Seeking a valid gold standard for an innovative, dialect-neutral language test.

Area of Science:

• Speech and Language Pathology
• Developmental Psychology
• Linguistic Anthropology

Background:

• Accurate language assessment is crucial for identifying developmental language impairments in children.
• Existing language assessments may not adequately account for linguistic variations in diverse populations.
• The Diagnostic Evaluation of Language Variation-Norm Referenced (DELV-NR) was developed as a dialect-neutral tool.

Purpose of the Study:

• To validate the DELV-NR using alternative gold standards.
• To compare the diagnostic accuracy of the DELV-NR against prior clinical diagnoses.
• To refine estimates of the DELV-NR's diagnostic accuracy.

Main Methods:

• The study involved 78 African American children (ages 5-6), including 20 with diagnosed language impairments.
• The DELV-NR and concurrent language samples (LSs) were administered.
• LS profiles served as a primary gold standard for comparison.

Main Results:

• The DELV-NR demonstrated superior diagnostic accuracy compared to prior diagnoses when validated against LS profiles.
• While specificity was adequate, the DELV-NR's initial sensitivity was somewhat low.
• Adjusted analyses, incorporating predictive power of prior diagnoses, yielded high sensitivity (.85) and specificity (.93).

Conclusions:

• The DELV-NR proves to be an improved diagnostic tool for language assessment in this population.
• Even with imperfect gold standards, the DELV-NR supports accurate clinical decision-making.
• The study highlights the importance of dialect-neutral assessments for equitable child language evaluation.