Psat Math
Read this first
Read the Overview and Method chapters once, keep the Trap Codex and By the numbers pages open while you practice, and work the Practice set as a mini-diagnostic. Every number here is verified against the official specification; every practice item is an AskSia original that mirrors the exam.
What PSAT/NMSQT Math is
The Math section of the PSAT/NMSQT is the second half of the test, delivered digitally in Bluebook after Reading & Writing. It runs as two modules of 22 questions and 35 minutes each — 44 questions and 70 minutes in all. Of those 44, 40 are scored; each module carries 2 unscored pretest items you cannot identify, so every question deserves the same effort.
A.How the section behaves
- Multistage adaptive: Module 1 mixes difficulties, and your performance there routes you to an easier or harder Module 2. Math adapts independently of Reading & Writing.
- Free movement within a module — flag, skip, and return as you like. But there is no backtracking once a module closes.
- At 35 minutes for 22 questions, the average budget is about 1 minute 35 seconds per question — generous by design, so the constraint is accuracy, not raw speed.
B.Two question formats
- Multiple choice: four options, one correct — the majority of the section.
- Student-produced response (SPR): you type a number or fraction with no options to lean on — roughly 25% of scored questions (8–12 of the 40). No choices means no backsolving and no elimination; your setup has to stand on its own.
- Stimuli span prose, tables, informational graphics, equations, and geometric figures — nothing beyond what the digital SAT suite already uses.
The blueprint tilt: how PSAT Math differs from SAT Math
PSAT/NMSQT Math tests the same four content domains as the digital SAT, and the two tests draw on a shared item bank — techniques transfer directly. What changes is the mix. The official framework tilts the PSAT toward data analysis and away from the hardest algebraic and geometric content.
| Domain | PSAT weight | Questions (of 40 scored) | vs digital SAT |
|---|---|---|---|
| Algebra | 35% | 13–15 | Same |
| Advanced Math | 32.5% | 12–14 | Lower (SAT: 35%) |
| Problem-Solving & Data Analysis | 20% | 7–9 | Higher (SAT: 15%) |
| Geometry & Trigonometry | 12.5% | 4–6 | Lower (SAT: 15%) |
1.What the tilt means for prep
- Algebra is still king at 35% — linear equations, systems, and linear functions remain the highest-yield territory on either test.
- Data analysis punches harder here. At 20% versus the SAT's 15%, ratios, rates, percentages, two-way tables, and probability earn a bigger slice of your study time than SAT-focused plans would give them.
- Geometry & trig is the smallest domain at 12.5% (4–6 questions) — worth learning well, not worth learning first.
2.What is deliberately left out
- Trigonometry is right-triangle only. There is no unit circle, no radians, no obtuse or negative angles anywhere on the PSAT.
- No "evaluating statistical claims" — that SAT skill is dropped from a faithful PSAT/NMSQT form, and margin of error is down-scoped to inference from sample statistics.
- These cuts are why the scale tops out at 760 instead of 800: the ceiling content simply is not asked.
What a Math point is worth
Math is scored 160–760 in 10-point steps and adds, unweighted, to Reading & Writing for the 320–1520 total. For National Merit, though, the two sections are not equals: the Selection Index doubles Reading & Writing and counts Math once.
Example: RW 690, Math 720 → (1380 + 720) ÷ 10 = 210
1.The honest arithmetic
Because Math enters the formula once, it can contribute at most 76 of the 228 Selection Index points, and a +10 Math gain moves the index by 1 where a +10 RW gain moves it by 2. If National Merit is your goal, Reading & Writing is the higher-leverage section — this book says so plainly.
2.Why Math still pays twice
A Selection Index point is a Selection Index point, whichever section supplies it — near a cutoff, a Math point counts exactly as much as an RW point. And every Math skill you build here transfers intact to the SAT: shared item bank, shared scale, same Desmos, same reference sheet. PSAT Math prep is SAT Math prep with the price of admission already paid.
- Learn the section's shape first: 2 modules × 22 questions × 35 minutes, adaptive routing, no backtracking between modules.
- Order your study by the blueprint: Algebra 35% → Advanced Math 32.5% → Problem-Solving & Data Analysis 20% → Geometry & Trig 12.5%.
- Train both formats: practice SPR items without answer choices until typing a bare number feels routine — they are roughly a quarter of the scored section.
- Make Desmos a habit, not a rescue: use it from the first practice set so calculator judgment is automatic on test day.
Accuracy first, Desmos early, type clean
PSAT Math is two modules of 22 questions in 35 minutes each — 44 questions, 70 minutes, of which 40 are scored. Divide it out and the section hands you about 1 minute 35 seconds per question: 35 × 60 ÷ 22 ≈ 95 seconds. That is a generous budget, and it dictates the whole method — this section is lost to unforced errors, not to the clock.
