|Title:||Assessing the Comprehension of Language in 2-Year-Olds Using Touch-Screen Technology|
|Principal Investigator:||Golinkoff, Roberta||Awardee:||University of Delaware|
|Program:||Early Intervention and Early Learning in Special Education [Program Details]|
|Award Period:||3 years (8/1/2016-7/31/2019)||Award Amount:||$1,599,998|
Co-Principal Investigators: Kathy Hirsh-Pasek (Temple University), Jill de Villiers (Smith College), and Aquiles Iglesias (University of Delaware)
Purpose: The purpose of this project is to develop a reliable and valid computer-based language assessment for children ages 24–36 months. Past research has demonstrated that early language skills are predictive of later language and academic skills. Therefore, early identification of children with language delays can lead to improvement of later outcomes. This project will be extending down in age a developed assessment, the Quick Interactive Language Screener or QUILS, from preschool children (3–5 years) to toddlers. The assessment will yield individual and group profiles in three areas of language — vocabulary, grammar, and process (strategies children use to learn language). Although the assessment will measure language development in all children, an important aim of the assessment will be to identify children with language delays who may be at risk for developing language impairment so that appropriate intervention can begin early.
Project Activities: In Year 1, item development will begin by generating twice the number of items needed on the final version, and pilot data will be collected through laboratory and field testing to help reduce the list of items. In Year 2, the item tryout phase will continue to field test the items with the goal of further reducing the length of the assessment to the desired number (40) for the final version. A sample of children will also be chosen this year to participate in a predictive validity study. In Year 3, the assessment will continue to be field tested, followed by additional analyses to finalize the items. In addition, test-retest reliability and convergent and predictive validity will be examined. The project will use Rasch and DIF analyses each year for item analyses.
Products: The products of this project include a fully developed, reliable, and valid assessment of language development for 24- to 36-month-old children, with the ability to help identify children with delays who are at risk for language impairment. Products will also include peer-reviewed publications and presentations.
Setting: The research will take place in laboratories and natural education settings (i.e., child care centers and homes, Early Head Start) in Delaware, Pennsylvania, and Massachusetts.
Sample: At least 880 children ages 24–36 months will participate in this research. Children with a large range of language abilities will be recruited, including those who are likely to be at risk for language impairment.
Assessment: The language assessment is based on the QUILS, which was developed with previous IES funding for preschool children. This project will extend the assessment to target children ages 24–36 months. This computer-based language assessment will use touch-screen technology to yield both individual and group profiles in three areas of language development. Two modules will measure language products (what the child knows) in the areas of vocabulary and grammar, and the third module will measure process (strategies children use to learn language, including both vocabulary and grammar learning). The assessment is designed for ease of use by the children (i.e., low response demands) and ease of administration by teachers, paraprofessionals, and other service providers. The final assessment is expected to have 40 items and take approximately 20 minutes to administer to a child.
Research Design and Methods: In Year 1, the research team will generate at least 80 items, more than double the expected final number, based on both the original QUILS and the creation of new items. These items will be pilot tested in a laboratory setting for content validity, with each child receiving only portions of the assessment in order to focus in depth on the items. Items from all three modules will then be tested in the field and analyzed with Rasch and DIF models, with the goal of reducing the number of items to 60. In Year 2, the researcher team will implement the item tryout phase in the field, with similar item analyses, to reduce the number of items to the best 40 items. In addition, the research team will identify the lowest performing children and a random selection of typically scoring children to participate in the predictive validity study. In Year 3, the research team will test the final assessment in the field. Different subsamples of children will take an additional assessment to determine test-retest reliability and convergent validity. Finally, the predictive validity study will determine whether a combination of the new QUILS and an existing language measure better predict later QUILS scores than either measure alone.
Control Condition: Due to the nature of the research design, there is no control condition.
Key Measures: The Peabody Picture Vocabulary Test (PPVT) and the MacArthur-Bates Communication Development Inventory (MCDI) will be used to examine the assessment's concurrent convergent validity, and the MDCI will also be used to examine predictive validity.
Data Analytic Strategy: For item analyses each year, the researchers will use Rasch analysis for each of the three modules and DIF analyses for gender and SES to assess item bias. Test-retest reliability and convergent validity will be examined using correlational analyses and predictive validity will be examined using regression analysis.
Related IES Project: Using Developmental Science to Create a Computerized Preschool Language Assessment (R305A110284)