Pacing and deciding when to move on
Reading position saved on this browser; this is not a completion record.
6.1 Choose a method by its next useful step
A useful method turns the current information into the requested quantity with manageable work. The fastest method is not always mental calculation, and the most sophisticated method is not always graphing. Decide by the structure of the problem, the form of the desired answer, and your ability to execute the method accurately.
Method | Strong reason to choose it | Main risk to manage |
|---|---|---|
Mental or short arithmetic | The target follows from a known group, simple percent, or clean ratio. | Skipping a sign or reporting an intermediate value. |
Direct algebra | Exact structure, parameter cases, or a simple equation matters. | Expanding unnecessarily or discarding a special case. |
Candidate testing | Few numerical options satisfy a readily evaluated model. | Checking the wrong quantity or missing a qualifier. |
Graphing | Intersections or approximate nonlinear solutions are the natural target. | A misleading window, wrong coordinate, or hidden domain. |
Hybrid method | Structure simplifies the model, but final arithmetic is inconvenient. | Letting an approximate output replace an exact condition. |
Estimate the remaining work, not the appearance of the question
A long story may reduce to one proportion. A short equation with a parameter may need several cases. After writing one good relationship, ask: What is the next useful step? Can I see a reliable path from it to the answer?
If a method is making progress, continue. If it creates cumbersome work, switch only when a specific alternative improves the situation. Opening a calculator, making several unplanned entries, and then starting the algebra again is not a strategy; it is repeated setup cost.
A useful three-way decision
Finish when the path is clear and the remaining work is small. Switch when you can name a better method and its first step. Defer when progress has stopped and other questions need attention.
The examples in this section ask not only “What is the answer?” but also “Why is this route appropriate here?” A method-selection explanation is a training task, not an extra response demanded by the SAT.
6.2 Build a module plan instead of a rigid per-question limit
Under standard timing, SAT Math has 44 questions across two 35-minute modules. Performance in the first module influences the mix of questions in the second.[1] Dividing 35 minutes by 22 questions gives about 95 seconds per question on average, but individual questions need different amounts of time.
Review is module-limited: a flag is a plan to revisit a question while that module remains active, not after moving to the next one. The official student guide describes flags and the question menu as tools for returning within a module.[11]
Use three phases as a flexible plan
First pass: secure clear work. Proceed through the questions, finishing those with a reliable path. Defer a stalled problem with a short restart note. Avoid skipping merely because an item looks unfamiliar; a brief translation may reveal an easy relationship.
Return pass: revisit with a purpose. Choose a deferred question for which you now have a concrete next step. A flagged item with a suspected unit error may be a better first return than one requiring a completely new method.
Completion pass: protect entered answers. Leave a small, practiced reserve to check the question menu, supply your best responses where needed, and audit high-risk entries. A useful starting rehearsal reserve is one to two minutes, adjusted to your actual practice results; it is not an official timing rule.
Check the clock at decisions, not after every line
Use occasional planning checks to compare time remaining with work remaining. A suggested rehearsal is a brief check around 12 and 24 elapsed minutes in a standard 35-minute module, then a final completion check. These are prompts to adapt, not deadlines for particular question numbers. Use your approved module duration when practicing with accommodations.
Do not try to infer the scoring path while testing
A question’s felt difficulty is not a reliable score report. College Board explains that the section score depends on work across both modules, and that equal numbers correct can yield different scores because the questions differ.[6] Use your time to answer the current questions, not to guess your route or reconstruct a scaled score.
A clock is a planning tool
The objective is accurate, completed responses across the module. Saving time on one question is useful only if it creates time for valuable work elsewhere; rushing a nearly finished solution can waste the work already done.
6.3 Use progress-based stopping rules
A universal rule such as “leave every question after exactly 60 seconds” ignores whether a solution is one line from completion or still lacks a model. Instead, use triggers that describe your progress.
What is happening | Recommended next action |
|---|---|
You have a correct model and one short calculation remains | Finish, then perform the relevant check. |
You have reread without creating a relationship | Name the target and one variable; if no useful step emerges, defer. |
The algebra expands rapidly | Check for a repeated group, an expression target, or useful choices. |
A graph does not show the expected feature | Check the entered equation, domain, and window before making a conclusion. |
A second method disagrees with the first | Locate the disagreement in the original conditions; do not average answers or guess between methods. |
You have spent time but cannot name the next step | Save your work and move to a question with a clearer path. |
Leave a restart line on scratch paper
A useful restart line contains the problem number, the model or key restriction, and the next intended action. Examples of note types are “check x ≠ 2,” “compare discriminant and linear case,” or “divide the two growth observations.” Do not rewrite the whole question.
