embellishing the story (suggesting the cat left the house), Luis emphasized what he was saying is true

Noguerón-Liu et al. 15

(line 16), as he tries to stick to the text. Luis described again the conflict and resolution, and in line 20, he embellished the cat’s inner dialogue (not in the text), repeating ‘meow’ in Spanish. The standard translation of ‘purr’ is ‘ron- ronear’, which is not a cognate.

(11) Luis: Este se conta (sic) de la niña querı́a despertar al gato porque querı́a jugar con él. (12) Dionisia: ¿Querı́a respertar o despertarlo? (13) Luis: Despertarlo. (14) Dionisia: Ah, para jugar con el gato. (15) Silvia: ¿Y luego que pasó? (16) Luis: Y luego, y luego que pasó, que él se paró. (17) Dionisia: Se le salió. [laughter] (18) Luis: No, si, si, ¡esto es de verdad Ma! (19) Dionisia: El gato. . . ¿no se despertó? (20) Luis: No se despertó, él dijo ‘‘nah, yo voy a seguir durmiendo,’’ y luego eh, y por fin, en la otra hoja, si, si se despertó a comer. Dijo ‘‘miau.’’ (21) Silvia: Oh, ¿dónde dijo miau? ¿Dijo miau? Y se despertó. (22) Dionisia: Ya le aumentó. (11) Luis: This tells us about the girl, who wanted to wake the cat up because she wanted to play with him. (12) Dionisia: Did she want to wake him up or respect him? (13) Luis: Wake him up. (14) Dionisia: Oh, to play with the cat. (15) Silvia: And then what happened? (16) Luis: And then, and then what happened, he stood up. (17) Dionisia: He escaped. [laughter] (18) Luis: No, if, if. . . this is true, mom! (19) Dionisia: The cat. . . did not wake up? (20) Luis: No, he woke up and said ‘‘Nah, I’m going to continue sleeping.’’ and then uh, finally, in the other page, he did, he did wake up to eat. He said ‘‘Meow.’’ (21) Silvia: Oh, where did he say meow? Did he say meow? And he woke up? (22) Dionisia: He added that.

This exchange illustrates the complexities in Dionisia’s participation, given her worries about Luis’ speech and his lack of attention to print. Yet, in line 20, Dionisia departed from the story – an effort Luis resisted – and she later pointed to the additional word in line 22 (‘miau’). Dionisia reported she and Luis enjoyed storytelling together, so the negotiation between accuracy, sticking to the text and playfulness shows up in this excerpt.

16 Journal of Early Childhood Literacy 0(0)

In further probing, Silvia asked Luis about his translation of ‘purr’. Luis re- read the page, pointing to the word. Silvia asked: ‘¿cómo dirı́a un gato en español?’ [what would a cat say in Spanish], and Luis answered ‘purr, ası́ dice ya, cuando yaaa. . . se acabaron todo’ [purr, that’s what you say, when they are done with all [their food]. When Silvia explained to Dionisia this was a sound cats make, Luis added ‘eso parece como un pájaro’ [that seems like a bird]. Hence, Luis inferred ‘purr’ may be related to a bodily function (analogous to burping), and then noticed the sound similarity between ‘bird’ and ‘purr’, missing the onoma- topoeic function. In both retellings, Luis was using a word he knew (meow), appropriate for his addressee (his mother), and syntactically and semantically acceptable. At the end of this interaction, Dionisia reinforced the appropriate- ness of Luis’ translation: ‘pero en español, nosotros, dice miau (. . .) él me lo dice bien a mı́’ [but in Spanish, we do, it says meow (. . .) he told it well to me], advocating for the recognition of Luis’ sense of audience.

Alicia’s retelling of ‘Bedtime for Nick’. In this book, the syntactic and semantic acceptability of Alicia’s miscues outnumbered graphic similarity (Table 2), showing she was reading for meaning. In her unassisted retelling in English, she identified the conflict in the text as ‘Nick was missing something. He was missing his dog.’ She inferred the dog ‘did not know it was time to sleep’ and the main character may have been ‘worried’ about the dog. When we invited Alicia to share the story with her mother, her approach was very different from Luis’, who only looked back at the last page of the book. Alicia took the book, opened it at the beginning, whisper read to herself in English, and then translated each page for her mother in Spanish. Because her voice was very low and she took long pauses, Silvia and her mom repeated her utterances, as shown in lines 22, 23, 25 and 27.