Rule 1 — Respect the 95-second reality
Almost no one runs out of time on PSAT Math the way they do on Reading & Writing (which allows roughly 71 seconds per question). Points die differently here: a sign dropped in step two, a percent taken of the wrong base, an answer to a question that was never asked. So the method is deliberately unheroic: work at the speed at which you do not make mistakes, and let the pacing math prove you can afford it.
Bank time where it is cheap. A one-step linear equation should cost 30–45 seconds; that surplus funds the multi-step word problems and data questions that genuinely need two to three minutes.
Rule 2 — Module 1 discipline: easy first, because the router is watching
The PSAT is multistage adaptive: Module 1 is a broad mix of difficulty, and your performance on it routes you to an easier or harder Module 2 — Math routes independently of Reading & Writing. Since the harder second module is where the top of the 160–760 scale lives, Module 1 accuracy is the highest-leverage 35 minutes of the section.
Run Module 1 in two passes. Pass one: answer everything you can solve cleanly in about 90 seconds, and flag the rest without agonizing. Pass two: return to the flags with your banked time and full attention. You may move freely and revisit any question within a module — but once Module 2 starts, Module 1 is sealed forever, so leave nothing blank before the clock ends.
Rule 3 — Desmos first, not last
The built-in Desmos graphing calculator is available for the entire Math section — both modules, every question, alongside the in-app geometry reference sheet. Treat it as a solving engine, not an arithmetic backstop:
- Plot both sides. For any equation, graph the left side and the right side as two separate curves — for x2 - 5x + 4 = 2x - 6, plot y = x2 - 5x + 4 and y = 2x - 6 together.
- Read the intersections. The first coordinate of each intersection point is a solution. Zeros, vertices, and intercepts are equally clickable — no factoring under pressure.
- Systems are a click. Graph both equations of a linear system and tap the crossing point; a two-minute elimination becomes a 30-second lookup.
- Verify typed answers. Before submitting a student-produced response, substitute your value back into the original expression in Desmos and confirm it.
- Know when to skip it. For a one-step equation, typing costs more than thinking — at ~95 seconds per question, Desmos on trivial algebra is a net loss.
Rule 4 — SPR hygiene: type the value, nothing else
About 25% of the scored Math questions — roughly 8 to 12 of the 40 — are student-produced responses (SPR): no choices, you type the answer. The box accepts a number or a fraction, full stop. No percent signs, no dollar signs, no units, no commas, no variables — strip everything but the value. A correct solve entered as 67.5% instead of 67.5 is a wasted point.
Before you submit, run the three-question sanity check from the nonlinear-functions playbook: could this answer legitimately be negative? Could it be a fraction? Could it be zero? SPR answers are allowed to be all three, and talking yourself out of an ugly-looking correct value is a classic unforced error.
| You found… | Type this | Not this |
|---|---|---|
| A cost of $123 | 123 | $123 |
| 67.5 percent | 67.5 | 67.5% |
| Seven twenty-fourths | 7/24 | a hand-rounded decimal — the fraction is exact |
Rule 5 — The method, applied
Three worked examples, one from each of the section’s load-bearing zones: the biggest domain (Algebra, 35%), the domain the PSAT tilts heavier than the SAT (Problem-Solving & Data Analysis, 20% vs 15%), and the ceiling of PSAT trigonometry.
Problem. A print shop charges a one-time setup fee plus a fixed price per poster. An order of 12 posters costs $69, and an order of 30 posters costs $150. How much does an order of 24 posters cost?
Lock the rate and the start before computing. A linear model f(x) = mx + b is a constant-rate machine: decide what the rate and the starting value mean in this problem’s units first. The rate is dollars per poster: (150 - 69)/(30 - 12) = (81)/(18) = 4.50. The starting value is the cost at zero posters — the setup fee: 69 - 12(4.50) = 69 - 54 = 15.
Then answer the question asked. The model is y = 4.50x + 15, so 24 posters cost 4.50(24) + 15 = 108 + 15 = 123 — the order is $123. Desmos check: graph y = 4.50x + 15 and confirm it passes through both given orders before trusting the readout at 24.
Problem. A jacket priced at $140 is marked down 25%. At checkout an additional 10% is taken off the reduced price. The final price is what percent of the original price?
Every percent acts on its own base. After the markdown the price is 140 × 0.75 = 105. The checkout discount applies to that reduced price, not the original: 105 × 0.90 = 94.50. As a share of the original, 94.50 ÷ 140 = 0.675 — the customer pays 67.5%. In an SPR box, that is typed 67.5, no symbol.
