Data-Based Decision-Making for Core Instruction¶

Strategy · All strategies
Evidence · 2 claims (1 for, 1 mixed) · 3 studies (2 quant-synthesis, 1 review), q2–q4 · 2 of 3 report an effect size · 1 claim rests on one study

Description¶

Data-based decision-making (DBDM) for core instruction is the systematic cycle of collecting student performance data during whole-class (Tier 1) teaching, analyzing it against learning goals, and using the results to adjust instruction — re-teaching, regrouping, changing representations, or accelerating. It treats core instruction itself as an object of continuous improvement rather than assuming that only intervention tiers need data.

Design Implications¶

DBDM operationalizes Assessment for Learning at the classroom level: frequent, curriculum-embedded measures give teachers evidence to close the gap between current performance and the goal Formative evaluation produces substantial learning gains when it feeds instructional adjustment. [+S]. The value lies not in the data but in the decision rules attached to it — data without a predefined action ("if fewer than 80% of students meet the criterion, re-teach with a different representation") rarely changes instruction Data use improves outcomes only when paired with explicit instructional decision rules. [+M].

Context¶

Requirements¶

• Brief, reliable, curriculum-aligned measures (e.g., Curriculum-Based Measurement) administered at regular intervals
• Predefined decision rules that map performance bands to specific instructional actions
• Time and structures for teachers to analyze data collectively (grade-team or PLC data meetings)
• Instructional alternatives ready to deploy — re-teaching scripts, alternative representations, flexible grouping plans

Constraints¶

• Data collection without protected analysis time produces compliance behavior, not instructional change [~M]
• Over-reliance on single high-stakes snapshots misleads; brief repeated measures outperform infrequent summative data for instructional decisions [~S]
• Teachers often interpret data to confirm existing beliefs; structured protocols and collaborative analysis reduce confirmation bias [-M]
• Poorly aligned measures (testing what was not taught) generate decisions that waste instructional time [-M]

Implementation Variability¶

• Team-based: grade-level teams analyze common assessment data in scheduled meetings (the professional learning community model)
• Individual: a single teacher uses exit tickets and quick checks daily to adjust the next lesson
• System-supported: platforms (e.g., AIMSweb, DIBELS) automate progress monitoring and flag students against decision rules
• DBDM plus inquiry: teachers frame a hypothesis, collect data, and evaluate the instructional change (action-research variant)

Target Learners¶

• Whole classrooms in Tier 1 settings — the goal is preventing failure rather than remediating it Progress monitoring with decision rules improves achievement, with the largest effects when instruction is adjusted based on the data. [+S]
• Students whose needs would otherwise go unnoticed until they require intensive intervention
• Teachers benefit as much as students: the cycle builds pedagogical content knowledge about which adjustments work [~M]

Target Learning Goals¶

• Mastery of core academic content (reading, mathematics) at the whole-class level
• Early identification of skill gaps before they compound
• Teacher learning: calibrating instruction to actual student performance rather than pacing guides

Instructions¶

1. Define the essential learning goal and the criterion for "enough" (e.g., 80% of students at mastery).
2. Select or design a brief, aligned measure; administer it at a regular interval (Assessment).
3. Analyze results against the decision rule — ideally with a team using a structured protocol (Check-In routines can structure the meeting).
4. Act: re-teach with a different representation, regroup for targeted practice (Practice), or accelerate if the criterion is met.
5. Re-measure after the adjustment and evaluate whether the change worked; feed results into the next cycle (Coaching supports fidelity of this step).

Claims¶

• Mathematics teachers differ from other teachers in data use, attitudes, and self-efficacy for DBDM [~W]

Related Strategies¶

• Formative Assessment — the assessment principle DBDM enacts at the program level
• Direct Instruction — a common Tier 1 core whose pacing and emphasis DBDM adjusts
• Weekly teacher team meetings to review each student's strengths, progress, and obstacles
• Begin improvement work by collecting data and close the cycle by evaluating progress
• Data-driven decision-making practices in tutoring programs

Examples¶

• Response to Intervention (RTI) — Tier 1 progress monitoring with decision rules (e.g., DIBELS benchmarks every 6–8 weeks) determines whether core instruction is sufficient or students need supplemental tiers (Fuchs & Fuchs, 2006).
• Data Wise (Harvard) — a structured 8-step cycle for school teams to analyze student work and adjust instruction (https://www.gse.harvard.edu/data-wise).
• CGI-informed math instruction — teachers analyze student strategy use on problem sets to decide which problem types and representations to feature next (Cognitively Guided Instruction).

Key Sources¶

• Mandinach, E. B. (2012). A perfect time for data use: Using data-driven decision making to inform practice. Educational Psychologist, 47(2), 71–85. doi:10.1080/00461520.2012.667064
• Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy & Practice, 5(1), 7–74. doi:10.1080/0969595980050102
• Fuchs, L. S., & Fuchs, D. (2006). Introduction to response to intervention: What, why, and how valid is it? Reading Research Quarterly, 41(1), 93–99. doi:10.1598/rrq.41.1.4
• Stecker, P. M., Fuchs, D., & Fuchs, L. S. (2005). Using curriculum-based measurement to improve student achievement: Review of research. Psychology in the Schools, 42(8), 795–819. doi:10.1002/pits.20113
• Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–112. doi:10.3102/003465430298487