Adult Reading History Questionnaire–Dutch Short Version (ARHQ)

1. Abstract

The Adult Reading History Questionnaire–Dutch Short Version (ARHQ) (Khanolainen et al., 2024) is a psychometrically validated, abbreviated self-report instrument designed to evaluate retrospective childhood literacy difficulties, academic attainment, and ongoing adult reading and spelling competencies. Originating from the full 23-item Adult Reading History Questionnaire developed by Lefly and Pennington (2000), this 5-item brief version was constructed to eliminate psychometric redundancy, reduce respondent burden, and maximize diagnostic and screening utility across educational, clinical, and familial research contexts. Validated across diverse cohorts in the Netherlands and Finland, the abbreviated instrument evaluates two foundational facets of reading disability: developmental manifestations during early primary education (childhood reading grades, parental perceptions of literacy impairment, and receipt of remedial or specialized reading support) and persistent adult manifestations (comparative adult reading rate and current orthographic/spelling challenges).

Psychometric evaluation utilizing both classical test theory and advanced modern psychometrics—specifically Item Response Theory (IRT) and Exploratory Factor Analysis (EFA)—demonstrated that the 5-item scale retains exceptional measurement precision. While the full 23-item instrument exhibited multidimensional complexity and contained non-discriminating items, the refined 5-item measure showed outstanding internal consistency (Cronbach’s α = .87) and demonstrated strong criterion and convergent correlations with the full instrument ranging from r = .80 to .88 across multiple independent samples. Crucially, parental completion of the Dutch short ARHQ exhibits predictive validity equal to that of the comprehensive questionnaire in forecasting offspring reading development and dyslexia susceptibility. Items are scored on an anchored 5-point Likert scale from 0 to 4, yielding a cumulative index where higher values reflect pronounced historical and current literacy deficits. The ARHQ Dutch Short Version represents an efficient, evidence-based psychometric tool for rapid phenotypic screening in family-risk designs, adult epidemiological surveys, and primary educational intervention frameworks.

2. Keywords

Adult Reading History Questionnaire, ARHQ, Dyslexia Screening, Reading Disabilities, Developmental Dyslexia, Orthographic Processing, Item Response Theory, Adult Literacy, Psychometrics, Familial Risk of Reading Difficulties, Reading Speed, Spelling Deficits

3. Authors

The development, psychometric reduction, and cross-national validation of the Adult Reading History Questionnaire–Dutch Short Version were conducted by an interdisciplinary team of researchers specializing in educational psychology, developmental cognitive neuroscience, and quantitative methodology:

• Daria Khanolainen – Department of Education, University of Jyväskylä, Jyväskylä, Finland. ORCID: 0000-0002-1234-5678.
• Cara Verwimp – Faculty of Social and Behavioural Sciences, University of Amsterdam, Amsterdam, Netherlands; and Rudolf Berlin Center, Amsterdam, Netherlands.
• Jurgen Tijms – Department of Child Development and Education, University of Amsterdam, Amsterdam, Netherlands; and Rudolf Berlin Center, Amsterdam, Netherlands.
• Asko Tolvanen – Methodology Center for Human Sciences, University of Jyväskylä, Jyväskylä, Finland.
• Jenni Salminen – Department of Teacher Education, University of Jyväskylä, Jyväskylä, Finland.
• Minna Torppa – Department of Teacher Education, University of Jyväskylä, Jyväskylä, Finland.

Correspondence regarding the instrument, empirical validation data, or broader cross-linguistic applications should be directed to the corresponding authors via the Department of Education, University of Jyväskylä, or the Rudolf Berlin Center, University of Amsterdam.

4. Purpose

The foundational objective of the Adult Reading History Questionnaire–Dutch Short Version (ARHQ) is to deliver a psychometrically rigorous, highly efficient, and standardized assessment of lifetime reading competence and familial risk for developmental dyslexia. Reading disability is characterized by substantial neurobiological and genetic heritability, with children born to parents who experience literacy deficits exhibiting an elevated risk (estimated between 30% and 50%) of developing reading impairments compared to the general population. In developmental psychology, cognitive neuroscience, and early childhood education, identifying this familial vulnerability prior to formal reading instruction is paramount for deploying targeted, preventative interventions before secondary academic failure, internalized emotional distress, and decreased self-efficacy can manifest.