The trap has a name. Adding the discounts — 25 + 10 = 35, "so 65% of original" — treats both percents as if they act on the $140 base. They do not, and that wrong value is exactly what a rushed solver produces. Most data-domain errors are base and denominator errors, not arithmetic errors: before touching numbers, say out loud what each percent is a percent of.
Problem. In a right triangle, the sine of one acute angle is (7)/(25). What is the tangent of that same angle?
Anchor the triangle before any ratio. Sine is opposite over hypotenuse, so build the cleanest triangle that fits: the side opposite the angle measures 7 and the hypotenuse measures 25 (any similar triangle gives the same ratios). The remaining leg — the side adjacent to the angle — comes from the Pythagorean theorem: √(252 - 72) = √(625 - 49) = √(576) = 24.
Now the ratio is a read-off. Tangent is opposite over adjacent: (7)/(24). Enter it as the fraction 7/24 — exact, no rounding decision. The whole solve is naming opposite, adjacent, and hypotenuse relative to the angle in use; the arithmetic is one theorem the in-app reference sheet already supports.
- 1 · Pace: ~95 s per question — accuracy is the strategy.
- 2 · Module 1: easy first, two passes; the router is watching.
- 3 · Desmos: plot both sides, read the intersections.
- 4 · SPR: type the bare value — fraction or decimal, no symbols.
- 5 · Apply: rate-and-start, mind the base, anchor the triangle. The Trap Codex (next chapter) names every way these go wrong.
The named traps — and how to catch them
A wrong answer on PSAT Math is rarely random — it is a designed trap with a name. The table below is drawn from AskSia’s curated trap graph: each entry names the wrong move, why it feels right in the moment, and the tell that catches it. Recognizing a trap by name is the fastest accuracy gain there is: you stop falling for a pattern, not just fixing one question.
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Linear Equations in One Variable | ||
| One side only | Performing an operation on only one side (e.g., subtracting 5 from the left but not the right). Feels right because: Students focus on 'getting rid of' a term and forget the mirror operation on the other side. | ✓ Substitute the final x back into the ORIGINAL equation; if the two sides differ, a balance step was skipped. |
| Wrong inverse | Using the wrong inverse — e.g., dividing to undo a subtraction, or adding to undo a multiplication. Feels right because: Under time pressure the operation pair (+/-, x/div) gets mismatched. | ✓ Name each term's operation on x and pair it with its true inverse before moving it. |
| Divide before subtract | Dividing by a before moving b: from 2x + 6 = 10, writing x + 6 = 5 (only divided the 2x term). Feels right because: Students see the coefficient and rush to divide, distributing the division to only one term. | ✓ If you divide early you must divide EVERY term on both sides; check that 6 became 3 too. Easier: subtract first. |
| Linear Equations in Two Variables | ||
| Swap x y | Substituting the y-value into x and vice versa when checking an ordered pair (a,b). Feels right because: Students grab the two numbers without tracking which is the input (x) and which is the output (y); the pair looks symmetric. | ✓ If a 'solution' fails the check, retry with the coordinates swapped before discarding it; always label (x, y) explicitly. |
| Only one solution belief | Believing a two-variable equation has a single numeric 'answer' the way a one-variable equation does, instead of infinitely many ordered-pair solutions. Feels right because: All prior practice ('solve for x') trained them that an equation yields one number; the jump to a whole line of solutions is conceptual. | ✓ Ask 'how many points are on a line?' — infinitely many; a single equation in two unknowns does not pin down one pair unless a second condition is given. |
| Substitute into wrong variable | When told 'when x = 4', plugging 4 in for y (or solving as if the given number were the output), so the wrong variable is fixed. Feels right because: Word problems often phrase the given quantity in the order y-then-x or bury 'x' far from its value, so students substitute by position, not by name. | ✓ Underline the named variable in 'when [var] = value' and put the value only where that letter sits; re-read to confirm which is the input. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Linear Functions | ||
| F times x | Reading f(x) as multiplication: treating f(3) as f·3 or thinking 'f' is a variable you can cancel/solve for. Feels right because: Parentheses-after-a-letter everywhere else in algebra means multiplication, e.g. 3(x+1), so the brain pattern-matches f(x) the same way. | ✓ Student writes f(3) = 3f or tries to divide both sides by f; or computes f(2)+f(3) by 'factoring out f'. |
| Input output swap | Confusing 'find f(5)' (input given, find output) with 'find x when f(x)=5' (output given, find input) and solving the wrong direction. Feels right because: The number 5 appears in both phrasings; students grab the number without checking whether it sits inside the parentheses (input) or after the equals (output). | ✓ For f(x)=2x+1, asked f(5) student answers 2 (solved 2x+1=5) instead of 11; or asked 'f(x)=5, find x' student answers 11. |
