6.2 Making Predictions from Textual Evidence
Key Takeaways
- Descriptor D2 asks you to make predictions based on several pieces of textual evidence — about what happens next, what a character will do, or how information applies in a new situation.
- A likely prediction is grounded in two or more textual clues; a wishful guess is grounded in personal preference, outside knowledge, or what usually happens in stories.
- The prediction must follow the passage's internal logic: a cautious character will not suddenly act recklessly without a textual trigger, and a slow process will not produce instant results unless the passage says so.
- When two predictions seem possible, choose the one with more independent textual support; the HiSET rewards the prediction the evidence most strongly supports.
- Reject any prediction that depends on an outcome the passage gives no evidence for, such as a council reversing a decision or a person changing a long-standing habit.
What Descriptor D2 Actually Tests
The second synthesis skill is make predictions based on several pieces of textual evidence. Like D1, D2 requires combining multiple pieces of evidence. The difference is the output: D1 produces a conclusion about what is true in the passage; D2 produces a prediction about what is likely to happen next, what a character is likely to do, or how the information is likely to apply in a new situation.
D2 questions on the HiSET typically take three forms:
- Plot prediction: "Based on the passage, what is the narrator most likely to do next?"
- Character prediction: "Which action would the character be most likely to take if the situation described continued?"
- Application prediction: "If the process described in the passage were applied to a new setting, what would most likely result?"
The common thread is that the prediction must be grounded in evidence the passage provides, not in what you would personally do, not in what usually happens in stories, and not in what you wish would happen.
Likely Prediction vs Wishful Guessing
The central discipline of D2 is separating a likely prediction (supported by textual evidence) from a wishful guess (supported by personal preference). The test writers know that readers naturally project their own hopes onto open-ended questions, so they build distractors that appeal to what a reader might want rather than what the text suggests.
Three tests distinguish likely from wishful:
- Evidence count: can you name two or more textual clues that point toward the prediction? If you cannot, the prediction is a guess.
- Alternative check: can you imagine a different outcome that the same evidence would also support? If yes, your prediction is too vague or too strong.
- Logic check: does the prediction follow the text's internal logic — the character's established motivations, the process's described mechanisms, the argument's trajectory — or does it impose outside logic?
A prediction that passes all three is a D2 answer. One that fails any is a distractor.
Following the Text's Internal Logic
A prediction is only as good as its fit with the passage's own rules. A character who has been cautious for three paragraphs will not suddenly act recklessly without a textual trigger. A scientific process described as slow and temperature-sensitive will not produce instant results in a new setting unless the passage says so. The HiSET rewards the reader who treats the passage as a closed system: the prediction must be derivable from the evidence and rules the passage itself establishes.
This is why personal experience is a weak basis for a D2 answer. Your experience of how a similar character acted in a different story, or how a similar process works in a different field, may or may not match the specific passage. Use the passage, not your outside knowledge, as the prediction engine.
A Worked D2 Example
Here is a short informational passage in the style of a HiSET informational text:
The volunteer fire department in Halverson runs on a rotation: three members on call each night, with a backup list for nights when more than one call comes in. Last winter, when a storm brought three calls in a single night, Captain Lugo called in four backups and still managed the responses without delays. This winter, the department's roster has dropped from twenty-two to fifteen, and two of the remaining members are recovering from surgery. The town council has declined to fund overtime pay for off-roster volunteers, citing budget constraints, even though Captain Lugo presented the data from last winter as evidence.
Question: Based on the passage, if a similar three-call night occurs this winter, what is most likely to happen?
- A. The department will handle the responses without delays, as it did last winter.
- B. Captain Lugo will personally respond to all three calls himself.
- C. The department will struggle to staff the responses, because the roster is smaller and backups are unavailable.
- D. The town council will reverse its decision and fund overtime pay.
The correct answer is C, and it rests on several pieces of evidence: (1) the roster has dropped from 22 to 15; (2) two members are recovering from surgery, further reducing available responders; (3) the town council declined to fund overtime, closing off the backup mechanism; (4) last winter's success depended on calling in four backups, which the current roster and funding make unlikely. The prediction follows the passage's internal logic — fewer responders, no overtime funding, and a demonstrated past need for backups combine to predict a staffing problem.
Option A ignores the changed conditions and simply extrapolates last winter's result. Option B is a wishful guess — nothing in the passage suggests one person can cover three calls, and the passage's own example shows backups were used last time. Option D is a different wishful guess — the passage gives no evidence the council will change its mind; in fact, it already declined despite Lugo's data. Both A and D project an outcome the evidence does not support.
A Note on Prediction Strength
D2 predictions come in different strengths. A prediction supported by three strong, independent clues is more confident than one supported by two indirect hints. The HiSET does not ask you to rank predictions by confidence, but it does require you to choose the prediction that the evidence most strongly supports. When two answer choices both seem plausible, look for the one with more independent textual support. The passage usually gives you enough clues to distinguish a likely prediction from a merely possible one.
Checklist for D2 Questions
- Identify the prediction type (plot, character, application).
- Name two or more textual clues that point toward the prediction.
- Check that the prediction follows the passage's internal logic, not outside knowledge.
- Reject any prediction that requires the character or process to act against established patterns.
- Reject any prediction that depends on an outcome the passage gives no evidence for (a council reversing a decision, a person changing a habit).
- When two predictions seem possible, choose the one with more independent textual support.
A D2 question asks: "What is the character most likely to do next?" Which is the strongest basis for a prediction?
A passage describes a student who has studied every night for two weeks, improved on each practice test, and is visibly exhausted the night before the final exam. A D2 question asks what the student is most likely to do. Which prediction is best supported?