Poster #58 - Finding a threshold for Dual-Language Learners’ L2 vocabulary learning on screen: An eye-tracking study

• In Event: 2-002 - Special Poster Session 05 with Continental Breakfast Reception
  In Poster Session: PS 05 - Ethnic and Racial Issues Section

Integrative Statement

Educational media is prominent in the lives of young children, providing them with rich vocabulary learning experiences (Barr & Linebarger, 2017). Research has begun to investigate how media promotes vocabulary learning among preschool-aged Dual-language Learners (DLLs) (Silverman & Hines, 2013). Drawing from Paivio’s (1986) dual-coding theory, which purport that screen-based media captures children’s attention and provides them with robust learning experiences, this study examines how certain screen-based instructional supports might facilitate vocabulary learning among young DLLs. It also responds to Goldenberg’s (2013) call for instructional approaches that specifically benefit DLLs whose learning is influenced by language proficiency (Cummins, 1979). Therefore, this study examines the extent to which specific instructional supports influence vocabulary identification and attention to screen.
The sample consisted of 45 four-year-old DLLs (44% female) from two Head Start programs in poverty-impacted neighborhoods. The average age was 4.15 years (SD=.22). Research assistants administered a screening tool and PPVT-4 to measure baseline English vocabulary knowledge. An adapted Alberta Language Environment Questionnaire (Paradis, 2011) was given to teachers to capture children’s exposure to the English language at home and school. Employing a within-subjects design, whereby each child was exposed to each condition and served as their own control, children watched 12 randomized video clips that adopted two categories of support: ostensive cues, designed to provide definitional information, and attention-directing cues, designed to orient children’s attention to a target word. Video clips were viewed on a Tobii Technology T120 eye-tracker. After viewing, children were administered a 24-item receptive word identification task: 12 words in context and 12 words in new context to capture depth of word knowledge (Nagy & Herman, 1987).
Dynamic Areas of Interest were drawn around the target items on screen for the entire time the item was on screen. In the analysis, a repeated measures ANOVA was used with the four pedagogical supports as the within-subjects factor. Child’s age in months was used as a covariate to account for potential developmental differences across the 4-year-old age span in the sample. Finally, a median split was used for analysis (High language proficiency mean = 3.77; low language proficiency mean = 1.23 on a five-point scale).
Findings revealed that children less proficient in English identified more words through the ostensive cues. There were, however, no significant differences between ostensive or attention-directing cues between children with high English language proficiency. Interestingly, in the new context measure, children with low English language proficiency demonstrated significant vocabulary gains in attention-directing cues rather than the ostensive cues. Again, no significant differences occurred between those with high language proficiency. Moreover, children with low English language proficiency were slower to orient their attention to vocabulary supports on screen. The role of language proficiency in young DLLs’ vocabulary learning will be discussed, highlighting the potential of using educational media to lower the threshold (Cummins, 1979; 1981) for language learning in young children, as well as the need to differentiate instructional approaches for DLLs who are often treated as one category.

Author

• Kevin M Wong, New York University
  Presenting Author