| Order of operations reverse | Dividing by m before subtracting b (e.g. for 3x+6=15 dividing everything by 3 incorrectly), or only subtracting b without dividing by m. Feels right because: Students rush the two undo-steps and skip or reorder them; with a leading coefficient the temptation to divide first is strong. | ✓ Plug the candidate x back into f; if f(x)≠c the inversion steps were wrong or out of order. |
| Systems of Two Linear Equations | ||
| Solve only one equation | Student finds a point on just one line (e.g. plugs in x=0) and calls it 'the solution' without checking the other equation. Feels right because: One equation alone has infinitely many (x,y) pairs that 'work', so any of them feels like an answer. | ✓ A true solution must satisfy BOTH equations — substitute back into the equation you didn't use. |
| Confuse xy roles | Reading the intersection as (y,x) — swapping which number is x and which is y. Feels right because: When skimming, the order of coordinates is easy to flip, especially copying off a Desmos popup. | ✓ Horizontal value is always x; substitute your (x,y) back to confirm it satisfies both lines. |
| Read axis intercept instead | Reading where ONE line crosses an axis (its intercept) as the solution, instead of where the TWO lines cross each other. Feels right because: Axis crossings are visually salient (they sit on the gridlines), so the eye lands on them first. | ✓ The solution is the point shared by BOTH lines, not where either line meets the x- or y-axis. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Linear Inequalities | ||
| Forgot flip | Student divides or multiplies both sides by a negative coefficient but does NOT reverse the inequality sign (e.g. -2x < 6 → x < -3 instead of x > -3). Feels right because: It mirrors equation-solving exactly, which students have practiced far more; the flip is the single rule that breaks the 'inequalities = equations' habit, so it is the most-forgotten step. | ✓ Whenever the variable's coefficient becomes negative at any step, check: did the sign flip exactly once? Sanity-test by plugging a number from your answer set back into the original inequality. |
| Flip on positive | Student flips the sign even when dividing/multiplying by a positive number, or flips it twice (over-correcting after learning the rule). Feels right because: After being drilled 'flip the sign,' some students over-generalize and flip on every multiply/divide, or flip an extra time when they also moved a negative term across. | ✓ Flip ONLY when the number you multiply/divide by is itself negative. Moving a term across (adding/subtracting) never flips. Count: at most one flip per negative-coefficient division. |
| Only one part | In a < bx + c < d, student applies an operation (subtract c, divide by b) to only the middle and one side, not all three parts. Feels right because: Students mentally treat it as two separate inequalities but then only finish one, or forget the leftmost bound exists. | ✓ Write the three parts in a column and apply each operation to all three rows simultaneously; the variable should end isolated in the middle with a number on each side. |
| Equivalent Expressions | ||
| Sets equal to zero | Student sets the expression equal to 0 (or to a number) and 'solves for x', producing a numerical answer instead of an equivalent expression. Feels right because: Most prior practice was solving equations, so the reflex is to set things equal and isolate x. | ✓ If your answer is a single number but the choices are all expressions in x, you solved instead of rewrote — back up. |
| Partial domain check | Student tests only x=0 or x=1, where many wrong forms accidentally agree, and concludes equivalence. Feels right because: x=0 and x=1 are easiest to compute, so they're the default test values. | ✓ Always test at least one 'awkward' value (e.g., x=2 or x=-1); if two forms agree at 0 and 1 but differ at 2, they're NOT equivalent. |
| Combine unlike powers | Adds terms of different degree, e.g., 3x²+5x=8x³ or 8x², treating any two x-terms as combinable. Feels right because: Both terms contain x, so they look like 'the same kind'. | ✓ Like terms must match BOTH the variable and its exponent exactly; x² and x are different species and cannot merge. |
Advanced Math & Problem-Solving
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Nonlinear Equations & Systems | ||
| Sign flip root | Reading the root with the wrong sign — writing x=3 from the factor (x+3) instead of x=−3. Feels right because: Students copy the number inside the parenthesis instead of setting the factor to zero and solving. | ✓ Substitute the root back: if (x+3) and you claim x=3 gives 3+3=6≠0, it fails. |
| Divide by x | For x²=5x, dividing both sides by x to get x=5 and losing the root x=0. Feels right because: Dividing looks like the natural simplification, and it does give one valid root. | ✓ Whenever you divide by a variable, check x=0 separately — move everything to one side and factor instead. |
| Sign of b | Plugging b instead of −b, or mishandling the sign of b when b is negative. Feels right because: The −b at the front is easy to forget, especially with a negative b creating a double negative. | ✓ Write a,b,c with their signs in a box first, then substitute; sum of roots should equal −b/a. |
