Amsterdam-Nijmegen Everyday Language Test

Abstract

The Amsterdam-Nijmegen Everyday Language Test (widely known by its Dutch acronym ANTAT, Amsterdam – Nijmegen Test voor Alledaagse Taalvaardigheid, and in international literature as the ANELT) is an ecologically valid, standardized psychometric and clinical assessment instrument engineered to evaluate functional verbal communication in individuals with aphasia secondary to cerebrovascular accident (CVA) or traumatic brain injury (TBI). Developed by Leo Blomert, Corry H. Koster, and Mary-Louise Kean in 1995, the instrument addresses a critical diagnostic gap left by conventional neurorehabilitation batteries, which frequently quantify discrete linguistic impairments (e.g., phonology, syntax, lexical retrieval) in artificial laboratory conditions rather than pragmatic effectiveness in authentic social discourse. The instrument consists of two equivalent, parallel forms (Version A and Version B), each comprising 10 open-ended everyday scenarios preceded by two standardized practice trials. Patients are presented with standardized situational prompts (e.g., speaking to a physician, shopping, resolving a neighborhood dispute) and instructed to respond in direct speech. Responses are evaluated along two distinct five-point rating scales: Scale A (Verbal Communicative Adequacy), which appraises the successful transmission of target communicative content regardless of surface-level linguistic errors; and Scale B (Clarity of Vocal Expression), which quantifies speech intelligibility and acoustic-phonetic execution. Psychometric validation studies demonstrate exceptional inter-rater reliability ($r > .90$; Intraclass Correlation Coefficients exceeding $.93$), high test-retest reliability ($r > .92$), robust parallel-form equivalence ($r = .94$), and high internal consistency ($lpha > .90$). Non-parametric Item Response Theory (Mokken scale analysis) and factor analytic inquiries confirm the construct unidimensionality of the communicative adequacy scale. The test serves as a sensitive measure of longitudinal therapeutic gains and social-communicative rehabilitation outcomes.

Keywords

Amsterdam-Nijmegen Everyday Language Test, ANTAT, ANELT, Aphasia Assessment, Functional Communication, Pragmatics, Speech-Language Pathology, Stroke Rehabilitation, Verbal Communicative Adequacy, Neurogenic Communication Disorders

Authors

The test was conceptualized, normed, and psychometrically validated through a research collaboration between cognitive neuroscientists, clinical aphasiologists, and psycholinguists in the Netherlands and the United States:

• Leo Blomert, Ph.D. (1955–2012): Formerly Professor of Cognitive Neuroscience and Psycholinguistics at Maastricht University and the Maastricht Brain Imaging Center (M-BIC), Netherlands. Dr. Blomert was an international authority on language acquisition, developmental and acquired dyslexia, and the neural substrates of functional communication in post-stroke aphasia.
• Corry H. Koster, M.Sc.: Senior clinical aphasiologist and speech-language researcher affiliated with the Department of Rehabilitation and Neurological Sciences at Radboud University Medical Center (Nijmegen) and collaborative aphasia rehabilitation centers in the Netherlands. Her clinical expertise spearheaded the ecological scenario selection and scoring taxonomy.
• Mary-Louise Kean, Ph.D.: Professor Emerita of Cognitive Sciences and Linguistics at the University of California, Irvine (UCI), United States. Dr. Kean contributed theoretical frameworks regarding agrammatism, theoretical linguistics, and psycholinguistic modeling of language breakdown following neurological trauma.

Purpose

The fundamental clinical and psychometric objective of the Amsterdam-Nijmegen Everyday Language Test is to measure the level and change in verbal communicative ability in everyday social interactions in patients exhibiting expressive language impairments due to acquired brain pathology. Historically, standardized aphasiology instruments—such as the Boston Diagnostic Aphasia Examination (BDAE) or the Aachen Aphasia Test (AAT)—were engineered primarily to localize cerebral lesions and delineate microscopic linguistic deficits across specific modalities (e.g., confrontation naming, auditory sentence comprehension, repetition, written spelling). Although clinically indispensable for neurological categorization, high impairment scores on formal psycholinguistic tasks do not systematically correlate with an individual’s operational success in daily life. For instance, a patient with moderate Broca’s aphasia may fail an isolated syntactic parsing task yet manage to order medication or report an emergency through non-canonical, compensatory lexical assemblies.

