Baseline Assessment¶

Strategy · All strategies
Evidence · 3 claims (3 for) · 8 studies (5 quant-synthesis, 2 causal, 1 review), q2–q4 · 5 of 8 report an effect size

Description¶

Baseline assessment measures what learners know and can do before instruction begins — prior knowledge, prerequisite skills, misconceptions, and sometimes attitudes or self-efficacy. It is carried out through diagnostics, pretests, placement tasks, or brief ungraded probes, and its primary purpose is to inform instructional decisions (pacing, grouping, content selection), not to grade learners.

Design Implications¶

Baseline assessment enables instruction to start where learners actually are rather than where the curriculum assumes they are, and it is the entry point for Assessment for Learning Assessment for learning practices improve achievement. [+S]. Its diagnostic value depends on alignment: the baseline must sample the same knowledge and task types the instruction will target, or the data cannot drive decisions. Results feed directly into Adaptive Learning pathways and inform whether Activation activities are needed to surface relevant prior knowledge Activating prior knowledge improves learning. [+M].

Context¶

Requirements¶

• Clear definition of the prerequisite knowledge and skills the instruction assumes
• Items or tasks that genuinely discriminate between levels of prior knowledge, including common misconceptions as distractors
• A decision rule: what instructional action follows each possible result (skip, remediate, regroup, proceed)
• Low-stakes framing so results reflect knowledge, not test anxiety

Constraints¶

• Pretests on material learners have not yet studied can produce poor performance that demotivates or misleads [~M] — keep baselines focused on prerequisites, not target content
• Baseline data go stale: prior-knowledge measures taken weeks before instruction lose predictive value as learners change
• Over-reliance on a single snapshot misclassifies learners, especially those with test anxiety or limited academic English [~M]
• If results are collected but not acted on, assessment consumes time and adds no instructional value [-M]

Implementation Variability¶

• Formal placement tests (e.g., math placement exams) that route learners into differentiated tracks
• Ungraded diagnostic quizzes embedded in the first session or LMS module
• Concept inventories targeting known misconceptions (e.g., the Force Concept Inventory in physics)
• Performance-based baselines: a short writing sample or coding task that reveals skill, not just declarative knowledge
• Adaptive diagnostics that adjust item difficulty to locate a learner's level efficiently

Target Learners¶

• Mixed-prior-knowledge cohorts, where a single starting point would leave some learners lost and others bored Adaptive learning improves outcomes. [+M]
• Novices with fragmented or naive prior knowledge whose misconceptions would otherwise go undetected
• Returning or adult learners whose skills have decayed and who need accurate re-entry points
• Less useful for homogeneous expert cohorts, where a baseline adds little information [~W]

Target Learning Goals¶

• Prerequisite identification: verifying readiness for new content
• Misconception diagnosis: surfacing naive models before they interfere with new learning
• Personalization inputs: setting initial difficulty and pathways in adaptive systems
• Growth measurement: establishing the "before" point against which learning gains are measured

Instructions¶

1. Define the prerequisite knowledge and skills the unit assumes; write or select items that sample them, including distractors based on documented misconceptions.
2. Administer the baseline as low-stakes and early as possible, ideally before the first instructional session.
3. Analyze results against a pre-specified decision rule: who proceeds, who needs Activation of prior knowledge, who needs remediation or Accommodations.
4. Use results to set starting difficulty and groupings; in adaptive systems, feed them into Adaptive Difficulty as the initial estimate.
5. Re-baseline at major transitions (new unit, new term) rather than treating the first snapshot as permanent.

Related Strategies¶

• Formative Assessment — baseline assessment is the first data point in an ongoing assessment-for-learning cycle
• Mastery Learning — baselines determine the starting point from which mastery criteria are applied
• Check-Ins — lightweight ongoing probes that continue the diagnostic function during instruction
• Start adult learners in homogeneous small groups and use individual instruction with gradual release
• Design participation deliberately: align entry points with partners' actual readiness rather than assuming a uniform baseline

Examples¶

• Force Concept Inventory (FCI) — a widely used physics diagnostic that identifies Newtonian vs. naive-mechanics reasoning before mechanics instruction; results are used to compare instructional approaches (e.g., interactive engagement vs. lecture).
• Khan Academy — course challenges and initial mastery checks place learners on a skill map, calibrating which practice sets are recommended first.
• ALEKS — an adaptive knowledge-space assessment that maps exactly which topics a learner is ready to learn, then builds a course path from that baseline.
• Diagnostic writing samples in first-year composition — a short ungraded essay in week one that informs feedback priorities and workshop grouping.

Key Sources¶

• Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy & Practice, 5(1), 7–74. doi:10.1080/0969595980050102
• Hestenes, D., Wells, M., & Swackhamer, G. (1992). Force Concept Inventory. The Physics Teacher, 30(3), 141–158. doi:10.1119/1.2343497
• Dochy, F., Segers, M., & Buehl, M. M. (1999). The relation between assessment practices and outcomes of studies: The case of research on prior knowledge. Review of Educational Research, 69(2), 145–186. doi:10.3102/00346543069002145
• Fuchs, L. S., & Fuchs, D. (1986). Effects of systematic formative evaluation: A meta-analysis. Exceptional Children, 53(3), 199–208. doi:10.1177/001440298605300301