| Nonlinear Functions | ||
| Linear vs exponential confusion | Treating a table with a constant ADDED difference as exponential, or a constant RATIO as linear. Feels right because: Both tables 'grow,' and students check the first two rows then assume the pattern without verifying difference vs ratio. | ✓ Compute BOTH successive differences and successive ratios across ALL rows: equal differences => linear, equal ratios => exponential. |
| Unequal x spacing | Reading differences/ratios straight down a table whose x-values are NOT equally spaced (e.g. x = 0, 1, 3, 4), so the 'rule' looks broken or is mislabeled. Feels right because: Tables 'look' like they always step by 1, so students never check the x column spacing before comparing y's. | ✓ First confirm the x-values increase by a CONSTANT step. The clean rule: only compare rows whose x-steps are equal. If you must use uneven steps, divide each y-DIFFERENCE by its x-step for linear (slope), or take the x-step-th ROOT of each y-RATIO for exponential — a raw ratio across a gap of 2 is not the per-unit factor. |
| Vertex h sign | Reading the vertex x-coordinate as +h directly from (x-h)² when it's written (x+3)², giving h=+3 instead of h=-3. Feels right because: The visible number is 3, and the minus-sign flip in the formula is easy to forget under time pressure. | ✓ Set the inside of the square equal to zero: x+3=0 => x=-3. The vertex x is whatever makes the squared term zero. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Ratios, Rates, Proportions & Units | ||
| Ignoring missing third category | When a ratio covers only some categories (e.g. red:blue = 2:3 but there are also green items), student assumes the whole = red+blue and forgets the unmentioned category. Feels right because: The given ratio feels complete; students assume every item is accounted for by the ratio. | ✓ Re-read for words like 'the rest', 'remaining', 'also', or a third color/group; if present, the whole is larger than the ratio sum. |
| Forgot to sum shares | Student multiplies the total by a ratio term directly (e.g. for 2:3 of 50, computes 50·2 or 50·2/3) instead of dividing by the SUM of terms first (50·2/5). Feels right because: The total and the ratio numbers are both visible; multiplying them feels like the natural next step, and the 'divide by the sum' step is invisible. | ✓ Count the shares first: 2:3 means 5 equal shares. Each part is (term/sum)·total. Verify the parts add up to the original total. |
| Adding instead of multiplying | To scale 2:3 so the first term becomes 6, student adds 4 to both terms (getting 6:7) instead of multiplying both by 3. Feels right because: Adding the same amount 'keeps things fair', and additive thinking is the default before proportional reasoning matures. | ✓ Check: does the ratio of the two new terms still equal the original? 6:7 ≠ 2:3, so it's wrong. Equivalent ratios require multiplication. |
| Percentages | ||
| Decimal off by 100 | Converting 35% to 3.5 or 0.035 instead of 0.35 — moving the decimal point one place too few or too many. Feels right because: 'Divide by 100' is done by eye; a single-digit slip in moving the decimal is easy under time pressure, and both 3.5 and 0.035 'look like' a percent answer. | ✓ Sanity-check: a 'normal' percent (1%–99%) must become a decimal between 0.01 and 0.99. If your decimal is ≥1 or has 3+ leading zeros, you moved the point wrong. |
| Over 100 pct | Treating a percent over 100% (e.g. 150%, 250%) as impossible or capping it at 100% — failing to write 150% = 1.5. Feels right because: Early schooling frames percent as 'a part out of a whole', so students intuit a 100% ceiling and don't accept 'more than the whole'. | ✓ Whenever a quantity grows beyond the original (e.g. 'is now 150% of last year'), expect a factor >1. Reread: 'of' the original is allowed to exceed 100%. |
| Multiply raw percent | Multiplying by the raw percent number instead of the decimal: computing 35·240 = 8400 instead of 0.35·240 = 84. Feels right because: Students grab the two numbers in the problem and multiply, skipping the ÷100 step because the '%' symbol got dropped mentally. | ✓ Estimate: 35% of 240 must be roughly a third of 240 ≈ 80, not thousands. If your answer is bigger than the base for a sub-100% percent, you forgot ÷100. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| One-Variable Data: Distributions & Measures | ||
| Mode for median | Reporting the mode (most-frequent value) when the median (middle value) was asked, or vice versa. Feels right because: On a dot plot or frequency table the tallest stack/biggest count visually dominates, so students grab 'the most common value' regardless of which measure is named. | ✓ Underline the measure word in the question; if it says 'median' you must order and count to the middle, NOT pick the tallest column. |
| Wrong count | Dividing the new sum by the old number of data points after a value was added or removed (e.g. dividing 6 values' sum by 5). Feels right because: The problem foregrounds the original count, so students anchor on it and forget that adding a value also changes the denominator. | ✓ Count the data values again AFTER the change; the denominator must equal the current number of values, not the starting number. |
