Abstract
Prediction of kindergarten and preschool children’s alphabetic knowledge was examined with respect to child characteristics (e.g., age, SES, emergent literacy skills), letter properties (e.g., alphabetic order, letter name structure, sound properties), and child x letter factors (e.g., letters used to spell a certain child’s name). A single, multilevel model including all types of factors was created, allowing for between-item, between-subject, and subject x item analyses. All zero-order and unique predictors of children’s letter name and letter-sound knowledge were identified. Factors of each type were reliably predictive of outcomes, although most of the model variance was attributed to differences among children. Implications of using the new modeling technique as well as practical application for early childhood classrooms are discussed.
|
