Sensitivity and specificity of Three Tier screening tools vary widely, and no studies report language-domain-only accuracy¶

Claim · All claims
Evidence · 1 study · quant-synthesis r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Across nine studies and 14 screening tools, sensitivity ranged from 0 to 100 and specificity from 15 to 100; in mathematics only one tool was evaluated, with perfect sensitivity (100) but poor specificity (15–32). → Keles 2025

Evidence¶

Keles 2025¶

Keles, S., ten Braak, D., Hagen, Å. M., & Melby-Lervåg, M. (2025). Validity of Early-Grade Screening Tools in the Three Tier (RTI/MTSS) Model: A Systematic Review and Meta-Analysis of Literacy, Language, and Mathematics. Educational Psychology Review. https://doi.org/10.1007/s10648-025-10077-5

q2 · i? · quant-synthesis · r2

Descriptive synthesis of sensitivity and specificity data from nine validation studies covering 14 screening tools used for tier assignment. The review reports "The sensitivity range for the tools across domains varied from zero and 100" and cut-offs from 7 to 40%. In literacy, sensitivity was 59–100 and specificity 38–84.2; the single mathematics tool showed specificity of 15–32.

"The sensitivity range for the tools across domains varied from zero and 100, while the specificity range across domains spanned from 15 to 100. The cut-off used ranged from 7 to 40% across the different studies."

Discussion¶

Related Claims¶

• DIBELS is the most frequently used screening tool for tier assignment, and literacy is the most assessed domain — related
• Very few studies (18 of 127) provide validity data for screening tools used in tier assignment, and none test whether screening itself improves achievement — related
• Screening domain is a significant moderator of validity: literacy-literacy correlations (r = .55) exceed language-literacy (r = .38) and mathematics/literacy-achievement (r = .27) — related
• Screening tools in the Three Tier model show a moderate pooled correlation with academic achievement outcomes (r = .51), with very high heterogeneity — related