| Arithmetic slip | Adding the values incorrectly (missing one, double-counting a repeat) so the sum is wrong. Feels right because: Repeated values like 15, 15 invite either skipping the duplicate or counting it once when it should count twice. | ✓ Cross off each value as you add it; on Desmos type every value individually including duplicates. |
| Two-Variable Data: Models & Scatterplots | ||
| Assume step one | Assuming each gridline equals 1 when the axis is actually labeled in steps of 2, 5, 10, 100, or 1000. Feels right because: Step-of-1 is the most common grid in homework, so students count squares instead of reading the printed numbers. | ✓ Always read two adjacent labeled tick numbers and divide by the number of squares between them BEFORE estimating any coordinate. |
| Swap xy | Swapping the x and y variables when interpreting (e.g., treating 'temperature' as x when it is on the y-axis). Feels right because: The variable named first in the sentence is not always the horizontal axis; students map word order to axis order. | ✓ Point to the physical axis label and re-read 'x = ___, y = ___' out loud before reading any point. |
| Direction vs strength | Confusing direction with strength — e.g. calling a steeply-sloped but loosely-scattered cloud 'strong' because the slope is big, or calling a tight near-flat cloud 'weak'. Feels right because: Students equate 'steep' with 'strong'; steepness is direction/rate, tightness is strength, and the two are unrelated. | ✓ Strength = how tightly points cluster around the trend (scatter), NOT how steep the trend is. A gentle slope can be a strong association. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Probability & Conditional Probability | ||
| Flip fraction | Writing total/favorable instead of favorable/total (the fraction upside down). Feels right because: Students grab the two numbers they see and divide in reading order, or put the 'bigger picture' number on top. | ✓ Two sanity checks: (1) a probability can never exceed 1, so a result over 1 (or over 100%) is flipped; (2) the denominator should be the larger 'whole' group, so confirm the bottom number is the total you can point to — this catches flips even when favorable happens to equal total. |
| Miscount total | Using the wrong total — e.g. counting only one column of a table, or forgetting to add a 'totals' row/column already given. Feels right because: The denominator is the part the question hides; the favorable count is usually obvious, the total is where the trap lives. | ✓ Circle the word that defines 'all' (every, total, the whole group) and recount; cross-check against a given grand total. |
| Grand vs row | Dividing by the grand total when the question restricts to one row or column (or vice versa). Feels right because: The grand total in the corner is the most visible number and feels like the default 'whole'. | ✓ Re-read for a restricting phrase: 'of the students who...', 'among the females', 'given that'. If present, the denominator is that subgroup's total, NOT the grand total. |
| Inference from Sample Statistics | ||
| Sample is population | Student treats the sample statistic as if it describes the whole population exactly, dropping the word 'estimate'. Feels right because: The sample number is the only concrete number given, so it feels like 'the answer'. | ✓ Check whether the chosen conclusion is about 'the sample' vs 'all/the population'; the question almost always asks about ALL members. |
| Statistic parameter swap | Mixing up which quantity is known (sample) and which is unknown (population), e.g. saying the population mean 'is' the sample mean. Feels right because: Both are called 'the mean,' so they blur together. | ✓ Read the noun attached to each mean: 'mean length of the sample' vs 'mean length of all newborns'. |
| One sided | Applying the margin of error to only one side (just add it, or just subtract it) instead of both. Feels right because: The estimate looks like a starting point you move from in one direction; one-sided answers are placed as distractors and look 'half right'. | ✓ Check that the answer interval is centered on the estimate with equal distance up and down; if one endpoint equals the estimate itself, it's one-sided. |
Geometry & Trigonometry
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Area & Volume | ||
| Diameter as radius | The problem gives the DIAMETER but the student plugs it in as r, doubling every linear dimension (area off by 4×, volume off by 8×). Feels right because: The number sits right there and 'r' feels like 'the round one's size'; diameter and radius are visually interchangeable in a hurry. | ✓ Whenever you see the word diameter, circle it and immediately write r = d/2 before touching any formula. If your area is suspiciously 4× an answer choice, you used diameter as r. |
| Cone pyramid drop one third | Dropping the ⅓ on a cone or pyramid, computing it like a cylinder or prism (answer comes out 3× too large). Feels right because: The base-times-height pattern is muscle memory from prisms; the ⅓ is easy to skip when copying fast. | ✓ If the shape comes to a point (cone, pyramid), the formula MUST start with ⅓. Cross-check against the reference sheet line, not memory. |