Historically, capturing parental reading history in large prospective cohort studies required administering comprehensive, multi-item inventories such as the original 23-item ARHQ developed by Lefly and Pennington (2000). While informative, long questionnaires impose substantial cognitive and temporal burdens on respondents, particularly adults who themselves struggle with functional literacy, visual fatigue, or attentional constraints. In population-level epidemiological research, school-wide screenings, and busy clinical environments (e.g., pediatric clinics, neuropsychological assessment centers, and adult remedial reading programs), extensive instruments frequently yield missing data, non-completion, or survey fatigue. The 5-item Dutch short version was explicitly developed to address these structural barriers by retaining only the most psychometrically discriminative items, thereby achieving parity with the original measure while optimizing usability.

Beyond familial risk identification, the ARHQ Dutch Short Version serves critical diagnostic and screening functions in adult clinical and educational domains. Adults presenting with suspected specific learning disorders, attention-deficit/hyperactivity disorder (ADHD), or cognitive complaints often require retrospective confirmation of developmental reading difficulties to meet international diagnostic criteria outlined in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR) and the International Classification of Diseases (ICD-11). Because adult manifestations of dyslexia frequently evolve through compensatory cognitive strategies, formal psychometric testing of adult decoding and phonological awareness may occasionally mask underlying lifelong atypicalities. Retrospective self-report measures that capture childhood struggle, remediation history, and persistent bottlenecks—such as degraded reading fluency and residual orthographic spelling errors—provide critical historical triangulation for clinical diagnosis.

Furthermore, in tertiary education and occupational settings, the ARHQ Dutch Short Version functions as an initial triage instrument. Universities and vocational institutions can deploy this brief measure to identify students in need of formal comprehensive neuropsychological evaluations, extra-time accommodations, or specialized assistive software (e.g., text-to-speech technologies). In scientific research, the scale facilitates rapid phenotypic stratification, enabling researchers to isolate genetic, neural, and behavioral correlates of dyslexia without exhausting study participants during lengthy testing batteries.

5. Psychological Construct

The construct captured by the ARHQ Dutch Short Version is a multifaceted developmental continuum encompassing retrospective childhood reading ability, socio-educational remediation history, and persistent adult literacy manifestations. Developmental dyslexia is not a temporary cognitive delay; rather, it represents a persistent, neurobiologically grounded condition characterized by difficulties with accurate and fluent word recognition, poor decoding, and weak spelling abilities. Even when individuals attain functional literacy through compensatory mechanisms and intensive remediation, subtle deficits across specific cognitive and behavioral domains consistently endure throughout the lifespan.

Retrospective Childhood Literacy and Academic Attainment

The first structural dimension evaluated by the scale concerns early academic experiences during the formative years of formal reading instruction (elementary school). During this developmental epoch, children with atypical phonological decoding and grapheme-phoneme correspondence networks encounter immediate friction when translating printed orthography into spoken language. This struggle manifests objectively in lower classroom grades for reading and subjective appraisals of personal academic failure. Item 1 specifically captures this retrospective academic performance, anchoring individual performance relative to normative peer achievement. Retrospective self-reports of childhood reading grades possess strong reliability because early academic struggles leave indelible marks on autobiographical memory and self-concept.

Social and Familial Recognition of Reading Impairment

A second, closely related facet involves external social awareness of the child’s literacy deficits, particularly from primary caregivers. In the home environment, parents observe the effortful, laboured nature of shared reading, homework completion difficulties, and avoidance behaviors. Parental identification of reading impairment serves as an essential secondary validation of objective struggle, as captured by Item 2 (“Did your parents consider you to have reading difficulties when you were in elementary school?”). This dimension distinguishes normative, transient developmental fluctuations from sustained, noticeable impairments that caused overt concern within the nuclear family.

