Young Children with Speech-Language Disabilities and Visual Attention to Oral Language Tasks

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Language, Communication

Abstract

Capturing eye-movement patterns with computer technology is a window into cognitive processing (Rayner, 1986; Rayner et al., 2013). For preschoolers and kindergarteners, eyetracking has revealed how attention is paid to early literacy tasks, such as receptive vocabulary or picture storybook comprehension (Justice et al., 2005; Liao et al., 2020; Thompson et al., 2018). Determining whether children focus on core content may affect both the design and validity of early language assessments. These considerations are especially important for children with speech/language production difficulties, who may not be able to discuss their perceptions of a task with examiners. Enhancing our understanding of children's early language processing is critical to monitoring progress and determining appropriate interventions. The current study considers the nature of gaze patterns for young children with speech/language impairment when taking a computer-based test of phonological awareness (PA), the sound structure of language.

Children (n=62; 3;0-8;4yrs) with speech/language impairment were recruited from a variety of schools in the Midwest from diverse educational, socioeconomic, and racial/ethnic backgrounds. Eighteen children also had secondary diagnoses (autism spectrum disorder, attention-deficit/hyperactivity disorder, etc.). All children took ATLAS-PA (Skibbe et al., 2020), an English PA assessment suitable for children with speech/language impairment. Items were presented on a plain white background one-by-one on a laptop computer, with a recorded voice asking children to blend syllables/sounds to make words (e.g., "Put cow... boy together" or "Put mas...k together"). Stimulus words were all highly concrete and reviewed by experts with experience in special education and early language. Children heard the audio stimulus twice while being shown three side-by-side images. Children pointed to the screen to indicate their choice, then an assessor mouse-clicked on their response. A Tobii X2-60 portable eyetracking unit, placed below the bottom center of the laptop screen, gathered videorecordings of infrared illumination of corneal reflections at 60 times per eye per second. Before the assessment began, children were presented with five calibration dots against a solid background color. Calibration of the eyetracker was difficult for some children on the autism spectrum and for some children who were restless during testing (n=12). Gaze fixation duration was captured for the screen areas of interest.

Eyetracker data revealed that children tended to pay attention longest to the middle of the three responses, different from the left gaze preference found in past work with mainly neurotypical participants (Foulsham et al., 2013; Proietti et al., 2015). Moreover, children tended to gaze at the answer options in left-to-right order, and often at all three options, similar to how typically-developing children behave (Skibbe et al., 2018), indicating that the key and two distractors were working as intended to draw attention. Nevertheless, children tended to look longest at the answer they later chose and focused more on task content than other areas of the screen (e.g., white space). Further analyses will address how fixation patterns relate to children’s PA ability scores. The results highlight the utility of eyetracking to understand engagement of children with diverse skills, building validity evidence for early language assessments.

Authors

• Sarah Goodwin, Michigan State University
  Presenting Author

• Lori E Skibbe, Michigan State University
  Non-Presenting Author

• Jacob Richardson, Michigan State University
  Non-Presenting Author

• Ryan Bowles, Michigan State University
  Non-Presenting Author

• Gary A. Troia, Michigan State University
  Non-Presenting Author