| Hypotenuse leg swap | Treating a given leg as the hypotenuse (or vice versa) — e.g., computing c = √(a² − b²) when c was actually given, so the recovered length is wrong before any area/volume step. Feels right because: In a tilted figure the longest-looking edge isn't always the labeled hypotenuse; students apply a² + b² = c² without identifying which side faces the right angle. | ✓ The hypotenuse is opposite the right-angle mark and is the longest side. If you're solving for a leg, the formula is leg = √(hyp² − otherleg²) — a subtraction, not an addition. |
| Lines, Angles & Triangles | ||
| Vertical vs supplementary | Treating two angles as equal (vertical) when they are actually supplementary (linear pair), or vice versa. Feels right because: The picture has four angles around the crossing and they look symmetric; students grab 'equal' as the default relationship. | ✓ Ask: are the two angles directly across the X from each other (equal) or sharing one ray on a straight line (sum 180°)? |
| Complementary supplementary swap | Confusing complementary (90°) with supplementary (180°). Feels right because: Both words start similarly and both describe 'two angles that add to a special number.' | ✓ Mnemonic: Complementary = Corner (90°, the C); Supplementary = Straight (180°, the S). |
| No parallel assumed | Applying corresponding/alternate-angle equalities when the lines are NOT stated or marked parallel. Feels right because: The diagram looks like two parallel lines, so students assume it. | ✓ Require an explicit 'parallel' statement or matching arrow marks before using these rules; otherwise the angles are unrelated. |
| Trap | The wrong move — and why it tempts | How to catch it |
|---|---|---|
| Right-Triangle Trigonometry | ||
| Hypotenuse misidentified | Student plugs a leg in as c (the hypotenuse) or solves c² = a² - b² instead of recognizing which side is the hypotenuse. Feels right because: When the figure is rotated or not drawn to scale, the longest/hypotenuse side may not look like the slanted bottom they expect; they grab whatever side is labeled c. | ✓ Check that the side you called c is opposite the right-angle mark and is numerically the largest; if your 'hypotenuse' is shorter than a leg, you mislabeled. |
| Forgot square root | Student computes a²+b² and grids that value as the side, forgetting to take the square root to get c. Feels right because: The arithmetic 'feels done' once the two squares are added; the final sqrt is an easy step to drop under time pressure. | ✓ Ask 'is this c or c²?' If your answer is suspiciously large compared to the legs, you skipped the root. |
| Scaling error | Student recognizes 3-4-5 pattern but forgets to multiply all three terms by the same scale factor (e.g., legs 6 and 8 but writes hypotenuse 5 instead of 10). Feels right because: Pattern recognition triggers an automatic '5' before the brain re-scales; the memorized number overrides the actual one. | ✓ Divide each given leg by its base-triple value; the factor must be identical for both legs and then applied to the hypotenuse. |
| Circles | ||
| R squared vs r | Reporting the right-hand constant as the radius (e.g. answering 49 for (x−5)²+(y+3)²=49 instead of 7). Feels right because: The number 49 is sitting right there and looks like 'the answer'; students forget the right side equals r², not r. | ✓ If the right side is a perfect square (4, 9, 16, 25, 36, 49, 64) and the student's radius is that exact number rather than its square root, they skipped the square root. |
| Sign flip center | Reading the center with the wrong signs — e.g. taking (x−5)²+(y+3)² to have center (−5, 3) instead of (5, −3). Feels right because: Students copy the signs they see in the equation rather than the sign of the value that makes the parenthesis zero. | ✓ Set each parenthesis to zero: x−5=0 gives x=5; y+3=0 gives y=−3. If the answer's signs match the literal signs in the equation instead, it's flipped. |
| Forgot balance | Adding (D/2)² to the left to complete the square but forgetting to add the same amount to the right (or to subtract the moved constant), giving a wrong r². Feels right because: Students focus on forming the perfect square and lose track of keeping both sides equal. | ✓ Track every added term: whatever you add to complete a square must reappear on the right (or be subtracted back). Re-graph in Desmos; if the radius doesn't match, the balance step is off. |
PSAT Math by the numbers
| Domain | Weight | Questions | vs the SAT |
|---|---|---|---|
| Algebra | 35% | 13–15 | same |
| Advanced Math | 32.5% | 12–14 | lower (SAT 35%) |
| Problem-Solving & Data Analysis | 20% | 7–9 | higher (SAT 15%) |
| Geometry & Trigonometry | 12.5% | 4–6 | lower (SAT 15%) |
| The tilt that changes your prep: MORE data analysis, LESS advanced math and geometry than the SAT. | |||
PSAT/NMSQT glossary
| Term | What it means |
|---|---|
| Algebra (domain) | The largest PSAT Math domain at 35% of scored questions (13–15 of 40): linear equations in one and two variables, linear functions, systems of two linear equations, and linear inequalities. Its weight matches SAT Math exactly. |
