Sentence priming in English and Spanish among bilinguals with good and poor reading comprehension

• In Event: Poster Session 08
  In Poster Session: PS 08 Topic - Language, Communication

Abstract

It has been suggested that context, combined with cognitive-perceptual skills such as lexical, phonological, and semantic processes, allows access to the underlying meaning of sentences and larger texts in reading (Christopher et al., 2012). Word knowledge is crucial for the development of other linguistic-related skills that, in turn, are crucial to gain proficiency in a language (Perfetti & Stafura, 2014), and language proficiency plays a crucial role in the development of both L1 and L2 reading skills among bilingual individuals (e.g., Droop & Verhoeven, 2003). This study aims to investigate whether there was evidence for the lexical identity component of the lexical quality hypothesis (LQH; Perfetti, 2017) in a sample of 67 Spanish-English bilingual children enrolled in grades 3 to 6 in Honduras.

Children were administered verbal and nonverbal working memory measures as well as Spanish and English standardized measures of vocabulary knowledge and reading comprehension skills. Children also completed a reading and listening sentence priming task to examine the effects of context on reading comprehension. The task consisted of two task modes (reading and listening), in two languages (Spanish and English), and three sentence context types (predictable, unpredictable, and neutral). Each task mode was comprised of fifteen trials per context types. Sixty concrete words (30 per language) were used to create sentences that consisted of three segments, the priming phrase, a display with four images (one corresponding to the target word and three distractors), and the final phrase. In the predictable context trials, the target words were semantically related to the words in the priming phrase (e.g., I went to the library to read a [apple, BOOK, clock, pillow] about animals), but the target words in the unpredictable context trials were not (e.g., At the store, I bought a [squirrel, beach, rock, BOOK] about plants). To examine the effect of context, the target words for the predictable context were the same as those for the unpredictable context.

Preliminary analysis from a linear mixed effects model revealed a significant main effect of context, with more correct responses for predictable than for unpredictable context (g = 0.18). These effects were larger for the lower grades than they were for the upper grades ( g = 0.21 to 0.46). Finally, there was a significant context, task mode (reading and listening), language, and grade interaction with no significant context effects across all grades showing for the English listening task, but significant context effects for the Spanish listening task for third and fifth grade. For the English and Spanish reading tasks, significant context effects were observed for all but for sixth grade children.

Results from the reading and listening tasks in both languages indicated that there were more significant context effects for reading than for listening. Also, the effects of context in the reading task were stronger for the younger children who presumably have weaker decoding skills, supporting the LQH. Follow-up analysis will examine whether differences in context effects vary as a function of individual differences in reading-related skills and working memory.

Authors

• Sergio Leiva, University of Nebraska-Lincoln
  Presenting Author

• John Marc Goodrich, University of Nebraska - Lincoln
  Non-Presenting Author