Children’s Real-Time Behaviors during a Model Replication Task

• In Event: Beyond Accuracy: Children’s Correct and Incorrect Strategies on Early STEM Tasks Provide Insights into Cognitive Development

Abstract

Block play provides a window into children’s spatial cognition. Children’s performance on block-building assessments relates to performance on other spatial tasks and math assessments and to later participation in STEM disciplines (Nath & Szucs, 2014; Verdine et al., 2014, 2017). Much research on children’s block building has been in this vein, focusing on block-building performance, or “outcomes”. Less attention has been directed to the processes—children’s real-time block-building behaviors—that contribute to block-building success. What behaviors do children exhibit during block building, and which behaviors facilitate successful outcomes? To address these questions, we identify and document children’s real-time block-building behaviors during a model replication task and relate children’s behaviors to one another and to task outcomes.

Participants were 120 (60 girls) 5-year-olds (M=5.6 years; SD = 0.15) from African American (n = 30), Dominican (n = 30), Chinese (n = 30), and Mexican (n = 30) backgrounds. Children were instructed to replicate 3D target block structures built by a researcher (one at a time). The task consisted of 12 structures (trials) that generally required increasingly more blocks and a greater variety of block combinations as testing progressed. The task ended when children either failed three structures in a row or reached the 12th structure. Children were video recorded during the task and video-recordings were coded using Datavyu, a frame-by-frame video coding software (www.datavyu.org).

Behavioral codes were developed based on researchers watching two randomly selected videos (1 boy, 1 girl) from each of the four ethnic groups and were also guided by prior research (e.g., Huttenlocher, 1967; Park et al., 2009; Thornton, 1999). This process resulted in 6 coding variables: 1) time per trial, 2) observing the researcher build, 3) checking against target structure, 4) alignment with target structure, 5) error placements, and 6) trial outcome. Error placements were further classified as intrinsic (i.e., incorrect shape) or extrinsic (i.e., incorrect location or orientation).

Children varied enormously on all behaviors and trial outcomes. Successful trial outcomes were not related to children spending more time on the task or observing the examiner build the target structure at the start of trials. Rather, children’s alignment of structure and checking while building aided in selecting the correct shapes, correctly orienting and placing blocks in proper locations, and children’s ultimate trial success. Notably, children from Chinese backgrounds surpassed children from Dominican, Mexican, and African American backgrounds in task performance, and produced fewer errors than children from other ethnic backgrounds. Chinese children’s relatively high performance corresponded with their block-building behaviors. They were most likely to align their structure with the target structure, had the highest rate of checking, and had the fewest errors compared to children from other ethnic backgrounds.

Findings shed light on how a range of strategic behaviors, such as alignment, checking, and error detection and avoidance interact and uniquely contribute to children’s success at block building. Although we focused on a specific block-building task, we will discuss how an emphasis on real-time behaviors has implications for the science of children’s learning more generally.

Authors

• Daniel Suh, New York University
  Presenting Author

• Jake McCallum, New York University
  Non-Presenting Author

• Catherine S Tamis-Lemonda, New York University
  Non-Presenting Author