Separate sunk time from useful progress
Time already spent cannot be recovered by spending more. But partial progress may reduce the cost of finishing. The right comparison is future work: How much useful work remains here, compared with the next question? Do not stay simply because you have already invested effort, and do not leave simply because the elapsed time feels large.
A provisional response is different from a flag
Mark for Review, option elimination, and selecting or entering an answer are different actions. Bluebook provides those tools and a question menu.[3] Make sure a selected or entered response actually exists when that is your intention. A flag does not supply an answer.
College Board advises that, for most students making their best effort, guessing is preferable to leaving a question blank, especially after eliminating options.[6] Use mathematical evidence to improve your choice; do not invent a supposed lucky option or assume a guessing pattern guarantees a score.
6.4 Review targeted risks and calibrate with evidence
A good review pass has a question to answer. “Does this root satisfy the original radical equation?” is useful. “Should I solve this again because I am nervous?” is less focused.
Flag reason | Focused return task |
|---|---|
Unsure about the requested quantity | Reread the final sentence and compare it with the entered value. |
Cleared a denominator or squared both sides | Test the candidate in the original equation. |
Used a rounded output near a threshold | Check the neighboring allowable values. |
Parameter changes a leading coefficient | Analyze the reduced-degree case separately. |
Selected from a graph | Confirm the requested coordinate, domain, and relevant window. |
Combined rates or percentages | Reconstruct the denominator, group sizes, or changing bases. |
Change an answer for a reason
Keep a response when it survives a relevant check. Change it when you find a specific error or a missing condition. There is no mathematical principle saying that a first answer is always best or that a longer solution is more trustworthy.
Track accuracy and avoidable time separately
After fresh practice, record whether each question was correct, whether the method was appropriate, where you stalled, and what a better first step would have been. A slow correct answer is a method-training opportunity; a fast wrong answer needs accuracy work, not more speed.
Try one change at a time in the next practice session: a target statement before modeling, a clearer flag reason, or an earlier switch from blind expansion to substitution. Small informal practice samples do not establish that one timing routine caused a score change; they help you identify hypotheses to test with further work.
Rehearse in the official environment
The mixed sets here exercise strategy without predicting a scaled score. Use a fresh full-length Bluebook test for adaptive digital rehearsal, and the official Student Question Bank for additional targeted skill practice.[10] [12] Keep some fresh material so that recalled answers do not masquerade as improved method selection.
Section 6 exit standard
You can choose a method deliberately, recognize progress, leave a useful restart line when deferring, and use the final review for specific risks. The five workflow scenarios are instructional examples, not one-size-fits-all test-day scripts.
15 worked examples
6.01. Finish a short structural calculation without tool setup
If 9(2x − 1) = 63, what is 2x + 3? Choose an efficient method and find the answer.
Show worked solutionHide worked solution for example 6.01
Recognize. The given group 2x − 1 differs from the target by 4. Direct arithmetic is enough.
Work. Divide by 9 to get 2x − 1
Answer: 11, by direct arithmetic on the known group.
Check. Solving for x would give x
Avoid the trap. Opening a graph or isolating x is not wrong, but neither is necessary. The efficient stopping point is the requested expression, followed by a quick target check.
6.02. Combine like percentage parts before calculating
Fifteen percent of a positive number n, plus 35% of the same number, equals 40. What is n?
Show worked solutionHide worked solution for example 6.02
Recognize. The two percentages have the same base and are being added, so their rates can be combined.
Work. Together they are 50% of n. Thus 0.50n
Answer: 80.
Check. Fifteen percent of 80 is 12, and 35% of 80 is 28; their sum is 40.
Avoid the trap. Combining percentages is valid here because both parts use the same number. It is not the same situation as a percentage increase followed by another percentage change.
6.03. Avoid finding an unneeded variable
The system x − 2y = 7 and 3x + 2y = 13 is given. Find 2x.
Show worked solutionHide worked solution for example 6.03
Recognize. Addition cancels y and gives a multiple of the requested quantity. Elimination is more direct than graphing two full lines.
Work. Add to obtain 4x
Answer: 10.
Check. The system’s solution is x
Avoid the trap. Reporting 5 would correctly identify x but fail to answer the question. The target should remain visible through the last line.
6.04. Use a graph when the requested result is approximate
The graphs of y = 2x2 − 3x − 7 and y = 4x + 5 intersect. What is the greater intersection x coordinate, to the nearest hundredth?