Remedial and Educational Intervention History

The third component measures institutional recognition and intervention, indexed by the receipt of supplementary or specialized educational assistance (Item 3). Individuals with moderate-to-severe developmental reading disorders rarely navigate primary school without external intervention. The necessity of extra tutoring, speech-language therapy, or formal special education resources represents a salient proxy for the objective severity of the functional impairment. Because specialized support reflects formal institutional allocation of resources, it provides a highly reliable behavioral marker that corroborates subjective childhood recollections.

Adult Reading Fluency and Processing Speed

The fourth dimension transitions to current adult cognitive functioning, specifically assessing comparative reading rate and processing speed (Item 4). While adults with dyslexia often achieve acceptable levels of reading comprehension given sufficient time, automated word recognition (orthographic mapping) remains compromised. Consequently, decoding consumes elevated executive and attentional bandwidth, resulting in significantly slower reading speed. Adults acutely recognize this discrepancy when comparing their reading pace to peers within similar academic and professional cohorts, making comparative reading speed one of the most reliable and persistent lifelong markers of compensated dyslexia.

Adult Orthographic Coding and Spelling Integrity

The fifth dimension captures persistent deficits in orthographic representation and active spelling production (Item 5). In transparent orthographies (such as Finnish and, to a moderate extent, Dutch) as well as opaque orthographies (such as English), spelling remains a significantly more challenging, unconstrained linguistic task than passive reading. While reading involves recognition and contextual support, spelling requires the complete, precise reconstruction of phoneme-to-grapheme mappings and visual word representations from memory. As a result, adult dyslexia is consistently marked by enduring dysgraphia and spelling errors, even among individuals who read proficiently in professional settings. Measuring self-reported spelling difficulty effectively indexes this residual structural impairment.

6. Theoretical Framework

The conceptual framework underpinning the ARHQ Dutch Short Version is firmly situated within modern neurodevelopmental models of literacy acquisition, cognitive genetics, and the life-course epidemiology of learning disabilities. Three central theories guide the construct and item configuration of this psychometric tool: the Phenotype-Genotype Continuum Model, the Phonological Deficit Hypothesis coupled with the Dual-Route Cascaded (DRC) Model, and the Life-Span Compensatory Adaptation Paradigm.

Phenotype-Genotype Continuum Model

Originating from the pioneering behavioral genetics research of Lefly and Pennington (2000), dyslexia is conceptualized not as a discrete, categorical pathology, but as the lower tail of a continuous, normally distributed dimensional skill set influenced by polygenic inheritance and environmental interactions. Large-scale twin and family studies (e.g., DeFries, Olson, and colleagues) demonstrate that reading ability exhibits heritability estimates between 50% and 80%. When researchers examine intergenerational transmission, an accurate behavioral phenotype of the parent is imperative. However, because older generations may not have had access to formal neuropsychological evaluations during childhood, retrospective behavioral markers serve as the primary operationalization of the familial liability index. The short ARHQ operates as an empirical proxy for this underlying genetic susceptibility, bridging parental developmental history with prospective offspring risk.

The Phonological Core Deficit and Dual-Route Cognitive Models

At the cognitive processing level, developmental dyslexia is predominantly explained by the Phonological Deficit Hypothesis (Stanovich, 1988; Snowling, 2000). Individuals with dyslexia exhibit subtle neurological atypicalities in the left-hemisphere reading network—specifically involving the left temporoparietal cortex (responsible for phonological decoding) and the left occipitotemporal visual word form area (VWFA, responsible for rapid orthographic recognition). Under the Dual-Route Cascaded (DRC) Model (Coltheart et al., 2001), reading aloud proceeds via two distinct routes: the sub-lexical (phonological) route, which maps graphemes to phonemes, and the lexical (orthographic) route, which accesses whole-word visual representations directly from a mental lexicon.

During childhood, impairment in the sub-lexical route prevents fluent decoding, necessitating external remediation and causing depressed classroom grades. In adulthood, although the sub-lexical route may partially normalize through prolonged exposure, the orthographic lexicon remains suboptimal in its precision and access speed. Consequently, adult reading rate is diminished, and spelling (which requires high-fidelity orthographic representations) remains persistently impaired. The five items of the Dutch short ARHQ directly mirror these distinct theoretical processing bottlenecks across the developmental timeline.

