Writing development: Predictors, profiles, and intervention

Writing development: Predictors, profiles, and intervention

First Author: Hope Gerde -- Michigan State University
Keywords: Writing development, Growth Modeling, Orthography, Early Literacy, Intervention
Abstract / Summary: 

This dynamic symposium examines children’s writing, a fundamental but understudied area of development related to later reading and writing success (Graham & Hebert, 2011; NELP, 2008), in preschool and adolescence. The studies offer new ideas and the potential to ignite insightful discussion regarding critical factors that influence writing development across time and the interrelations existing within profiles of early and later writers. The works utilize multiple theoretical perspectives and a range of methodologies to answer questions of writing development for diverse students including children living in poverty and speaking varied languages. The symposium begins with projective and retrospective examinations of predictors of writing/spelling growth, followed by the exploration of writing profiles in both preschool and middle grades, noting the unique patterns of component skills within the groupings. The presentations conclude with the teacher and child-level outcomes identified in a RCT of an innovative early writing intervention for Head Start teachers.

Symposium Papers: 

Level and growth of children’s language and decoding skills predict later writing achievement

First Author/Chair:Sonia Cabell -- Florida State University
Additional authors/chairs: 
Hope Gerde; HyeJin Hwang; Ryan Bowles; Lori Skibbe; Shayne Piasta

Purpose: The purpose of this study was to examine the extent to which level and growth in children’s early language and literacy skills predict later writing achievement.

Method: Data were individually collected for 334 children as part of a larger study at four time points, approximately six months apart, beginning when children were 3- to 4-years old. At all timepoints, oral language, phonological awareness, and decoding skills were measured. At the fourth timepoint (when children were 5- to 6-years old), writing (namely transcription) was measured across a variety of subskills: name writing, letter writing, word writing, and sentence writing. A series of non-linear growth models were estimated, using FIML to account for missing data, and covarying family income, child age, race, and gender.

Results: Results indicated that the initial level (Timepoint 1) of most skills was significantly associated with later writing achievement across measures (Timepoint 4). Moreover, the growth rate of decoding skills significantly predicted children’s word and sentence writing, and the growth rate of oral language skills significantly predicted all four measures of children’s writing. Children’s growth in phonological awareness did not predict subsequent writing achievement.

Conclusions: Although children’s early language and literacy skill level is an important predictor, the rate at which children acquire these skills over time makes a unique contribution to their later writing success. In other words, how quickly children codify early skills matters to their subsequent writing ability.

Do bedtime stories build the foundation for later reading preferences, spelling skills, and vocabulary?

First Author/Chair:Sandra Martin-Chang -- Concordia University
Additional authors/chairs: 
Brittany Tremblay; Gene Ouellette

Purpose - Few studies have examined the links between storybook reading before kindergarten and academic abilities in adolescence. Here, we examined whether storybook reading was associated with differences in high school spelling, word-recognition, and print exposure in English and French.

Method – Forty-five adolescent-parent dyads independently answered retrospective Title Recognition Tests (TRT) about storybook reading, containing the names of popular English and French storybook titles published before 2007. Adolescents also completed a bilingual Author Recognition Test (ART), as well as spelling and word-recognition tests in both languages.

Results – English: Adolescents’ TRT scores were correlated with both their parents’ TRT scores and their own ART scores. Adolescents’ TRT scores accounted for 9% of unique variance in their ART scores, after controlling for parental education, word-reading, word-recognition, and spelling. Hierarchical regressions revealed that adolescents’ TRT scores also accounted for unique variance in spelling (10%) and word recognition (12%) when entered into the regressions first, however, the contribution of the TRT was no longer significant after accounting for their ART scores. French: Adolescents’ ART scores accounted for 14% of unique variance in French spelling, after controlling for parent’s education.

Conclusions – Print exposure in high school is associated with spelling and word-recognition abilities in both languages. Interestingly, shared storybook reading is linked to print exposure and continues to be associated with English spelling, and word-recognition abilities in both languages. Thus, we encourage parents to continue reading storybooks as they seem critical to the future reading behaviors and abilities of their children.

Profiles of preschool writers: Addressing emergent writing with attention to transcription, composing, and executive function

First Author/Chair:Margaret Quinn -- University of Tennessee, Knoxville
Additional authors/chairs: 
Ryan Bowles; Gary Bingham; Hope Gerde

Despite recent frameworks for understanding the skills and competencies that comprise early writing (e.g., Kaderavek, Cabell, & Justice, 2009; Puranik & Lonigan, 2014), limited research examines the profiles of emergent writers with attention to children’s composing and transcription (handwriting and spelling) skills.
Purpose: This study examined preschoolers early writing developmental profiles to 1) consider how children varied on foundational early writing skills, and 2) to better understand the competencies and processes that theoretically comprise early writing.