She used reported speech (instead of direct quotes) to paraphrase dialogue: some of it with similar meaning: ‘it’s time for bed’ (text) / ‘his mom told him to go to sleep’ (translation, line 22); some of it was verbatim, ‘Will you read me a story?’ (text)/ ‘he asked her to read him a story’ (translation, line 26). In line 27, Alicia used her original miscue from her oral reading, the substitution of ‘Nick liked the story about the magic fish’ for ‘music fish’— a miscue with some graphic similarity, but less semantically and syntactically appropriate. In her translation, she attempted to make the miscue work syn- tactically in Spanish:

(21) Alicia: [inaudible] (22) Teresa: Que la mamá le dijo que ya se duerma.

Noguerón-Liu et al. 17

(23) Silvia: Que ya se duerma. ¿Y luego? (24) Alicia: Y se lavó los dientes. (25) Silvia: Se lavó los dientes. (26) Alicia: Y. . . vino a la cama. [turns page] Y dijo que, que, que. . .. Que lea un. . . una historia. (27) Silvia: Una historia. (. . .) ¿Y después? (28) Alicia: Y. . . dijo algo a él, a él, la mamá. Tenı́a música, qué tenı́a música con un pescado. (21) Alicia: [inaudible] (22) Teresa: That the mom said go to bed. (23) Silvia: That he goes to bed. And then? (24) Alicia: And he brushed his teeth. (25) Silvia: He brushed his teeth. (26) Alicia: And. . . he came to bed. [Turns page]s And he said that that, that. . . that he reads a. . . a story. (27) Silvia: Yes, a story. (. . .) And then? (28) Alicia: And, she said something to him, to him, the mom. He had music, he had music with a fish.

This excerpt shows Alicia’s negotiation of multiple tasks: translating a text for her mother, including as many details as possible, and repairing a decoding error, making it ‘fit’ syntactically in her translation. While the music/magic substitution was not central for overall meaning, it indicated how, given the opportunity to re-read, Alicia still relied more on graphic similarity. Her focus on graphic cues aligns with her decision to re-read and translate verbatim several dialogue lines and events, sticking to her initial decoding of the text. Unlike most English retellings, when Alicia summarized stories in one or two lines, here she re-read and translated more comprehensively for her audience, perhaps aware that her mother had not understood the story from her English retelling, and that she did not have access to print.

These bilingual retellings show the ways children managed their various linguistic repertoires to retell and translate for their parents, as they built on the knowledge they constructed in their English retelling. Alicia negotiated the syntactic features of a miscue in her efforts to repair meaning, while Luis inferred the meaning of ‘purr’, comparing it with a similar onomatopoeic expression in Spanish. Both children paraphrased, embel- lished or translated dialogue in the stories, demonstrating understandings of events and character’s motives. These translated retellings also illustrated features of retelling at home, where Alicia seems to stick to the text to translate and decode for her mother, and Dionisia negotiates an emphasis

18 Journal of Early Childhood Literacy 0(0)

on accuracy with storytelling and language play, in order to support Luis’ speech development.

Discussion and implications

Our analysis of the linguistic patterns, strategies and resources in emergent bilinguals’ reading illustrates the potential of including home practices when interpreting data. We show how their emerging mastery of their second language and their use of pooled linguistic resources shaped their perfor- mances. The analysis of their bilingual retellings displayed students’ negotia- tion of the tasks of translating, re-reading and retelling for multiple audiences, with opportunities to repair miscue gaps, and to reflect on the translation of vocabulary items, with assistance from their mothers and researchers. These Spanish retellings also provide insights into home retelling practices, as inter- actions show how children talk about books when mothers say ‘Dime de que se trató’ [Tell me what this was about].

Since translanguaging views are based on interactive, fluid language prac- tices in family contexts (Garcı́a and Wei, 2014), applying this concept to assessment (Garcı́a et al. 2017) helps illustrate the complexity of pooled language resources. When children negotiate the appropriateness of ‘purr/ meow/miau’ or ‘carro/car/cart’, they are balancing their home language with content in a highly decodable text, designed to represent ‘familiar experi- ences’ – experiences they lived in Spanish. The cross-checking of their miscue syntactic patterns and language patterns in retelling, together with their trans- lation/retelling efforts, can paint a fuller picture of the ways they negotiate comprehension gaps, and the connections they make across languages, back- ground knowledge and new knowledge. These findings align with past work identifying linguistic resources in retelling events, where students can demon- strate their comprehension of texts across languages (Hopewell, 2011; Kabuto, 2016a), as well as the value of exploring second-language approx- imations in retelling (Croce, 2014).