Show worked solutionHide worked solution for example 6.04
Recognize. The target is an approximate nonlinear intersection. Graphing the two equations and reading the greater x coordinate is reasonable; exact algebra can verify it.
Work. Equating outputs gives 2x2 − 7x − 12
x
The greater value is approximately 4.7604, so it rounds to 4.76.
Answer: 4.76.
Check. Its corresponding y value is approximately 24.0416. A graph window must include that height to show the intersection.
Avoid the trap. The requested coordinate is x, not y. A default window may miss the point even though the equations were entered correctly.
6.05. Test a few radii instead of expanding a cubic
A cylinder has radius r and height 2r + 1, in centimeters. Its volume is 144π cubic centimeters. Which value is r? A) 2 B) 3 C) 4 D) 5
Show worked solutionHide worked solution for example 6.05
Recognize. Testing four positive candidates in r2(2r + 1)
Work. At r
Answer: C, 4.
Check. The height is 9, and π(4)2(9)
Avoid the trap. If you start expanding and then stall, switch to the finite choices with a clear reason. A switch should simplify the remaining work rather than merely restart it.
6.06. Choose algebra when a default graph window hides the roots
What is the smaller zero of f(x) = (x − 40)2 − 9?
Show worked solutionHide worked solution for example 6.06
Recognize. The vertex form already isolates a square. A graph’s default window near the origin may show neither root.
Work. Set (x − 40)2
Answer: 37.
Check. f (37) = (−3)2 − 9 = 0. The two zeros are symmetric about x
Avoid the trap. No visible intercept in a small window does not imply no real zero. Here algebra is both exact and shorter than adjusting the display. A window containing x
6.07. Reconstruct a total instead of entering a data list
A data set of n values has mean 18. After the values 23 and 27 are added, the new mean is 19. What is n?
Show worked solutionHide worked solution for example 6.07
Recognize. The original data values are not needed. The mean and count determine their total.
Work. The original total is 18n. After adding two values, the total is 18n + 50 and the count is n + 2. Therefore 18n + 50 = 19(n + 2) n = 12.
Answer: 12.
Check. The initial sum is 216; the new sum is 266 across 14 values, and 266/14
Avoid the trap. A statistics calculator cannot reconstruct an unspecified list from its mean alone, but the requested count is determined by the total relationship. Modeling is the bottleneck, not arithmetic.
6.08. Use cases instead of guessing a parameter from a slider
For which real k does (k + 1)x2 + 4x + 4 = 0 have exactly one real solution in x?
Show worked solutionHide worked solution for example 6.08
Recognize. The leading coefficient can be zero. A graph or slider may suggest some cases, but algebra gives the exact values and catches a reduced-degree equation.
Work. At k
Answer: k
Check. The equations become 4x + 4
Avoid the trap. A single tangency found by graphing does not prove that every relevant parameter case has been found. Classify the degree first.
6.09. Recognize a target square that avoids irrational roots
A real number x satisfies x2 − 7x + 3 = 0. What is (2x − 7)2?
Show worked solutionHide worked solution for example 6.09
Recognize. Expanding the target creates four times the variable part of the given equation. There is no need to calculate either irrational root.
Work. The equation gives x2 − 7x
(2x − 7)2 = 4x2 − 28x + 49 = 4(x2 − 7x) + 49 = −12 + 49 = 37.
Answer: 37.
Check. The quadratic formula gives x
Avoid the trap. Taking a decimal root first and squaring a rounded expression creates needless precision risk. The target is exact and the same for both roots.
6.10. Combine approximate solving with an original-sign check
Solve
= x and report the solution to the nearest tenth. Choose a method that keeps the domain visible. 2x + 9
Show worked solutionHide worked solution for example 6.10
Recognize. The solution must have x ≥ 0. Either graph the two sides or square carefully; the final answer is approximate, but the sign restriction is exact.
Work. Squaring gives x2 − 2x − 9
Answer: 4.2.
Check. For the exact positive candidate, 2x + 9 = 11 + 2
Avoid the trap. A graph can help locate the positive intersection, but a rounded value is not an exact algebraic solution. Verify the exact candidate, then round as requested.
6.11. Allocate the remaining time without imposing equal deadlines — Workflow scenario
In a standard-timing practice module, six minutes remain. Four questions have no response, and two answered questions are flagged. You plan to protect the final minute for a completion scan. How much solving time remains, and what is a reasonable order of work?
Show worked solutionHide worked solution for example 6.11
Recognize. The reserve leaves a working budget, but the remaining questions may have very different costs.
