Project: Play-based Narrative Elicitation

The aim of this research is to create developmentally appropriate, play-based storytelling elicitation procedures to collect language samples of young children aged 18-48 months, tools for evaluating the magnitude and quality of narrative language produced in play-based storytelling sessions, and examine the psychometric properties of these new tools. This research project will employ descriptive and correlational research designs using a sample of 200 children and 200 adults (parents, caregivers, educators, etc.). The reliability and validity of a researcher-developed set of storytelling elicitation materials and procedures, as well as scoring procedures, will be examined through field tests. A research assistant (RA) will conduct the research activities in a child's natural environment, eliciting stories from children for up to four different time points (e.g., Weeks 1, 2, 12 and 24). During story elicitations, language samples will be audio recorded using a hand-held recording device or via a device that attaches to the child's worn clothing. Each audio record file will be transcribed at a later date by a RA and analyzed using a narrative scoring rubric. Two independent scorers will score 30% of the samples to document the reliability of scoring. The transcribed information will allow for correlational research to assess the elicitation and scoring procedures’ sensitivity to narrative language development over time, as well as allow within-group reliability and validity comparisons. The researchers are interested in the play-based storytelling elicitation and scoring procedures’ feasibility and validity of capturing naturally occurring oral narrative language during play-based activities. Covalidation of the measure developed against existing language assessments, such as the ECI (Juniper Gardens Children’s Project, 2009) and PLS-5 (Preschool Language Scales-5th Ed.; Zimmerman, Steiner, & Pond, 2011), will support the goal of ensuring that the procedures being developed is aligned with constructs used in other assessments of narrative language in young children.
Test-retest reliability will be evaluated by comparing data collected at Weeks 1 and 2 (inter-response time of 1 week), and predictive validity outcomes will be analyzed by comparing the remaining repeated measures collected again Weeks 12 and 24 with those collected in Week 2. A two-level hierarchical growth curve model with repeated measures nested within individual child participants will be used to analyze individual child participant growth patterns, as well as for children grouped in cohorts by age. The strength of findings, interpretation of measurement results, and significance of effect will assess the criterion validity of the storytelling measurement tool for children aged 18-48 months. Quantitative statistical analyses using structural equation modeling will be conducted to explore which components can be reliably measured, through an exploratory factor analysis will provide preliminary support for the inclusion of story grammar, sequence, generative versus familiar story content, mean length of utterance, choice measures, and conversational turns using analysis of narrative language samples collected during test conditions. This study will be the first to examine the validity and reliability of a set of play-based storytelling elicitation and scoring procedures. We expect to document adequate validity, reliability, and feasibility of the procedures. If acceptable validity and reliability are found, the play-based storytelling elicitation and scoring procedures will be used in subsequent intervention studies to determine the effect of language promotion interventions implemented in young children’s homes. If the procedures do not have adequate validity and reliability, revisions to the elicitation and scoring procedures will be made and additional research will be completed.