Handwriting Legibility and Its Relationship to Spelling Ability and Age: Evidence From Monolingual and Bilingual Children.

Markéta Caravolas,Cameron Downing,Catrin Leah Hadden,Caspar Wynne

doi:10.3389/fpsyg.2020.01097

Abstract

Studies of the relationship between spelling and handwriting concur that spelling skills influence the dynamic processes of handwriting. However, it remains unclear whether variations in spelling ability are related to variations in the legibility of handwriting, how important spelling skills are relative to the amount of handwriting experience afforded by an individual’s age and number of years of schooling, or to what extent this relationship may be task- and orthography-specific. We investigated these questions in a study comparing spelling and handwriting legibility in a group of N = 127 Welsh-English bilingual children matched in age and number of years of schooling to a group of N = 127 English-monolingual children, as well as to a group of N = 127 younger, English monolingual children matched to the bilingual group in spelling ability. All groups completed the Spelling and Handwriting Legibility Test (SaHLT) and a broader battery of literacy measures. The bilingual children were found to have poorer handwriting legibility than same age peers, and in some cases, than their younger, spelling-ability peers, suggesting that spelling ability, more so than amount of handwriting experience and years of schooling impacts handwriting legibility. This was corroborated in a series of multi-group path models, where all children’s handwriting was predicted by spelling ability more strongly than by age, and, the effect of spelling generalized across two different spelling tasks in all groups. Finally, bilingual children seemed to draw on general (Welsh) as well as on orthography-specific (English) knowledge when handwriting in English.

Highlights

Spelling and handwriting skills, called transcription skills (Juel et al, 1986; Berninger and Swanson, 1994), form a crucial, but to date understudied skill set in children’s writing acquisition
We probed whether associations between spelling and handwriting among bilingual writers depended on the orthography in which they were writing or whether they reflected general spelling skill (i.e., Welsh and English). We addressed these questions in a series of multi-group path analyses, in which spelling ability, age, non-verbal ability and reading skills were included as potential predictors of handwriting legibility
The analysis of covariance on the raw scores revealed a main effect of the covariate F(1,357) = 25.88, p < 0.001, η2ρ = 0.07, as well as a main effect of group such that, after controlling for the effect of non-verbal ability, the bilingual group read English words more efficiently than their monolingual peers (BIL vs. chronological age (CA): t(251) = 4.78, p < 0.001, d = 0.68; BIL vs. SA: t(252) = 10.53, p < 0.001, d = 1.21) – a finding to which we return in the Discussion – and in turn, the younger SA group read less efficiently than their older monolingual peers (t(251) = −5.35, p < 0.001, d = 0.60)

Summary

Introduction

Called transcription skills (Juel et al, 1986; Berninger and Swanson, 1994), form a crucial, but to date understudied skill set in children’s writing acquisition. The ANCOVA on the standardized English Picture-Word Matching scores demonstrated that, after controlling for the significant effect of non-verbal ability (F(1,357) = 27.23, p < 0.001, η2ρ = 0.07), the groups’ performances did not differ from each other, and all groups were reading well within the normal range relative to their normative populations. This was mainly done to ascertain that the bilingual group’s relatively strong performance on the English reading test did not reflect a group with strong reading skills in their language of education This ANCOVA on the standardized reading scores revealed that after controlling for a significant effect of non-verbal ability (F(1,356) = 27.51, p < 0.001, η2ρ = 0.07), the three groups did not differ from each other, and all were reading well within the normal range

Methods

Results

Conclusion