Work. Six minutes is 360 seconds. After reserving 60 seconds, 300 seconds remain for solving, or an average of 300/4
Answer: 300 working seconds; 75 seconds per unanswered question on average, not a mandatory limit.
Check. The full plan fits the remaining 360 seconds and preserves time to confirm responses. Adjust if a nearly finished item can be completed quickly.
Avoid the trap. Spending all six minutes on the first difficult question sacrifices the planned completion pass. A flag and a response are separate states.
6.12. Respond to stalled progress rather than sunk time — Workflow scenario
You have spent 150 seconds on a parameter problem. You have written a correct equation but have repeated the same unsuccessful manipulation twice. Several unvisited questions remain. What is a sensible next action?
Show worked solutionHide worked solution for example 6.12
Recognize. The reason to defer is the absence of a useful next step, not the number 150 itself.
Work. Briefly inspect the equation for a degree-changing parameter, a repeated group, or a useful target combination. If no concrete route emerges, preserve the equation and a short restart note on scratch paper, mark the question for review, record your best available response, and continue.
Answer: Defer with a restart line unless a specific, short path has now become clear.
Check. The saved work reduces the setup cost when you return. The decision protects opportunities on unvisited questions without discarding useful progress.
Avoid the trap. “I have already spent too long to stop” treats past time as recoverable. Compare remaining work, not past effort.
6.13. Protect answer transfer in the final pass — Workflow scenario
With 85 seconds remaining, one numeric-response field is blank even though your scratch work has established the answer 7/12. One flagged item has a suspected unit-conversion error. All other questions have responses. Propose a completion plan.
Show worked solutionHide worked solution for example 6.13
Recognize. An established answer that has not been entered is an immediate, specific task. The flagged item has a defined check, not an open-ended need to start over.
Work. First enter 7/12 in the blank field and confirm it appears. One reasonable rehearsal allocation is 20 seconds for that entry and check, up to 40 seconds for the suspected conversion, and the remaining 25 seconds for a final question-menu scan. These are planning examples, not required timings.
Answer: Transfer the known answer first, check the named risk second, and preserve a final scan.
Check. The allocations total 85 seconds. If either task finishes early, use the saved time deliberately rather than starting an unbounded second solution.
Avoid the trap. A correct result on scratch paper does not submit a response. Keep calculation, flagging, and answer entry distinct.
6.14. Change a response when a missing case is verified — Workflow scenario
A question asks for the sum of all real k for which (k − 3)x2 + 4x + 4 = 0 has exactly one real solution. You entered 4 after using the discriminant. During review, you notice that k = 3 makes the equation linear. Should the response change?
Show worked solutionHide worked solution for example 6.14
Recognize. There is a specific mathematical omission, not merely a feeling of doubt.
Work. At k
Answer: Yes. Change the response to 7.
Check. The two reduced equations are 4x + 4
Avoid the trap. “Never change your first answer” is not a mathematical rule. A verified missing case is evidence for correction; vague anxiety is not.
6.15. Use practice evidence without inventing a score prediction — Workflow scenario
You complete three fresh, nonadaptive 12-question strategy sets. Your results are shown below. What do these results support, and how should you plan the next review?
Set
Minutes used
Correct out of 12
A
18
10
B
21
11
C
19
10
Show worked solutionHide worked solution for example 6.15
Recognize. The results describe performance on these practice questions. They do not supply a calibrated SAT scaled score or prove a causal effect of taking more time.
Work. Across the sets, you answered 10 + 11 + 10
Review the five missed questions and any unusually slow correct ones. Classify each issue as modeling, method choice, computation, constraint checking, or answer transfer. Choose one specific adjustment for the next fresh set, such as writing a target statement or checking a vanishing leading coefficient before using a discriminant.
Answer: 31/36 correct and a mean time of 19 minutes 20 seconds describe this practice; use the error patterns to choose the next intervention.
Check. Set B was both slower and more accurate, but the questions differed. Three informal sets do not establish that extra time caused the difference or that a particular SAT score will follow.
Avoid the trap. Do not convert 86.1% into a predicted SAT score or assume that memorized retest answers demonstrate new strategy skill. Use official adaptive practice for a fuller rehearsal.[6] [10]
Transfer. A productive review produces a testable next behavior: “Before clearing a denominator, I will record its excluded input.” “Be more careful” is too vague to guide the next decision.
6.6 Section exit check
For E6.1–E6.4, name a suitable method and solve. E6.5 is a planning task, not an SAT mathematical item.
Review prompt
Could you state the next useful step before committing to a tool? Was your stopping point the actual requested quantity? Did you preserve time to transfer and check responses?