5.1 Assessment of Child Language Disorders
Key Takeaways
- Developmental Language Disorder (DLD) presents as a persistent impairment in language comprehension and/or production across form, content, and use, occurring without associated biomedical conditions like intellectual disability or sensorineural hearing loss.
- Psychometric evaluation requires analyzing Standard Error of Measurement (SEM) and 90–95% confidence intervals; diagnostic cutoffs typically fall between -1.25 and -2.0 standard deviations (standard score ≤ 81–77) below the mean.
- Diagnostic sensitivity and specificity must meet or exceed 80% (ideally ≥ 90%) to minimize false positives and false negatives during norm-referenced testing.
- Language Sample Analysis (LSA) offers criterion-referenced granularity, measuring Mean Length of Utterance in morphemes (MLU-m), Type-Token Ratio (TTR; normative baseline 0.42–0.50), and syntactic error rates.
- Dynamic Assessment (Test-Teach-Retest framework) evaluates learning potential and modifiability, serving as the gold standard for differentiating true language disorders from language differences in culturally and linguistically diverse (CLD) populations.
Assessment of Child Language Disorders
A comprehensive assessment of child language disorders requires a multidimensional evaluation of form (phonology, morphology, syntax), content (semantics), and use (pragmatics) across both receptive and expressive modalities. Speech-language pathologists (SLPs) must synthesize evidence from norm-referenced standardized tests, criterion-referenced measures, language sample analyses, dynamic assessment paradigms, and observational data to establish diagnostic eligibility and formulate evidence-based intervention plans.
1. Clinical Taxonomy and Differential Diagnosis
Developmental Language Disorder (DLD) vs. Secondary Language Impairments
The term Developmental Language Disorder (DLD)—formerly referred to as Specific Language Impairment (SLI)—describes children who present with persistent language difficulties that significantly impact daily communication, academic attainment, and social interaction, in the absence of differentiating biomedical conditions such as:
- Intellectual Disability (ID): Defined by nonverbal IQ scores < 70 alongside adaptive behavior deficits.
- Sensorineural Hearing Loss (SNHL): Auditory acuity thresholds worse than 20 dB HL across speech frequencies (500–4000 Hz).
- Autism Spectrum Disorder (ASD): Primary impairments in social-emotional reciprocity, nonverbal communication, and restricted/repetitive behavior patterns.
- Acquired Brain Injury (ABI) / Structural Neurological Lesions.
Children with DLD consistently struggle with grammatical morphology, particularly tense and agreement markers. Hallmark clinical markers in English include omissions of:
- Third-person singular -s (e.g., "He run fast")
- Past tense -ed (e.g., "Yesterday we walk home")
- Auxiliary and copula forms of be (e.g., "She running" or "They happy")
Language Difference vs. Language Disorder
A language difference stems from normal variations secondary to cultural background, regional dialect (e.g., African American English [AAE], Southern White English [SWE]), or second-language acquisition processes (e.g., dual language exposure, sequential bilingualism). A language disorder is diagnosed only when the child exhibits impairment in ALL languages spoken.
2. Psychometric Principles of Norm-Referenced Assessment
Norm-referenced tests compare a child's performance against a representative, nationally stratified normative sample. SLPs must critically evaluate test manuals for sound psychometric properties before clinical administration.
| Psychometric Metric | Standard Definition | Clinical Target / Cutoff Threshold |
|---|---|---|
| Mean & Standard Deviation | Distribution baseline centered at Mean = 100, SD = ±15 (or Scaled Score Mean = 10, SD = ±3). | Clinical cutoff typically set at -1.25 to -2.0 SDs (Standard Score ≤ 81 to 77; Percentile ≤ 10th to 3rd). |
| Standard Error of Measurement (SEM) | Estimate of test score error variance based on reliability coefficient. | Used to calculate Confidence Intervals (CI) (e.g., 90% or 95% CI) around the observed score. |
| Sensitivity | True Positive Rate: Percentage of children with a language disorder correctly identified by the test. | Acceptable: ≥ 80% (0.80); Good/Excellent: ≥ 90% (0.90). |
| Specificity | True Negative Rate: Percentage of typically developing children correctly identified as normal. | Acceptable: ≥ 80% (0.80); Good/Excellent: ≥ 90% (0.90). |
| Construct & Content Validity | Degree to which the instrument measures the theoretical construct of language abilities. | Requires empirical factor analysis and expert item review. |
Clinical Warning on Score Interpretation: Relying solely on age-equivalent or grade-equivalent scores is psychometrically invalid and clinically misleading. Age equivalents do not account for score variance, lack equal intervals, and misrepresent true performance distance from the normative mean. SLPs must report standard scores, percentile ranks, and confidence intervals.