Method: A diverse sample of 427 preschoolers (age ranging from 40 – 74 months) were assessed on a variety of early writing assessments, including: name writing, letter writing, invented spelling, and bubble/narrative writing. Children’s writing samples were scored for the following features: handwriting quality and fluency, spelling, quality of discourse/ideas, cohesion, and planning, aligning with current conceptual understandings of writing (e.g., transcription, composing, and executive function [EF]; Berninger & Winn, 1996).

Results: Latent profile analysis in MPlus was used to determine writing profiles. Six profiles emerged from the data revealing distinctive qualities and skill levels across theoretical writing components – transcription, composing, and EF (e.g., Profile 1 – low transcription, low composing, low EF; Profile 3 – high transcription, high composing, high EF; Profile 5 – low transcription, mid/high composing, high planning; see Figure 1).

Conclusions: Results suggest variable performance across writing components leading to distinctive writing profiles across children. Further examination will lead to powerful implications for understanding development, impacting factors contributing to profile membership, and instruction.

Diagnostic profiles of written expression in middle grades

First Author/Chair:Adrea Truckenmiller -- Michigan State University
Additional authors/chairs: 
Ryan Bowles; Margaret Quinn

Writing is a complex process that requires students to coordinate knowledge, abilities, and skills in multiple areas. Students may struggle with one or more of these skills and teachers need to know which skills students need more explicit instruction. The current study examines patterns of component skills of writing that typically predict writing achievement in middle grades. We conducted a profile analysis of 290 students in Grades 6 through 8 on measures that represent the teachable components of the Simple View of Writing – total amount written, writing fluency, spelling, planning, and the levels of language (word, sentence, discourse) within text generation (Berninger et al., 2002; Wagner et al., 2011). Three distinct profiles emerged: a small group of lower performers (n = 28), and two groups of typical performers. The typical performers had similar scores except the strong writers (n = 116) had higher scores on planning, total amount written, writing fluency, and discourse cohesion relative to the average group (n = 146). The profiles were strongly related to overall writing skills from the Test of Written Language: lower performers scores 1 SD below the mean of the norming sample; low planners score at the mean; and high planners score 1 SD above the mean. The profiles were also related to low, average, and high performance on the end of the year state test. Results suggest that instruction targeting planning, production, and cohesion may address the needs of many students.

Teacher and child-level outcomes of the iWRITE professional development intervention

First Author/Chair:Hope Gerde -- Michigan State University
Additional authors/chairs: 
Gary Bingham; Ryan Bowles; Alexa Meier; Xiao Zhang

Purpose: Instruction in early writing—an essential predictor of later reading (NELP, 2008)—can promote a range of literacy skills (Hall, Simpson, Guo, & Wang, 2015; Hofslundsengen, Hagtyet, & Gustafsson, 2016); however, quality instruction in early writing is rare in preschool (Bingham, Quin, & Gerde, 2017), Thus, our team developed the Improving Writing Resources and Instruction in Teaching Environments (iWRITE), an online professional development (PD) focused on enhancing language and literacy outcomes for young children through quality writing environments and supports. The intervention included access to eight online learning modules focused on early writing and an expert coach who provided individualized feedback on teacher-posted videos.

Methods: Fifty-four diverse Head Start teachers participated in a RCT evaluating the iWRITE PD with three pre-test and post-test assessments spanning seven months after intervention commencement.

Results: Results using linear latent growth curve analysis indicated that treatment teachers increased more rapidly on the Early Language and Literacy Classroom Observation (ELLCO) writing subscale over the course of the study (3.0 times faster, p<.001), leading to a higher score at post-test (d=1.4, p=.001). Similarly, treatment teachers increased more rapidly (2.3 times faster, p=.002) and had higher final scores (d=.9, p<.001) on the WRITE (Gerde, Bingham, & Pendergast, 2015) measure of classroom writing materials (6.4 points higher, p<.001). In addition, treatment teachers exhibited higher scores on the ELLCO total (d=.8, p=.05).

Conclusions: Findings highlight the iWRITE PD’s capacity to improve classroom-level writing environments and the broader literacy environment. Examination of the child writing and literacy outcomes is ongoing and will be presented.