Conversely, patients displaying fluent anomic or Wernicke’s aphasia might achieve moderate scores on basic structured tests yet fail to convey essential semantic propositions during conversational exchanges because of perseveration, circumlocution, or empty speech. The ANELT directly isolates and quantifies this functional communicative competence. By placing the individual inside a simulated verbal scenario requiring direct quotation (e.g., “You are at the bakery and want to buy half a loaf of whole wheat bread. What do you say?”), the instrument evaluates whether the patient can formulate and articulate a message that is intelligible and pragmatically successful to an interlocutor.

The primary clinical applications include:

• Establishing a normative communicative baseline during subacute and chronic phases of neurorehabilitation following cerebral ischemia, hemorrhage, or cranial trauma.
• Tracking longitudinal recovery profiles and measuring the real-world ecological efficacy of speech-language therapies (e.g., Constraint-Induced Aphasia Therapy [CIAT], Semantic Feature Analysis [SFA], or Pragmatic Communication Therapy).
• Assisting multi-disciplinary neurorehabilitation teams, patients, and caregivers in managing discharge planning by yielding an objective metric of independent social communication.
• Serving as a primary or secondary functional outcome measure in pharmacological, neurostimulation (e.g., repetitive Transcranial Magnetic Stimulation, transcranial direct current stimulation), and linguistic clinical trials.

Psychological Construct

The instrument is grounded in the construct of functional verbal communication, conceptualized as the capacity to convey relevant semantic information through oral expression such that an interlocutor accurately decodes the speaker’s illocutionary intent, irrespective of the formal grammatical correctness, phonological precision, or lexical complexity of the output. In the ANELT, this overarching construct is operationalized through two orthogonal yet functionally interactive dimensions:

Dimension 1: Verbal Communicative Adequacy (Scale A)

Scale A reflects the pragmatic, informational, and communicative success of the utterance. It evaluates whether the critical communicative core (the semantic kernel necessary to solve the everyday dilemma) has been conveyed to the listener. Crucially, the rating system separates linguistic form from communicative function. An utterance marked by severe agrammatic telegraphic syntax (e.g., “Car… stop… garage… help” in response to an automobile breakdown) communicates the essential proposition successfully, earning a high adequacy score (Score 4 or 5), because a native interlocutor would comprehend the speaker’s need. Conversely, a structurally fluent, grammatically pristine utterance that fails to supply the requested information (e.g., “Well, you know, things happen like that when the day comes around, but it is what it is”) receives a low adequacy score (Score 1 or 2) because no informative semantic content is delivered. The scale is scored from 1 (not understandable / no relevant information conveyed) to 5 (fully understandable / completely relevant information conveyed without ambiguity).

Dimension 2: Clarity of Vocal Expression (Scale B)

Scale B captures the motoric, acoustic, and phonological intelligibility of the utterance. This dimension reflects how easily an interlocutor can recognize the physical speech tokens produced by the patient, assessing the degradation introduced by dysarthria, apraxia of speech, or phonemic paraphasias. Rated on an equivalent five-point continuum from 1 (completely unintelligible) to 5 (completely clear, effortless acoustic comprehension), Scale B prevents motor speech disturbances from contaminating the pure informational assessment of Scale A. While an individual may achieve complete functional adequacy by selecting robust communicative strategies, their articulation may remain dysarthric; conversely, an individual with jargon aphasia may display crystalline articulatory clarity with zero communicative adequacy.

Theoretical Framework

The architecture of the ANELT rests upon three theoretical paradigms: modern functional aphasiology, Speech Act Theory, and the International Classification of Functioning, Disability and Health (ICF) formulated by the World Health Organization.

1. Speech Act Theory and Pragmatics

Grounded in the linguistic philosophy of John L. Austin (1962) and John Searle (1969), Speech Act Theory posits that language is not merely a descriptive formal system but an instrument of human action. Utterances execute specific acts categorized into locutionary acts (the phonological and syntactic production of words), illocutionary acts (the real communicative intention or force behind the utterance, such as requesting, complaining, or warning), and perlocutionary effects (the psychological or behavioral outcome achieved on the listener). Traditional neuropsychological testing measures the locutionary layer (syntax, lexicon, repetition). In contrast, the ANELT targets the illocutionary force: can the patient produce a speech act capable of eliciting the appropriate perlocutionary response from the listener? Furthermore, H. Paul Grice’s (1975) Cooperative Principle and conversational maxims—specifically the Maxims of Quantity (be as informative as required) and Relation (be relevant)—directly govern the scoring rubrics for Scale A, penalizing off-target circumlocutions while crediting concise, goal-oriented linguistic formulations.

