Current Project: Tallying Reference Errors In Narrative (TREIN)

Tallying Reference Errors in Narrative (TREIN): A tool for assessing expressive discourse capacity in children. TREIN is currently underdevelopment by Dr. John C. Thorne.

The ability to create effective expressive discourse enables children to be successful at many social and academic tasks on a daily basis. The ability to produce grammatically correct sentences is a skill that will support effective expressive discourse, but it is not sufficient for managing the efficient flow of information to meet the processing needs of listeners in many discourse situations. Tallying Reference Errors in Narrative (TREIN) is an analysis protocol designed to quantify discourse level behaviors that help or hinder efficient information flow in narrative discourse by children. Two central constructs are quantified by a TREIN analysis: 1) how effectively children manage the introduction of concepts into a narrative, and 2) how effectively they maintain reference ties (Halliday & Hasan, 1976) to those concepts as the narrative develops. To understand the TREIN, we will first examine the introduction of concepts into a narrative and how to code strategies commonly used by children, and then follow this with a discussion of how reference ties to concepts are achieved and coded.

In English, concepts can be introduced into a narrative in a limited number of ways which will vary according to the discourse context in which the narrative is embedded. Likewise, there are a limited number of strategies in English for making reference ties to concepts that have been introduced into a narrative that will also vary according to the discourse context.  The TREIN protocol utilizes a very carefully specified discourse context in order make the impact of certain introduction and reference strategies predictable. This carefully controlled discourse context allows clinicians and researchers using the TREIN to distinguish between appropriate and inappropriate strategies when they occur in the narratives of children. The pattern of use of appropriate and inappropriate discourse strategies as defined in the TREIN protocol, therefore, becomes a proxy measure of a child’s capacity for producing effective and efficient expressive discourse that can be considered separately from their ability to produce grammatical sentences.

The TREIN narrative Context

The TREIN protocol utilizes a story generation task (frogwayinstructionsmodified_protocol) to generate a narrative discourse for analysis. The task has the child look through a series of wordless story pictures, and then the child uses those pictures as prompts while they tell a narrative to a listener that is (presumably) unfamiliar with the pictures and unable to see them while the narrative unfolds.  This discourse context emphasizes the need for the narrator to carefully track the difference between their own knowledge of the story (provided by their access to the story pictures), and that of their listener who only has the verbal narrative produced so far with which to form an understanding of narrative concepts. It requires the narrator to avoid introduction and reference strategies that would only be appropriate in a context where both narrator and listener could see the story pictures (e.g., pointing to pictured concepts). By not allowing both the listener and narrator to have the same degree of access to the story pictures, this task limits appropriate strategies for introduction and reference to concepts to the spoken modality, making it unnecessary to quantify non-linguistic strategies in the discourse.

To lean more about Dr. Thorne’s work with TREIN, visit his website!


Parenting in Prison: Evaluating an Intervention for Incarcerated Mothers and their Infants

It all started with an article.  Kate Krings, MS CCC-SLP, Lecturer in the UW Department of Speech and Hearing Science, was reading The Atlantic one night in the Summer of 2015.  Written by Sarah Yager, the article tells the story of prison nursery programs in the United States.  Upon learning that one of these programs was one within 50 miles of UW’s campus, she said to herself, “I’m going. And I’m taking my students with me.”

Kate reached out to Felice Davis, Associate Superintendent at the Washington Correctional Center for Women (WCCW), to offer programs and workshops to incarcerated mothers and their babies at the Residential Parenting Program. These workshops utilized the Hanen Centre’s It Takes Two to Talk® and provided preventative strategies to bolster the communication development of the children through high quality parent-child interactions. Kate pursued Hanen certification and got permission from Elaine Weitzman, director of the Hanen Centre, to adapt this program for this unique population and environment.

As word spread about this program, Carol Hoerster, a UW SPHSC alum and former colleague of Kate’s at Seattle Public Schools, introduced Kate to Dr. Jean Kelly, founder and director of Promoting First Relationships, an evidence-based program promoting responsiveness and attachment in parent-child relationships.

In January 2016, Dr. Amy Pace joined the faculty in the Department of Speech and Hearing Sciences, and a research-practice partnership was formed. Our innovative collaboration between researchers, educators, and clinicians significantly increases the odds of positive maternal and child outcomes in a challenging and complex prison environment.

Our mission is to positively impact the well-being of young children born in prison through empowerment and education of their incarcerated mothers.  We support and nurture caregiver-mother-child interactions to establish a strong communication foundation for language and social-emotional development.  We investigate and emphasize the importance of evidence-based methods for intervention with incarcerated mothers.  We are committed to serving a highly vulnerable and marginalized population with compassion, respect, and scientifically-based education programs.


Another aspect of this project is the impact of service learning on graduate student clinicians’ experience. Service learning opportunities have been shown to provide deeper understanding of classroom learning and increase empowerment to create change in a community. We are examining if this experience provides graduate student clinicians with this type of in-depth learning and informs their future clinical practice. Our 2016-2017 team looks froward to their visit to WCCW on November 9th, 2016!

Preschool Language Assessment

Language ability at kindergarten entry is one of the strongest predictors of literacy development and academic achievement. For this reason, early identification of language delay or impairment is critical. Unfortunately, many existing standardized tests are time-consuming and cannot be easily administered by teachers, parents, or paraprofessionals.

The purpose of this project is to test a new, computer-administered language assessment – the Quick Interactive Language Screener (QUILS) – for use with preschool children ages 3 through 5.  The QUILS was developed by Roberta Golinkoff and Aquiles Iglesias (University of Delaware), Kathy Hirsh-Pasek (Temple University), Jill de Villiers (Smith College), and Mary Sweig Wilson (Laureate Learning Systems).

The UW Child Language Lab will be working with the University of Washington Department of  Speech and Hearing Science and the Haring Center Experimental Education Unit to determine how accurately the QUILS identifies preschool children’s vocabulary, syntax, and language processing skills. Specifically, we want to look at how the QUILS measures the language of preschool children that have been diagnosed with, or may be at risk for, language delays.

The Child Language Team is currently recruiting preschool children with language impairments for this study.  Please contact us at if you are interested in learning more!