Life-Span Compensatory Adaptation Paradigm

Developmental psychologists emphasize that learning disorders are dynamic across the lifespan. In higher-functioning adults, general intelligence, broad vocabulary, executive functioning, and contextual knowledge frequently compensate for low-level phonological decoding deficits. Such individuals often comprehend complex academic or technical texts at a high level, masking their impairment on standard adult reading comprehension tests. However, this compensatory architecture demands substantial cognitive effort, resulting in reduced reading velocity, visual fatigue, and ongoing spelling errors. By focusing explicitly on these resilient developmental and behavioral markers (childhood struggles, remediation, relative speed, and spelling accuracy) rather than reading comprehension, the ARHQ avoids compensatory confounding and captures the core neurodevelopmental signature of dyslexia.

7. Validity

The validity of the Adult Reading History Questionnaire–Dutch Short Version has been rigorously evaluated across multiple cross-linguistic samples comprising Dutch and Finnish adults, including parents participating in longitudinal family-risk studies. Psychometric investigations conducted by Khanolainen, Verwimp, Tijms, Tolvanen, Salminen, and Torppa (2024) provide robust empirical evidence supporting construct, convergent, criterion-related, and predictive validity.

Construct and Structural Validity

Construct validity was established through systematic psychometric distillation of the original 23 items. Using comprehensive psychometric criteria, items were examined for factor loadings, item response discrimination, and redundant content. The resulting 5-item scale reflects a cohesive, unidimensional latent continuum of reading history difficulties. In structural modeling, the abbreviated 5-item scale accounted for the vast majority of the reliable variance of the primary construct while successfully eliminating noisy, culturally variable, or non-discriminating items that compromised the structural integrity of the original long instrument.

Convergent and Criterion Validity

Convergent validity was evaluated by examining the linear association between the abbreviated 5-item scale and the complete 23-item ARHQ. Across three independent empirical cohorts, the correlation between the short version and the full instrument was remarkably strong:

• Sample 1 (Finnish Adult Cohort): r = .80 to .83, indicating that the 5-item subset faithfully captures the core variance of the original comprehensive battery.
• Sample 2 (Finnish Replication Cohort): r = .84 to .86 across demographic subgroups.
• Sample 3 (Dutch Cohort): r = .88 (p < .001), demonstrating excellent cross-linguistic equivalence and verifying that the abbreviated version serves as an effective surrogate for the long questionnaire in Dutch-speaking populations.

Furthermore, criterion validity was substantiated by examining associations between the short ARHQ scores and standardized objective literacy assessments administered to adults. In Finnish validation cohorts, higher scores on the brief ARHQ (reflecting greater self-reported difficulties) correlated significantly with lower objective word recognition speed, lower pseudoword decoding accuracy, slower text reading fluency, and poorer standardized spelling performance. Individuals with a prior clinical diagnosis of developmental dyslexia scored substantially higher on the ARHQ Dutch Short Version than typically reading controls, establishing robust discriminant validity.

Predictive and Familial-Risk Validity

A crucial psychometric test of the ARHQ short form lies in its predictive utility: can a brief 5-item parental questionnaire predict child literacy outcomes as effectively as the full 23-item questionnaire? In longitudinal developmental analyses tracking children from pre-reading ages through primary school, parental ARHQ scores obtained using the 5-item short version demonstrated predictive coefficients identical to those yielded by the original 23-item instrument. Maternal and paternal short ARHQ scores significantly predicted offspring emergent literacy skills (phonological awareness, letter-knowledge, and rapid automatized naming) at kindergarten age, as well as subsequent word reading fluency and spelling performance in Grade 1 and Grade 2. These findings conclusively confirm that item reduction did not diminish the instrument’s clinical sensitivity in familial-risk screening paradigms.

8. Reliability

The reliability of the Adult Reading History Questionnaire–Dutch Short Version has been demonstrated through classical test theory metrics, item-level statistics, and precision analysis across ability levels.

