Skip to main content
Account

Table 1 Summary of psychometric property indicators and the associated common acceptability thresholds

From: Measuring Mathematical Skills in Early Childhood: a Systematic Review of the Psychometric Properties of Early Maths Assessments and Screeners

Psychometric Evidence

Example Analysis Methods

Common Acceptability Thresholds

Content validity

Expert panel of experts and users

Agreement across experts, with adjustments made to items when required

Structural validity

Confirmatory factor analysis (CFA)

RMSEA < .06; CFI > .95; TLI > .95 (Hu & Bentler, 1999)

 

Rasch model

0.5–1.5 (Linacre, 2017)

Internal consistency

Cronbach’s alpha; Kuder-Richardson (KR-20) coefficient; split-half reliability correlations

 ≥ .70 (Prinsen et al., 2018)

Reliability

Correlations for test–retest and inter-rater reliability

 ≥ . 70 (NCII, 2019)

Criterion validity

Diagnostic accuracy

Sensitivity ≥ .90; Specificity ≥ .70 (Jenkins et al., 2007; Kilgus et al., 2014)

 

Concurrent, divergent, and predictive correlations between the evaluated measurement tool and ‘Gold Standard’ measurement tools

 ≥ .60 (NCII, 2019)