Spelling Matters! Revealing the complexity of the relation between spelling and reading

Spelling Matters! Revealing the complexity of the relation between spelling and reading

First Author: Nicole Conrad -- Saint Mary's University
Keywords: reading and writing relationship, Learning to read, Learning to spell, Spelling, Reading
Abstract / Summary: 

Research shows that spelling and reading are related. Both are components of literacy development and both depend on many shared cognitive and linguistic skills. However, if we are to bridge the gap between research and practice, we must better understand how these two skills are related. The diverse set of research studies included in this symposium each contribute a piece to understanding this larger theoretical question. Four studies examine transfer between spelling and reading at the individual word level. First, Conrad demonstrates that children in Grade 2 can establish in memory strong word-specific representations during spelling that facilitate word reading, but differences in reading speed and accuracy occur depending on the quality of the established representations. Second, Ouellette demonstrates that for children in Kindergarten, practice spelling a word enhances vocabulary learning, and vocabulary learning enhances learning to read that word; yet the direct relation between spelling practice and word reading was not evident at this age. Ruberto and colleagues next demonstrate that gains in spelling skill for French-speaking children are greater following explicit spelling instruction combined with reading practice than reading practice alone. Tucker and Deacon then demonstrate that accurately decoding a word in text enhances the learning of the spellings of new words for children in Grade 2, but not in Grade 1. Breadmore and colleagues take a step back from transfer at the individual word level to test whether individual differences in spelling onset latency predict individual differences in reading and spelling skill a year later. Together, these studies reveal the complexities of the relation between spelling and reading, complexities that are amplified by the diversity of these studies, which together provide ideas of how and why spelling and reading are related. A better understanding of the how leads to improvements in instructional practices

Symposium Papers: 

Variation in lexical quality on word reading outcomes: Assessing the quality of word-specific representations established during spelling practice

First Author/Chair:Nicole Conrad -- Saint Mary's University

Purpose: Skilled reading develops through the establishment of word-specific representations in memory, which can occur incidentally during spelling. Spelling accuracy and consistency gauge the quality of these established representations (Perfetti, 1997). The current study examines the effect on pseudoword reading of variation in quality of representations established during spelling; specifically, how consistency and accuracy of spelling attempts indicate lexical quality and affect subsequent pseudoword reading outcomes.

Method: Grade 2 children learned one of two target spellings (e.g., choom/chume) then practiced spelling these pseudowords four times within stories. Spellings were coded as consistent target: consistently producing the target spelling provided, consistent phonological error: consistently producing an alternate phonologically plausible spelling, consistent error: consistently producing an incorrect spelling, or inconsistent error: producing multiple erroneous spellings. One day later, participants spelled and read the target pseudowords to assess quality of the lexical representations established during spelling practice, and the effect of quality on target pseudoword reading speed and accuracy.    

Results: Chi-square analysis revealed that spellings produced during spelling practice were unevenly distributed. Over 50% of spellings were incorrect, but 95% were consistent spellings. Reading errors were infrequent but most were from pseudowords in the consistent error category. Preliminary item-based analysis revealed that children were faster reading “consistent target” than “inconsistent error” pseudowords. There was no difference in reading times for “consistent target” and “consistent error” pseudowords.  

Conclusion: Children establish strong representations when spelling that may or may not be “correct” and this can have implications for word reading.

Capitalizing on shared storybooks in kindergarten: The interplay between semantics, spelling, vocabulary acquisition and learning to read

First Author/Chair:Gene Ouellette -- Mount Allison University

Purpose: To examine how vocabulary learning through storybook exposure can be heightened by elaborated semantic teaching and spelling practice, and how this in turn may impact learning to read. There is a robust literature demonstrating that children learn vocabulary from repeated exposure within storybooks; a smaller number of studies have found that learning can be increased with the addition of elaborated semantic teaching. What remains less clear, is whether this vocabulary learning offers an advantage when it comes to learning to read those same words and if student-directed spelling practice would provide further benefit.

Method: The present study uses the well-established paradigm of embedding (10) non-words within a storybook to instantiate word learning in a natural context. Students in kindergarten (N=56) were seen in small groups (4-6), three times in one week, where the same modified story was read in a dialogic fashion each time. The teacher provided additional semantic enrichment for 5 of the embedded non-words. Some children also practiced the new words through guided invented spelling (Gentry & Ouellette, 2019). On the fourth day, participants were assessed for vocabulary learning and word reading.

Results: There was clear vocabulary learning from the storybook exposure. Importantly, participants showed superior performance on a learn-to-read task for the items exposed within the stories. There were no additional benefits to word reading derived from the additional semantic teaching or spelling practice, although each did result in increased vocabulary learning.

Conclusion: Vocabulary words learned through storybook exposure are at a benefit when it comes to learning to read those same words. Initial lexical representations built through vocabulary learning facilitate successful word reading on a word-specific basis; semantic depth and spelling may strengthen this learning at this time in development.

