Predicting¶

Strategy · All strategies
Evidence · 2 claims (1 for, 1 mixed) · 5 studies (3 causal, 1 quant-synthesis, 1 review), q2–q3 · 1 of 5 report an effect size

Description¶

Predicting asks learners to anticipate what will happen next in a text, demonstration, or problem before the answer is revealed. In reading, learners forecast plot, argument, or content; in mathematics, they identify what type of question is being asked, which operations might apply, and what a reasonable answer would look like. The act of committing to a prediction engages interest, exposes gaps in knowledge, and prepares learners to evaluate incoming information against their expectations.

Design Implications¶

Prediction functions as a form of pretesting: even when predictions are wrong, generating them activates relevant prior knowledge and improves retention of the correct information that follows Attempting to answer questions before instruction, even unsuccessfully, enhances subsequent learning. [+M]. Predictions create curiosity and a need to confirm or disconfirm, which increases attention during the subsequent explanation or reading. The strategy works best when learners must commit to a specific, checkable prediction rather than a vague guess, and when feedback follows quickly enough to resolve the uncertainty it created.

Context¶

Requirements¶

• Content with a genuine predictive structure — a narrative, a causal process, a problem with a determinate answer
• Prompts that require a specific commitment ("What will the next step be? What do you think the answer is?") rather than open speculation
• Timely confirmation or disconfirmation, so the prediction is resolved rather than left dangling (Practice, Assessment)

Constraints¶

• Requires some prior knowledge of the topic; learners with none can only guess randomly, and random guessing produces little activation Activation improves learning only when relevant knowledge exists to be activated. [~M]
• Predictions based on strong misconceptions can entrench errors if the correction is weak or delayed — high-confidence wrong predictions need explicit disconfirmation High-confidence errors, once corrected, improve retention more than low-confidence ones. [~M]
• Overuse of "guess what happens next" in narrative reading can fragment comprehension if predictions interrupt the flow of the text
• In well-structured domains, prediction before any modeling can impose unproductive search on novices; a Demonstration or worked example may be needed first

Implementation Variability¶

• Text prediction: pause at chapter titles, headings, or mid-narrative points and ask learners to forecast content before reading on
• Mathematical prediction: before solving, ask what the problem is asking, what operation fits, and whether the answer will be large or small, positive or negative — an estimation and plausibility check
• Demonstration prediction: before an experiment or worked example, learners predict the outcome, then compare against the actual result (Anchored Instruction contexts)
• Reciprocal teaching: prediction is one of four strategies students rotate through in small-group reading discussion

Target Learners¶

• Learners with moderate prior knowledge, who have enough schema to generate meaningful predictions Activation improves learning. [+M]
• Reluctant or passive readers, for whom prediction points create a purpose for reading
• Novices in problem-solving domains benefit from the "what kind of problem is this?" variant, which builds classification skill before computation
• Less effective for complete novices, who lack the knowledge base to predict anything but arbitrarily

Target Learning Goals¶

• Comprehension monitoring: noticing when text violates expectations
• Prior knowledge activation: connecting new information to what is already known
• Problem classification: recognizing problem types and appropriate operations before executing them
• Motivation and engagement: curiosity about whether one's prediction was right

Instructions¶

1. Select a prediction point — a heading, a pause in a demonstration, or the statement of a problem before its solution.
2. Prompt a specific, committed prediction: "What do you think will happen? What will the answer be close to?" (Activation)
3. Have learners record or share predictions so they are accountable and comparable to the outcome.
4. Reveal the outcome through reading, Direct Instruction, or completing the problem.
5. Compare prediction to result and discuss discrepancies — the discussion of why the prediction was wrong is where much of the learning occurs.
6. Follow with Practice in which learners generate their own predictions independently, fading the prompts.

Related Strategies¶

• Activating Prior Knowledge — prediction is a specific mechanism for activation; it forces retrieval rather than merely prompting for it
• Think-Aloud Modeling — expert prediction ("I expect this to…") models the strategy for learners
• Pretesting — the broader family of before-instruction attempts that prediction belongs to
• Present problems before explanatory texts to arouse curiosity and situational interest in learners lacking prior knowledge
• Use small-group discussion of a phenomenon-describing problem before text study to bridge prior knowledge and new information

Related Elements¶

• Activation — the cognitive function prediction serves
• Advance Organizers — a related pre-reading structure; organizers frame content, predictions commit learners to expectations
• Practice — where independent prediction becomes a habitual self-regulation move

Patterns That Use This Element¶

• Cognitive Apprenticeship — learners predict expert moves before observing the modeled performance
• Reciprocal Teaching — prediction is one of the four cycled comprehension strategies
• Flipped Classroom — pre-class prediction prompts prepare learners to check their expectations against instruction

Examples¶

• Reciprocal teaching (Palincsar & Brown) — small reading groups predict upcoming content at designated points before summarizing, questioning, and clarifying; the routine raised sixth-graders' comprehension substantially in the original studies.
• Physics "predict–observe–explain" — students predict the outcome of a demonstration (e.g., which mass hits the ground first), observe it, and explain any discrepancy; widely used to confront misconceptions.
• Estimation in mathematics — before computing 47 × 63, students predict whether the answer is closer to 300 or 3,000, building number sense and a plausibility check for their eventual answer.

Key Sources¶

• Palincsar, A. S., & Brown, A. L. (1984). Reciprocal teaching of comprehension-fostering and comprehension-monitoring activities. Cognition and Instruction, 1(2), 117-175. doi:10.1207/s1532690xci0102_1
• Kornell, N., Hays, M. J., & Bjork, R. A. (2009). Unsuccessful retrieval attempts enhance subsequent learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(4), 989–998. doi:10.1037/a0015729
• Richland, L. E., Kornell, N., & Kao, L. S. (2009). The pretesting effect: Do unsuccessful retrieval attempts enhance learning? Journal of Experimental Psychology: Applied, 15(3), 243–257. doi:10.1037/a0016496
• Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58. doi:10.1177/1529100612453266