2. The ICF Neurorehabilitation Paradigm

Within the WHO ICF framework, neurological conditions induce consequences across three distinct tiers:

1. Body Functions and Structures (Impairment): Structural lesions of the left perisylvian language network causing anomia, agrammatism, or phonological assembly failure.
2. Activities (Limitation): The restriction in executing purposeful tasks, such as formulating a phone call or conveying a retail transaction request.
3. Participation (Restriction): The exclusion from social roles, interpersonal relationships, and community integration.

The ANELT was specifically developed to measure the Activity tier. Blomert and colleagues argued that an assessment tool must measure whether the patient can utilize whatever neurofunctional linguistic resources remain available to resolve real-world dilemmas, thereby mediating the gap between neurobiological lesion severity and social participation.

Validity

Extensive psychometric investigations have established the construct, criterion, convergent, and discriminant validity of the ANELT in diverse international clinical cohorts.

Content and Face Validity

The 20 scenarios comprising Versions A and B were extracted from an initial corpus of over 70 real-world situations identified by clinical panels of speech-language therapists, patients with chronic aphasia, and their healthy relatives. Scenarios were systematically evaluated for ecological representative frequency, linguistic naturalness, emotional neutrality, and cultural cross-applicability across socioeconomic strata. Low-frequency or culturally idiosyncratic situations were pruned, yielding 20 matched items characterized by high situational typicality.

Convergent and Criterion Validity

During its initial Dutch and subsequent international psychometric evaluations, the ANELT demonstrated moderate-to-high correlations with recognized neuropsychological aphasia batteries, confirming both convergence on language competence and independence as a functional measure:

• Aachen Aphasia Test (AAT): Blomert et al. (1994) documented strong correlations between the ANELT Scale A total score and the AAT Spontaneous Speech Subscales: Communicative Success ($r = .86, p < .001$) and Syntactic Structure ($r = .79, p < .001$). Correlations with isolated impairment subtests (such as the AAT Token Test, $r = -.71$, and AAT Confrontation Naming, $r = .74$) demonstrated that while basic receptive and expressive abilities facilitate functional verbal output, functional communicative performance diverges meaningfully from isolated test performance.
• Boston Diagnostic Aphasia Examination (BDAE): Correlation with the BDAE Aphasia Severity Rating Scale was reported at $r = .81$ to $.88$, while correlation with the BDAE Complex Ideational Material subtest reached $r = .68$.
• Functional Independence Measures: When correlated with the Communicative Effectiveness Survey (CES) and the Functional Communication Profile (FCP), the ANELT Scale A demonstrated strong convergence ($r = .78$ to $.84$), confirming that the structured role-play method accurately models communicative effectiveness in home and community settings.

Discriminant and Known-Groups Validity

The instrument clearly separates healthy control participants from brain-injured clinical groups. Healthy older adults consistently hit near-ceiling performance (mean Scale A scores between 49.2 and 50.0 across versions, with variance approaching zero). Furthermore, the scale distinguishes between clinical severity grades of aphasia classified via standardized syndromic algorithms (Global < Wernicke < Broca < Conduction < Anomic), demonstrating strong known-groups discriminant validity.

Reliability

The ANELT demonstrates high measurement precision across internal, temporal, and inter-examiner psychometric metrics.

Internal Consistency

Internal consistency analyses conducted across both normative developmental samples and clinical cohorts of chronic stroke survivors demonstrate high homogeneity:

• Cronbach’s Alpha ($lpha$): For Scale A, Cronbach’s alpha ranges from $.91$ to $.96$ for Version A, and from $.92$ to $.95$ for Version B, indicating exceptional item coherence without redundancy.
• Split-Half Reliability: Spearman-Brown corrected split-half coefficients consistently exceed $.90$ for both functional adequacy and vocal clarity.

Inter-Rater and Intra-Rater Reliability

Because the ANELT depends upon clinician scoring of open-ended conversational replies, establishing rater objectivity was an essential component of its standardization:

• Inter-Rater Reliability: Inter-rater agreement between independent trained speech-language pathologists evaluating audiotaped or videotaped test administrations yielded Pearson correlation coefficients ranging from $r = .94$ to $.98$. Cohen’s weighted kappa ($\kappa_w$) values across individual scenario adequacy ratings range from $.79$ to $.91$, denoting high inter-observer reliability.
• Intra-Rater Reliability: Clinicians re-scoring identical taped sessions after a 3-month washout period displayed intra-rater stability ranging from $r = .96$ to $.99$.