In response to calls for equitable literacy assessment (Jiménez, 2004; Stahl and Garcı́a, 2017), this analysis shows ways to include students’ home litera- cies. Our findings provide insights into evaluating emergent bilinguals who are decoding in their second language, because they have not received aca- demic literacy instruction in their first language at school. By recruiting parents in the assessment process, we gained a window into family practices, which are shaped by the distribution of skills and roles at home. Dionisia and Luis were partners in oral storytelling and language play at home (given Luis’

Noguerón-Liu et al. 19

history with speech development), and their interaction showed Dionisia’s efforts to identify the differences between languages, the text and his retelling, all while advocating recognizing Luis’ translation. Alicia was aware of the importance of accurate decoding in addition to retelling, undertaking the task of line-by-line translation to make meaning accessible for her mother. At a surface level, these mothers’ backgrounds may not be recognized by institutions due to their lack of formal schooling and English proficiency. But their participation in assessment sessions allowed us to interpret the situated meaning of retelling at home, and their valuable contributions to this process.

Findings also illustrate the complexities of comparing retelling perfor- mances across languages when retelling and translating are culturally situated family practices. Their cases align with Baird et al.’s (2015) work, showing how family members engage with multilingual resources differently, focusing on meaning-making with Spanish texts, and on pronunciation with English texts. In similar ways to Reyes’ findings (2006), the role of interlocutors in retelling also emerged as relevant in language choice, as we saw how Luis and Alicia interacted with English-dominant researchers, and then with their Spanish-dominant mothers in a similar task. Overall, their performances call for further research on the features of first-language retelling in home set- tings, when emergent bilinguals are not socialized in retelling tasks in bilin- gual classrooms – which may be the case in some new migration destinations, or contexts with rapidly changing demographics and limited first-language educational resources.

This study also provides some insights into validity in oral reading assess- ments, especially when reading level decisions are mostly based on quantita- tive measures of accuracy and comprehension. Valid testing for emergent bilinguals should account for the linguistic factors involved in the scoring of tests, as well as their development and adaptation (Solano-Flores, 2008). When oral retellings rely heavily on the oral language production of students developing proficiency in English, the impact of their approximations and emergent skills must be considered in data interpretation (Croce, 2014). Stahl and Garcı́a (2017) point to the BAS Field Study to determine reliability and validity, and the lack of emergent bilinguals in such studies. Conducted in 2006–2007, their study included Latinx children in metro areas in traditional migration destinations in California, Texas and Florida, but only included those who were ‘proficient speakers of English’ reading at grade level (Heinemann, n.d.). As mentioned above, the BAS guidelines encourage

20 Journal of Early Childhood Literacy 0(0)

administrators to let students use their first language to respond in a comprehension conversation, or to sketch or write to express their ideas. Yet, our findings showcase additional considerations to explore repertoires across miscue patterns, oral retelling responses and ways in which children engage in retelling with their caregivers.

Implications for practice showcase the potential of providing assessment administrators with an enquiry stance towards translanguaging, even when they do not have the same linguistic repertoires of their students. For instance, drawing attention to repeated miscues (such as Alicia’s substitu- tion of car/cart) can inform retrospective miscue analysis conversations, directed towards teacher’s learning about students’ potential cognates, in the same way, Courtney asked: what do you call this in Spanish? Furthermore, Luis’ inclusion of the conjunction ‘porque’ illustrates the need to recognize code-switching as a resource, instead of being disre- garded as erroneous (Corella Morales and Lee, 2015). The composition of the research team also shows the relevance of collaboration between English-dominant practitioners, translators/cultural brokers and families. Understanding children’s linguistic repertoires can be a collective effort where the adults in the child’s life can explore how various resources come together when children learn to read. Even if parents cannot be physically present during assessment observations, retelling at home can be explored in home surveys about how children use both languages to complete academic tasks (Garcı́a et al., 2017). When children and families engage with various discursive communities in new migration settings, teachers also need new ways of seeing, evaluating and celebrating their growing and diverse repertoires.

Acknowledgements

The authors would like to express their gratitude to the families who participated in this study, as well as to the staff in schools and libraries committed to supporting immigrant caregivers and their children. They are also thankful for the thorough comments and recommendations provided by the editors and reviewers of this manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the William T. Grant Foundation (Grant Number 184469), Reducing Inequality Initiative.

Noguerón-Liu et al. 21

Note

1. We use the gender-inclusive term Latinxs instead of Latino/as to refer to indivi- duals whose backgrounds come from Spanish-speaking nations.

