Poster #122 - Effects of Background Context Salience on Word Learning

• In Event: 2-110 - Poster Session 07
  In Poster Session: PS 07 Section - Language, Communication

Integrative Statement

Children learn words amidst a broad range of ambient context features which can vary widely across situations. While children are capable of tracking co-occurrence statistics of labels and referents across situations, the question of whether context variability affects word learning has recently received attention (e.g., Goldenberg & Sandhofer, 2013; Tippenhauer & Saylor, 2018; Twomey, Ma, & Westermann, 2017; Vlach & Sandhofer, 2011; Werchan & Gómez, 2014; Wojcik, 2017). These studies provide conflicting accounts for the effects of context on word learning. In some cases, two-year-olds are worse at learning words if training and test contexts vary (e.g., Goldenberg & Sandhofer, 2013; Vlach & Sandhofer, 2011; Werchan & Gómez, 2014). However, there is additional evidence suggesting that context variability does not reliably affect word learning (e.g., Tippenhauer & Saylor, 2018; Twomey et al., 2017; Wojcik, 2017).

It is possible that differences in background context composition across different studies can account for conflicting findings. In some studies that have found supporting evidence for context-dependent word learning (e.g., Vlach & Sandhofer, 2011), background contexts included images of nameable objects. In studies that have not demonstrated an effect, contexts were simply plain colors or patterns without nameable objects (e.g., Tippenhauer & Saylor, 2018; Wojcik, 2017). There is a body of work demonstrating that variation in competitor referents during exposures to a novel label can disrupt children’s word learning (e.g., Axelsson & Horst, 2010). Thus, differences in context composition (or salience) may account for differences in the context-dependent word learning literature.

The present study investigated whether background context salience affected 2.5-year-olds’ (N = 24, M = 2 years, 6 months) label acquisition. In one experiment, we tested children’s word learning in variable salient and variable non-salient contexts. Salient contexts were cloth backgrounds that included images of nameable objects, which could act as competitor referents to novel targets. Non-salient contexts were cloth backgrounds that were solid colors or patterns without images of nameable objects. There was a significant difference in target selection on immediate test trials following training in the two contexts. Children selected targets a greater proportion of the time when they were exposed to labels in the non-salient context (M = 0.57, SD = 0.40) than in the salient context (M = 0.40, SD = 0.41), F(1, 22) = 5.11, p = .03. However, this condition difference was not present after a 10-minute delay, suggesting that context salience may not affect longer-term retention of object/label pairings. These results may off one explanation for why there are disparate findings in the context-dependent word learning literature. In addition, the results draw into question what information children are reliably attending to and encoding during a word learning episode.

Authors

• Nicholas Tippenhauer, Vanderbilt University
  Presenting Author

• Megan Michelle Saylor, Vanderbilt University
  Non-Presenting Author