Parallel-Form and Test-Retest Stability

The strict psycholinguistic matching of Version A and Version B was validated by administering both forms in a counterbalanced design to stable chronic aphasia cohorts:

• Parallel-Form Equivalence: Pearson product-moment correlation between Version A and Version B exceeded $r = .94$ ($p < .001$), with no statistically significant differences between mean group scores ($t$-test, $p > .30$). This equivalence enables repeated longitudinal measurements without re-test inflation or item memorization artifacts.
• Test-Retest Stability: In untreated, medically stable chronic patients tested across a 4- to 6-week interval, test-retest reliability was documented at $r = .92$ ($p < .001$), verifying high temporal stability.

Factor Analysis

The dimensional structure and scaling properties of the ANELT have been investigated using classical factor analysis and modern Item Response Theory (IRT).

Exploratory and Confirmatory Factor Analysis

Principal Component Analysis (PCA) and Confirmatory Factor Analysis (CFA) conducted on the 10 items of Scale A reveal an unequivocal unidimensional construct for functional communicative adequacy. The primary unrotated factor accounts for over $68%$ to $74%$ of the total item variance across clinical cohorts, with an eigenvalue greater than $6.8$. Subsequent factors consistently display eigenvalues well beneath $1.0$, refuting multidimensionality within Scale A. Standardized factor loadings across all 10 items are exceptionally uniform, ranging between $.72$ and $.89$ (e.g., Scenario 1 loading $= .81$; Scenario 5 loading $= .86$; Scenario 7 loading $= .84$), demonstrating that each scenario functions as an equivalent indicator of the latent functional communication trait.

Item Response Theory & Mokken Scale Analysis

Blomert and colleagues subjected the ANELT to non-parametric Item Response Theory using Mokken scale analysis for polychotomous items. The analysis demonstrated high scalability coefficients:

• Loevinger’s $H$ Coefficient: The overall scalability coefficient $H$ for Scale A ranged between $.62$ and $.71$ across validation studies. In Mokken scaling, an $H$ value above $.50$ indicates a strong, highly scalable unidimensional metric.
• Item Scalability ($H_i$): Individual item scalability coefficients ($H_i$) were uniformly distributed between $.58$ and $.74$, confirming that the scenarios conform to a cumulative Guttman-like hierarchy where individuals possessing lower communicative ability successfully pass easier items, while only patients with higher communicative competence succeed across demanding scenarios.
• Rasch Rating Scale Modeling: Parametric Rasch modeling confirmed acceptable item fit statistics (Infit and Outfit mean square values between $0.78$ and $1.18$), demonstrating absence of severe item bias or multidimensional distortion across patient age, gender, or etiology.