ORCID iD

Silvia Noguerón-Liu http://orcid.org/0000-0003-0892-5289

References

August D, Francis DJ, Hsu HA, et al. (2006) Assessing reading comprehension in bilinguals. The Elementary School Journal 107(2): 221–238.

Baird A, Kibler A and Palacios N (2015) ‘‘Yo te estoy ayudando; estoy aprendiendo también/I am helping you; I am learning too:’’ A bilingual family’s community of practice during home literacy events. Journal of Early Childhood Literacy 15(2): 147–176.

Clay M (2000) Running Records for Classroom Teachers. Portsmouth: Heinemann. Corella Morales M and Lee JS (2015) Stories of assessment: Spanish–English bilingual

children’s agency and interactional competence in oral language assessments. Linguistics and Education 29: 32–45.

Croce KA (2014) Assessment of Burmese refugee students’ meaning making of sci- entific informational texts. Journal of Early Childhood Literacy 14(3): 389–424.

Delgado Gaitán CD (2012) Culture, literacy, and power in family–community– school–relationships. Theory into Practice 51(4): 305–311.

Fountas I and Pinnell G (2011) Benchmark Assessment System 1. Portsmouth: Heinemann. Freeman AE (2001) The eyes have it: Oral miscue and eye movement analyses of the reading of fourth

grade Spanish/English bilinguals. PhD Thesis, University of Arizona, USA. Garcı́a O (2009) Bilingual Education in the 21st Century: A Global Perspective. Malden: Wiley-

Blackwell. Garcı́a O and Wei L (2014) Translanguaging: Language, Bilingualism, and Education. New York:

Palgrave. Garcı́a O, Ibarra Johnson S and Seltzer K (2017) The Translanguaging Classroom: Leveraging

Student Bilingualism for Learning. Philadelphia: Caslon. Gillanders C and Jimenez RT (2004) Reaching for success: A close-up of Mexican

immigrant parents in the USA who foster literacy success for their kindergarten children. Journal of Early Childhood Literacy 4(3): 243–269.

Goodman Y, Watson D and Burke C (2005) Reading Miscue Inventory: From Evaluation to Instruction. Katonah: Richard C. Owen.

Hamann E and Harklau L (2015) Revisiting education in the new Latino diaspora. In: Hamann E, Wortham S and Murillo E (eds) Revisiting Education in the New Latino Diaspora. Charlotte: Information Age Publishing, pp. 3–28.

Heinemann (n.d.) Field study of reliability and validity of the Fountas & Pinnell Benchmark Assessment Systems 1 and 2. Portsmouth: Heinemann.

22 Journal of Early Childhood Literacy 0(0)

Heinemann (2017) The role of recording forms across the Fountas & Pinnell Literacy brand. Retrieved from http://www.fountasandpinnell.com/resourcelibrary/ resource?id=379.

Hopewell S (2011) Leveraging bilingualism to accelerate English reading compre- hension. International Journal of Bilingual Education and Bilingualism 14(5): 603–620.

Hopewell S and Escamilla K (2014) Struggling reader or emerging biliterate student? Reevaluating the criteria for labeling emerging bilingual students as low achieving. Journal of Literacy Research 46(1): 68–89.

Hurtig J and Dyrness A (2011) Parents as critical educators and ethnographers of schooling. In: Levinson B and Pollock M (eds) A Companion to the Anthropology of Education. Malden: Wiley-Blackwell, pp. 530–546.

Jiménez R (2004) More equitable literacy assessments for Latino students. Reading Teacher 57(6): 576–578.

Kabuto B (2016a) A socio-psycholinguistic perspective on biliteracy: The use of miscue analysis as a culturally relevant assessment tool. Reading Horizons 56(1): 25–44.

Kabuto B (2016b) The social construction of a reading (dis)ability. Reading Research Quarterly 51(3): 289–304.

Reyes I (2006) Exploring connections between emergent biliteracy and bilingualism. Journal of Early Childhood Literacy 6(3): 267–292.

Saldaña J (2013) The Coding Manual for Qualitative Researchers, 2nd ed. London: Sage. Solano-Flores G (2008) Who is given tests in what language by whom, when, and

where? The need for probabilistic views of language in the testing of English language learners. Educational Researcher 37(4): 189–199.

Stahl K and Garcia G (2017) Using assessments to map and evaluate the comprehen- sion development of young children. In: Israel S (ed.) Handbook of Research on Reading Comprehension. New York: Guilford, pp. 214–237.

Vertovec S (2007) Super-diversity and its implications. Ethnic and Racial Studies 30(6): 1024–1054.

Noguerón-Liu et al. 23