Does formal explicit instruction promote learning of words’ visual aspects? Results from a quasi-experimental study conducted on second-grade French-speaking children.

First Author/Chair:Noémia Ruberto -- Université de Montréal
Additional authors/chairs: 
Daniel Daigle; Ahlem Ammar

Purpose: The majority of spelling mistakes in French are related to words’ visual aspects such as multigraphemic phonemes and silent letters that do not carry meaning (Daigle et al. 2016). Recent studies have also shown that formal explicit instruction leads to more gains in spelling (Daigle et al. 2019, Marin and Lavoie, 2017). To define the most efficient formal teaching context to learn words’ visual aspects, we conducted a quasi-experimental study on 131 French-speaking children (mean age: 7.5).

Method: Two teaching devices were tested in four elementary school classes: 1) teaching of visual and semantic aspects of words (TVS, n=43) and 2) teaching of visual aspects of words only (TV, n=44). Two other classes formed the control condition (CC, n=44). In both teaching conditions, 24 words were taught: half of these words was included in a reading-only context while the other half was incorporated in a reading and writing context. Participants performed a gap dictation of the 24 targeted words; performances were analyzed at a lexical, sublexical and also at a specific level. To determine the potential differences between groups, ANOVAs were conducted to compare the performances before and after the intervention.

Results: Spelling gains were only observed for children in the TVS and TV conditions and words included in a reading and writing context are better learned than those in a reading-only context. Both teaching devices promoted the learning of multigraphemic phonemes and silent letters.

Conclusion: Results provide interesting insights into the practices that teachers should focus on to promote spelling acquisition.

An evaluation of phonological recoding as the key mechanism for learning new spelling patterns during emerging readers’ independent reading.

First Author/Chair:Rebecca Tucker -- Dalhousie University
Additional authors/chairs: 
S. Hélène Deacon

Purpose: The Self-Teaching Hypothesis (Share, 2008) posits that, for readers of all skill-levels, phonological recoding is the key mechanism to learning the spelling patterns of novel words encountered within passages of text; however, research on this prediction has focused on older children. The current study focuses on emerging readers, evaluating the role of phonological recoding in learning accurate spelling patterns during independent reading and the ability to transfer that learning to the processing of novel related words.

Method: Ninety English-speaking children in Grades 1 and 2 were asked to read nonwords (e.g., lurb) embedded in short stories. Children subsequently completed an orthographic choice task containing the same items as in the stories, as well as two novel forms of the nonwords not seen in the stories (e.g., lurber, lurble).

Results: Preliminary results suggest a clear facilitation effect for children in Grade 2: children who decoded a nonword accurately at least once performed better on the orthographic choice task for both target and novel nonwords. There was no evidence of this facilitation effect for children in Grade 1.

Conclusion: These findings suggest that phonological recoding facilitates emerging readers’ learning of the spelling patterns of new words encountered during independent reading, as well as their ability to transfer that learning to their processing of related words. Notably, the role of phonological decoding appears to change over the course of reading development, playing less of role for new readers (Grade 1) than for more experienced readers (Grade 2; Grades 3 & 5 as in Tucker et al., 2016).

Spelling fluency as a predictor of children’s reading and spelling outcomes

First Author/Chair:Helen Breadmore -- Coventry University
Additional authors/chairs: 
Emily Côté; S. Hélène Deacon

Purpose: Spelling offers insight into children’s progress connecting their oral and written language. Metrics of children’s spellings (and misspellings) can predict their progress in learning to read. Here, we examine whether the speed with which children begin to spell words could offer additional insight. We test the contribution of spelling onset latency, accuracy and the nature of errors in Grade 1 in predicting word reading and spelling in Grade 2.

Methods: We conducted a longitudinal study following 100 English-speaking Canadian children. At Grade 1, children completed the spelling task alongside standardised measures of non-verbal ability, phonological awareness, working memory, and vocabulary. Responses to the 10 item spelling task (Treiman et al., 2016) were digitised using an Inking Pen on a graphics tablet. Spelling onset latency was extracted using Eye and Pen 2. Spellings were also coded for accuracy, orthographic and phonological plausibility. Preliminary results show that children are ~50% accurate, ensuring that there is both sufficient variability in accuracy for errors analyses and adequate correct spellings for use of latency data. Grade 2 outcome measures include standardised measures of word reading and spelling.

Results: Hierarchical linear regressions will test which measures—accuracy and/or error coding and/or speed—contribute longitudinally to children’s reading and spelling outcomes (after accounting for control variables).

Conclusions: These analyses will offer the first insight into the utility of temporal dynamics of children’s spelling to predict broader literacy development. Findings could inform future conversations about spelling fluency parallel to reading fluency.