3. Criterion-Referenced Evaluation & Language Sample Analysis (LSA)
While norm-referenced tests identify the presence of a delay, Language Sample Analysis (LSA) provides ecological validity and detailed profiles of functional communication in naturalistic contexts.
Quantitative Language Sample Metrics
- Mean Length of Utterance (MLU):
- Calculated by dividing the total number of morphemes (MLU-m) or words (MLU-w) by the total number of fully intelligible utterances (minimum 50–100 utterances).
- Serves as a reliable index of syntactic complexity up to MLU 4.0 (typically ages 18 to 48 months).
- Type-Token Ratio (TTR):
- Measures lexical diversity by dividing the number of different words (NDW) by the total number of words (TNW).
- For children aged 2;0 to 8;0, normative TTR values consistently hover between 0.42 and 0.50. A TTR < 0.42 signifies limited semantic range and word-finding difficulty.
- Grammatical Error Rate & Complex Sentence Index:
- Quantifies the percentage of utterances containing morphosyntactic errors and evaluates the density of subordinate, relative, and infinitive clauses.
4. Dynamic Assessment & Multicultural Assessment Frameworks
When assessing culturally and linguistically diverse (CLD) children or dual-language learners, standardized tests routinely yield high false-positive rates due to content, construct, and linguistic bias.
The Test-Teach-Retest Paradigm
- Pre-Test: Assess baseline performance on a language task (e.g., novel word learning, story retelling).
- Mediated Learning Experience (MLE): Provide short-term, targeted intervention focused on teaching learning strategies, metacognitive awareness, and self-monitoring.
- Post-Test: Re-assess performance and evaluate child modifiability (how easily the child learns and transfers newly taught skills).
Diagnostic Decision Rule:
- High Modifiability / Rapid Learning: Indicates a language difference (lack of prior exposure, not an intrinsic deficit).
- Low Modifiability / Resistance to Prompting: Indicates a true Developmental Language Disorder (DLD).
5. Clinical Decision-Making & Differential Diagnosis Scenario
Case Presentation: A 4-year, 6-month-old child is referred for an evaluation due to teacher concerns regarding low intelligibility and difficulty following directions. Standardized administration of a commercial language battery yields a Core Language Standard Score of 74 (95% CI: 69–79; 4th percentile).
Diagnostic Workflow
- Rule Out Sensory/Medical Factors: Confirm passing results on a 20 dB HL bilateral pure-tone hearing screening and oral-peripheral structural exam.
- LSA Verification: Analyze a 100-utterance conversational sample. The child demonstrates an MLU-m of 2.8 (expected for age: 4.5+), with persistent omissions of tense/agreement markers ("Yesterday he drop book", "She like ice cream"), yielding an error-free utterance rate of only 22%.
- Pragmatic Differential: Assessment reveals intact joint attention, communicative intent, eye contact, and turn-taking, ruling out ASD.
- Diagnostic Conclusion: Primary Developmental Language Disorder (DLD) involving expressive morphosyntax and receptive sentence comprehension.
A speech-language pathologist administers a standardized language test to a 6-year-old child. The child receives an overall standard score of 78 (Mean = 100, SD = 15) with a Standard Error of Measurement (SEM) of 3 points. What is the 95% confidence interval for this child's true score?
An SLP conducts a dynamic assessment with a sequential bilingual 5-year-old child who performed below average on a standardized norm-referenced test. During the Mediated Learning Experience (MLE), the child demonstrates rapid strategy acquisition, high flexibility, and significant performance improvement on the post-test. Which diagnostic conclusion is most appropriate?
During a 100-utterance language sample analysis of a 4-year-old child, the SLP counts a total of 400 words (TNW) and 180 different words (NDW). What is the child's Type-Token Ratio (TTR), and how should it be clinically interpreted?
Which psychometric property of a standardized diagnostic test measures the test's ability to correctly classify children who DO NOT have a language disorder as typically developing?