Screening, Assessment, Evaluation, and Diagnosis
Screening versus assessment
Screening is brief and cheap; its job is to flag who needs a closer look, and it is judged by sensitivity (catching true cases), specificity (clearing true negatives), and predictive value, which depends heavily on base rate. False negatives miss children; false positives cost time. Child find obliges districts to locate children suspected of disabilities.
The evaluation toolkit
A comprehensive evaluation is multi-source: case history, oral mechanism examination, hearing screening, norm-referenced standardized tests, criterion-referenced and curriculum-based measures, language sampling (MLU in morphemes, type–token ratio, narrative macrostructure and microstructure), speech measures (percent consonants correct, phonological process use, intelligibility, stimulability), dynamic assessment, and functional observation.
Reading scores
Standard scores have a mean of 100 and SD of 15: 85 is −1 SD (about the 16th percentile) and 70 is −2 SD (about the 2nd percentile). Confidence intervals reflect measurement error; if a 95% CI overlaps the eligibility cutoff, the score alone cannot decide the case. Age- and grade-equivalents are not valid for eligibility decisions. Tests must be chosen and interpreted with cultural and linguistic fairness: assess multilingual learners in the native language and English as appropriate, use trained interpreters, and use dynamic assessment to separate difference from disorder. Dialect features are not errors.
Diagnosis and eligibility
Eligibility is an IEP-team decision requiring a disability category, adverse educational impact, and need for specially designed instruction. A medical diagnosis is neither necessary nor sufficient.
Traps
Letting one low score confirm a disorder; ignoring the SEM; calling dialect a disorder; treating an age-equivalent as mental age; assuming an IQ–achievement discrepancy is the only route to a language impairment label.
Worked example
A 7-year-old earns SS 78 (95% CI 74–82) on a language test with a cutoff of 85. Because the whole interval falls below the cutoff and language samples agree, the score supports - but does not by itself establish - eligibility; educational impact must still be documented.