Content and construct validity of assessment tools
| Test | Content | Construct |
| Bayley-III | Expert opinion for standard and low verbal version.31 35 Literature reviews. Gross motor score correlated with Motor component 0.70.31 | Factor analysis. Difference in mean scores with pervasive developmental disorder, and specific language impairment.31 Hi (gross motor subset)=0.52–0.97 for children with language impairment and 0.82–0.99 in control group.35 |
| BOT-2 | Focus groups, product survey, pilot, national tryout and standardisation studies, professional reviews.12 | Factor analysis, scores increase with age, discriminates between normal and children with DCD (n=50), high-functioning ASD (n=45) and mild-to-moderate ID (n=66).12 |
| MABC-2 | Expert panel, stakeholder feedback, literature review.23 Expert panel—clarity (validity content index 71.8–93.9, κ=0.76–0.88) and pertinence (98.5–99.3 and κ=0.83–0.92), p<0.001.41 | Factor analysis, correlation coefficients.37 Subtest correlations 0.65–0.76, p<0.001. Discriminates between ASD and control group.23 Structural equation modelling (for each age group).40 Expert panel—adequate face validity.41 Significant difference between TD, DCD and at risk DCD scores (η2=0.63), p<0.0001.41 The UK norms not appropriate to use with Dutch/Flemish children as underestimate/overestimate risk of motor impairment.20 In Chinese population: CFA initially rejected. Acceptable fit achieved after 2 items removed.19 Age band 2 shows good validity in Japanese population.38 |
| MAND | Based on neuropsychological theory. Several rounds of revision/trials of tasks during development.32 | Factor analysis.32 43 Scores increase with age, and discriminate between typically developing children and those with head trauma or neurological dysfunction as well as gender.32 43 |
| NSMDA | Literature review. Developed by an experienced paediatric physiotherapist.46 | Factor analysis (up to 2 years of age).46 47 Stability of test results over time (up to 2 years).46 47 |
| PDMS-2 | Literature review. Created by experts in the field. Revised with feedback from therapists guided revision. Hierarchical sequence of items.34 | Item response modelling. Factor analysis. Differential item functioning analysis. Scores correlated with age (r=0.80–0.93).34 |
| TGMD-2 | Expert panel (3 PE teachers with postgraduate qualifications).15 Translated version (Brazilian Portuguese) language clarity 0.96, pertinence>0.89. Experts CVI for clarity and pertinence were also strong—α=0.93 clarity and α=0.91 pertinence.14 | Exploratory and CFA.13–18 High and significant correlation of increasing age and increasing scores.48 Age and disability differentiation.15 18 Subtest correlation 0.41.15 Galloping, running and leaping not well correlated with locomotion subscale. Object control significant and highly correlated.50 ANOVA—significant age effect for object control.18 Moderate correlation between items and subset scores, and between subset scores and total score.18 |
ANOVA, analysis of Variance; ASD, autism spectrum disorder; Bayley-III, Bayley Scale of Infant and Toddler Development third edition; BOT-2, Bruininks-Oseretsky Test of Motor Proficiency second edition12; CFA, Confirmatory Factor Analysis; Hi, scalability coefficient; ID, intellectual disability; MABC-2, Movement Assessment Battery for Children second edition29; MAND, McCarron Assessment of Neuromuscular Development32; NDI, Neurodevelopmental Index; NSMDA, Neurological Sensory Motor Developmental Assessment33; TD, typically developing; TGMD-II, Test of Gross Motor Development second edition15; WISC-R, Wechsler Preschool and Primary Scale of Intelligence-R; WPPSI, Wechsler Preschool and Primary Scale of Intelligence31.