Instrument / Measurement Tool

• Test Type: Performance-based behavioral assessment via simulated communicative role-play; standardized clinician-administered observation.
• Target Population: Adults and older adults diagnosed with expressive aphasia, cognitive-communication disorders secondary to stroke, traumatic brain injury (TBI), neurodegenerative conditions, or neurosurgical interventions.
• Test Structure & Item Count: Total of 20 authentic open-ended scenarios divided into two psychometrically parallel forms (Versie A: 10 items; Versie B: 10 items). Both versions are preceded by two standardized practice/training scenarios to ensure patient comprehension of the direct-speech response expectation.
• Administration Protocol:
  • The examiner reads each scenario aloud following a standardized script, setting the context and ending with the standardized prompt: “Wat zegt u?” (“What do you say?”).
  • The patient is expected to respond directly in character (direct quotation). If the patient responds with indirect speech (e.g., “I would tell him that my car broke down”), the examiner delivers one standardized prompt: “Say it directly, as if you are talking to the person right now.”
  • Repetitions of the scenario are permitted if the patient requests or if auditory comprehension failure is suspected, but prompts cannot be paraphrased or simplified.
  • The entire verbal response is recorded (audio or video) and transcribed verbatim for objective clinical rating.
  • Administration time is approximately 15 to 25 minutes per version.
• Scoring Architecture & Response Scales: Each scenario response is rated independently across two 5-point ordinal scales:
  • Scale A: Begrijpelijkheid / Verbal Communicative Adequacy:
    • 1 = Not understandable / No relevant information conveyed: The response is incomprehensible, irrelevant, or absent; fails to convey any required semantic components.
    • 2 = Poorly understandable / Minimally relevant: Only a minimal fragment of the communicative intent is conveyed; the interlocutor cannot solve the situation without extensive external inference.
    • 3 = Moderately understandable / Partially relevant: Core information is partially transmitted; significant ambiguity remains, or critical details are missing, but the general direction of the message is recognizable.
    • 4 = Understandable with minor effort / Largely relevant: The essential communicative goal is attained; minor linguistic distortion or omission of trivial detail occurs, but an interlocutor requires minimal effort to deduce intent.
    • 5 = Fully understandable / Completely relevant: The message is delivered completely, effectively, and without communicative ambiguity; the pragmatic demand of the scenario is fulfilled.
  • Scale B: Verstaanbaarheid / Clarity of Vocal Expression (Intelligibility):
    • 1 = Unintelligible: Acoustic speech signal is completely unintelligible; severe phonemic disintegration or articulatory distortion.
    • 2 = Poorly intelligible: Severe phonetic/phonological errors; only isolated syllables or words can be recognized acoustically.
    • 3 = Moderately intelligible: Speech is intelligible with moderate concentration; distortion or paraphasias are frequent but do not obliterate all acoustic boundaries.
    • 4 = Largely intelligible: Speech is clear with minor articulatory or phonological imprecision.
    • 5 = Completely clear: Articulatory precision and acoustic intelligibility are unimpaired and effortless to process.
• Scoring Computation & Ranges:
  • Primary Outcome Score: The Total Functional Communicative Adequacy score is the sum of Scale A ratings across the 10 scenarios within a single version.
  • Theoretical Score Range (Scale A): 10 to 50 points per version (where 10 indicates absolute communicative failure across all items, and 50 reflects optimal functional communicative adequacy).
  • Theoretical Score Range (Scale B): 10 to 50 points per version (yielding an independent acoustic-motor clarity index).
  • Normative cutoff tables allow converting raw sum scores to percentile ranks and $T$-scores stratified by clinical severity.

Permissions & Fee and Test Year

The Amsterdam-Nijmegen Test voor Alledaagse Taalvaardigheid was formally published in 1995 by authors Leo Blomert, Corry H. Koster, and Mary-Louise Kean following comprehensive normative trials between 1991 and 1994. The original Dutch manual, response forms, and audio stimulus materials were published and distributed commercially by Swets & Zeitlinger Publishers (Lisse, Netherlands). Following historical publishing mergers, commercial publishing and licensing rights transitioned through Pearson Assessment and Information B.V. and subsequent distribution partners such as Hogrefe Publishing Group.

The test manual, scenario administration booklets, and standardized scoring protocol sheets are copyrighted clinical instruments subject to intellectual property protections. Access to official testing kits requires purchase through licensed psychometric test publishers. Use of the instrument in clinical healthcare or sponsored clinical research trials requires proper qualification (Level B or Level C psychological/speech-pathology assessment competence) and authorized materials. Bona fide researchers seeking to translate, adapt, or utilize the ANELT in non-profit academic research initiatives should consult standard rights-holder permissions channels.

References

• Austin, J. L. (1962). How to do things with words. Oxford University Press.
• Blomert, L., Kean, M. L., Koster, C., & Schokker, J. (1994). Amsterdam-Nijmegen Everyday Language Test: Construction, reliability and validity. Aphasiology, 8(4), 381–407. https://doi.org/10.1080/02687039408248666
• Blomert, L., Koster, C. H., & Kean, M. L. (1995). Amsterdam-Nijmegen Test voor Alledaagse Taalvaardigheid (ANTAT): Handleiding. Swets & Zeitlinger.
• Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semantics: Vol. 3. Speech acts (pp. 41–58). Academic Press. https://doi.org/10.1163/9789004368811_003
• Mokken, R. J. (1971). A theory and procedure of scale analysis: With applications in political research. De Gruyter Mouton. https://doi.org/10.1515/9783110813203
• Searle, J. R. (1969). Speech acts: An essay in the philosophy of language. Cambridge University Press. https://doi.org/10.1017/CBO9781139173438
• Snoek, J. W. (1996). Amsterdam-Nijmegen Test voor Alledaagse Taalvaardigheid (ANTAT) [Book Review]. Nederlands Tijdschrift voor Geneeskunde, 140(24), 1301.
• World Health Organization. (2001). International Classification of Functioning, Disability and Health: ICF. World Health Organization.

13. Items of the Scale (Questionnaire)