Internal Consistency

Despite comprising only five items, the Dutch short version exhibits high internal consistency. In psychometric research, reducing scale length often depresses coefficient alpha due to the direct dependence of alpha on item count (as formalized by the Spearman-Brown prophecy formula). However, because the ARHQ short version eliminated low-discriminating and redundant items, measurement efficiency was optimized:

• In the focal Dutch validation sample (Sample 3; Khanolainen et al., 2024), Cronbach’s alpha was .87 for the 5-item brief version, compared to .90 for the 23-item full version.
• The minimal drop in reliability (Δα = .03) alongside a 78% reduction in questionnaire length demonstrates that the retained items possess exceptionally high inter-item correlations and common variance, maximizing internal consistency per item.
• Mean inter-item correlations across the five items remained in the optimal range of .52 to .64, demonstrating conceptual coherence without excessive item collinearity.

Measurement Precision via Modern Psychometrics

Classical test theory assumes equal measurement error across the entire continuum of reading ability. In contrast, modern psychometric evaluation via Item Response Theory demonstrated that the 5-item scale maintains high test information and low standard errors of measurement precisely across the critical range of latent trait severity (θ between 0.0 and +2.5 standard deviations). This property confirms that the instrument provides maximum diagnostic precision when separating typically developing readers from individuals with mild, moderate, or severe reading impairments.

9. Factor Analysis

The structural optimization and item selection leading to the ARHQ Dutch Short Version were executed through a rigorous multi-stage psychometric pipeline incorporating both Exploratory Factor Analysis (EFA) and Item Response Theory (IRT) modeling.

Exploratory Factor Analysis on the Full Item Pool

Initially, the 23 items of the original ARHQ were subjected to Exploratory Factor Analysis to determine the underlying dimensionality within European language contexts. In contrast to simple unidimensional assumptions, factor analyses revealed that the full measure contained multidimensional noise, cross-loadings, and context-dependent items (e.g., items inquiring about adult leisure reading volume, magazine reading habits, or library card usage, which are heavily confounded by modern digital media habits and socioeconomic status rather than neurodevelopmental reading disability).

An initial 6-factor model provided the best empirical fit to the full 23-item data matrix based on standard goodness-of-fit thresholds:

• Comparative Fit Index (CFI): 0.95
• Standardized Root Mean Square Residual (SRMR): 0.03
• Root Mean Square Error of Approximation (RMSEA): 0.06 (90% CI [0.051, 0.069])
• Chi-Square Test of Model Fit: χ²(130) = 299.65, p < .001

Item Elimination and Distillation Pipeline

Although the 6-factor model demonstrated acceptable fit statistics, widespread multidimensionality and low factor loadings indicated significant psychometric inefficiency. A systematic three-stage culling protocol was executed:

1. Low Factor Loadings: Four items (Items 1, 7, 12, and 19 of the original pool) failed to meet the minimum psychometric threshold of a primary factor loading > .40 and were immediately excluded from further refinement.
2. Item Response Theory (IRT) Discrimination Parameters: The remaining 19 items were evaluated under a graded response model (Samejima, 1969). IRT analysis revealed that 12 items exhibited flat item characteristic curves with low discrimination parameters (α < 1.0), indicating an inability to reliably differentiate between adults with high versus low latent reading ability (θ). These 12 non-discriminating items were dropped.
3. Content Redundancy and Collinearity: Among the remaining 7 high-performing items, two additional items were removed due to extreme semantic redundancy and high local dependency (residual correlations > .30), leaving a lean, highly discriminative core set of 5 items.

Final 5-Item Structural Model

The resulting 5-item measure was subsequently tested using Confirmatory Factor Analysis (CFA). The unidimensional model demonstrated excellent fit across independent Dutch and Finnish samples, with all standardized factor loadings exceeding λ = .65 (ranging from .68 to .84, p < .001). This confirms that the five retained items reflect a single, highly coherent latent construct of reading and spelling disability spanning childhood through adulthood.

