Technical Features of Sentence-Level Curriculum-Based Measures and Language Sample Analysis for Students With Writing...

Area of Science:

• * Educational Psychology
• * Linguistics
• * Special Education

Background:

• * Oral language proficiency is crucial for developing strong writing skills.
• * Existing measurement tools for oral and written language have limitations.
• * Language sample analysis (LSA) offers potential for more nuanced assessment.

Purpose of the Study:

• * To evaluate complementary scoring metrics in LSA for the Picture Word Writing Curriculum-Based Measure (CBM-W).
• * To investigate the reliability, validity, and sensitivity to growth of LSA metrics.
• * To determine if LSA can provide more specific oral language skill estimates.

Main Methods:

• * Utilized language samples from 123 students with writing difficulties using the Picture Word CBM-W.
• * Assessed alternate form reliability and criterion-related validity against existing metrics and a standardized measure.
• * Examined sensitivity to growth from fall to spring for LSA and Picture Word CBM-W scoring mechanisms.

Main Results:

• * Two LSA metrics demonstrated technical quality and sensitivity to growth: mean length of T-Unit in morphemes (MLTU-M) and number of different words (NDW).
• * MLTU-M showed reliability when using the mean of two forms in the fall.
• * NDW demonstrated reliability and sensitivity to growth using the mean of two forms in fall and spring.

Conclusions:

• * Results support the integral role of oral language in sentence-level writing proficiency.
• * MLTU-M and NDW are effective complementary scoring mechanisms for Picture Word CBM-W.
• * These LSA metrics offer more specific insights into grammar/morphosyntax and semantics/lexical diversity than current CBM-W scoring.