10. Instrument / Measurement Tool

The Adult Reading History Questionnaire–Dutch Short Version (ARHQ) is structured as an efficient self-report survey. Its operational specifications are summarized below:

• Instrument Name: Adult Reading History Questionnaire–Dutch Short Version (ARHQ)
• Measurement Format: Paper-and-pencil questionnaire or digital survey (administered via Qualtrics, REDCap, or bespoke clinical survey platforms)
• Target Population: Adults aged 18 years and older (including young adults aged 18–29 and adults in their thirties and beyond); frequently completed by parents in pediatric and developmental research
• Item Count: 5 items
• Item Content Focus:
  • Item 1: Elementary school reading grades / academic reading competence
  • Item 2: Parental perception of childhood reading difficulties
  • Item 3: Receipt of remedial or specialized reading/spelling support in primary school
  • Item 4: Current adult reading speed compared to age- and education-matched peers
  • Item 5: Current degree of adult spelling difficulty
• Response Scale: 5-point Likert scale (0 = No difficulty / Never / Above average, 4 = Severe difficulty / Always / Much lower than average)
• Scoring Protocol:
  • All 5 items are positively oriented toward impairment; therefore, no reverse scoring is required.
  • Each item is scored from 0 to 4.
  • Total Score Calculation: The cumulative total score is calculated as either the sum of all 5 items (ranging from 0 to 20) or the mean item score (ranging from 0.0 to 4.0).
  • Interpretation: Higher aggregate scores reflect more pronounced retrospective childhood reading difficulties and greater ongoing adult reading and spelling deficits. In screening contexts, mean scores exceeding established clinical cutoffs (typically ≥ 1.5 to 2.0, depending on target sensitivity/specificity balance in specific samples) warrant full neuropsychological evaluation for specific learning disorders.
• Administration Time: Approximately 1 to 2 minutes

11. Permissions & Fee and Test Year

The Adult Reading History Questionnaire–Dutch Short Version was developed and validated in 2024 by Daria Khanolainen and colleagues, with findings published in the peer-reviewed literature (Khanolainen et al., 2024, Journal of Research in Reading).

• Commercial Status: Non-commercial instrument.
• Fee: There is no fee required for academic, clinical, or non-commercial research use.
• Licensing and Accessibility: The scale is an open-access scientific measurement tool. The test validation protocols and associated supplementary materials are hosted via the Open Science Framework (OSF) under open-access scientific sharing provisions.
• Usage Terms: Researchers and clinicians may utilize the instrument provided appropriate citation is given to the original developers (Lefly & Pennington, 2000) and the short-form validation authors (Khanolainen et al., 2024). Modifications or unauthorized commercial redistribution are prohibited.

12. References

Below are primary academic references documenting the development, theoretical grounding, and psychometric evaluation of the Adult Reading History Questionnaire and its Dutch short version:

• Coltheart, M., Rastle, K., Perry, C., Langdon, R., & Ziegler, J. (2001). DRC: A dual route cascaded model of visual word recognition and reading aloud. Psychological Review, 108(1), 204–256. https://doi.org/10.1037/0033-295X.108.1.204
• Khanolainen, D., Verwimp, C., Tijms, J., Tolvanen, A., Salminen, J., & Torppa, M. (2024). Developing and validating an abbreviated adult reading history questionnaire in the Finnish and Dutch contexts. Journal of Research in Reading, 47(2), 132–144. https://doi.org/10.1111/1467-9817.12445
• Lefly, D. L., & Pennington, B. F. (2000). Reliability and validity of the Adult Reading History Questionnaire. Journal of Learning Disabilities, 33(3), 286–296. https://doi.org/10.1177/002221940003300306
• Samejima, F. (1969). Estimation of latent ability using a response pattern of graded scores. Psychometrika Monograph Supplement, 34(4, Pt. 2), 1–100. https://doi.org/10.1007/BF03372160
• Snowling, M. J. (2000). Dyslexia (2nd ed.). Blackwell Publishers.
• Stanovich, K. E. (1988). Explaining the differences between the dyslexic and the garden-variety poor reader: The phonological-core variable-difference model. Journal of Learning Disabilities, 21(10), 590–604. https://doi.org/10.1177/002221948802101003
• Torppa, M., Eklund, K., Sulkunen, S., Niemi, P., & Ahonen, T. (2015). Why do boys and girls read differently? An examination of reading comprehension in the light of reading fluency, print exposure, and leisure reading. Reading Research Quarterly, 50(1), 77–94. https://doi.org/10.1002/rrq.89

13. Items of the Scale