| Advanced Math (domain) | The second-largest PSAT Math domain at 32.5% of scored questions (12–14 of 40): equivalent expressions, nonlinear equations and systems, and nonlinear functions. Slightly lighter than the SAT's 35% weighting of the same domain. |
| Problem-Solving & Data Analysis PSDA | The PSAT Math domain covering ratios, rates, proportions, units, percentages, one- and two-variable data, probability, and inference from sample statistics. At 20% of scored questions (7–9 of 40) it is weighted heavier than the SAT's 15%. |
| Geometry & Trigonometry | The smallest PSAT Math domain at 12.5% of scored questions (4–6 of 40): area and volume, lines, angles and triangles, right-triangle trigonometry, and circles. Lighter than the SAT's 15%, and trig is right-triangle only — no unit circle. |
| Student-produced response SPR | A Math question with no answer choices: you type a numeric or fraction answer. About 25% of PSAT Math questions are SPR (roughly 8–12 per test); the rest are four-option multiple choice. There are no SPR questions in Reading and Writing. |
| Module | One of the two equal parts of the Math section: 22 questions in 35 minutes each. You can move freely within a module but cannot return to Module 1 once Module 2 begins. |
| Term | What it means |
|---|---|
| Multistage adaptive testing MST | The PSAT's adaptive design: your performance on Math Module 1 — a broad difficulty mix — routes you to a second module of higher or lower average difficulty. Adaptation is module-level, not question-by-question, and Math adapts independently of Reading and Writing. |
| Desmos | The graphing calculator built into the Bluebook testing app, available for the entire PSAT Math section. It graphs equations, finds intersections and zeros, and builds tables — often faster than algebraic manipulation on multiple-choice items. |
| Reference sheet | The in-app sheet of geometry formulas (areas, volumes, special right triangles) available on every PSAT Math question, so those formulas never need to be memorized for test day. |
| Pretest question | An unscored question embedded for field-testing: 4 of the 44 Math questions (2 per module) do not count toward your score. They are indistinguishable from scored items, so treat every question as live. |
| Section score | The 160–760 score for Math, reported in 10-point intervals. It adds, unweighted, to the Reading and Writing section score for the 320–1520 total — but counts only once (versus twice for RW) in the National Merit Selection Index. |
| Selection Index SI | The National Merit qualifying number: (RW section score × 2 + Math section score) ÷ 10, ranging 48–228. Math is single-weighted, so a 10-point Math gain adds 1 index point; the Class of 2027 Commended cutoff is 208. |
Frequently asked questions
How many questions are on PSAT Math, and how long is the section?
PSAT Math has 44 questions in 70 minutes, split into two modules of 22 questions and 35 minutes each. Only 40 questions are scored; 4 are unscored pretest questions (2 per module) that you cannot identify. That works out to about 95 seconds per question — noticeably more time per item than Reading and Writing gives you.
What topics does PSAT Math test, and in what proportions?
Four domains, weighted over the 40 scored questions: Algebra 35% (13–15 questions), Advanced Math 32.5% (12–14), Problem-Solving & Data Analysis 20% (7–9), and Geometry & Trigonometry 12.5% (4–6). Linear equations, systems, and functions alone make up over a third of the section, so algebra fluency is the highest-leverage skill.
How is PSAT Math different from SAT Math?
Same digital engine, shifted weights: PSAT Math has more data analysis and less advanced math. Problem-Solving & Data Analysis is 20% on the PSAT versus 15% on the SAT; Advanced Math drops to 32.5% from 35%; Geometry & Trigonometry drops to 12.5% from 15%; Algebra stays at 35%. Trigonometry is right-triangle only — no unit-circle trig appears on any PSAT.
Can you use a calculator on PSAT Math?
Yes — on every question. The Desmos graphing calculator is built into the Bluebook testing app and available for the entire Math section; there is no calculator-free part. An on-screen geometry reference sheet with common formulas is also available throughout. You may bring an approved handheld calculator, but Desmos handles graphing, intersections, and tables directly on screen.
What are student-produced response (SPR) questions on PSAT Math?
About 25% of Math questions are student-produced responses: you type a numeric or fraction answer with no choices offered — roughly 8 to 12 of the section's questions. The rest are four-option multiple choice. SPR items remove the option of testing answer choices, so you must produce the value and enter it in an accepted form.
Is PSAT Math adaptive?
Yes — by module, not by question. It uses multistage adaptive testing: Module 1 is a broad mix of difficulty, and your performance routes you to a second module that is on average easier or harder. Math adapts independently of Reading and Writing, and you can move freely among questions within a module but never back across modules.
How is PSAT Math scored?
PSAT Math is scored 160–760 in 10-point intervals, and it adds to the Reading and Writing section score (also 160–760) for a 320–1520 total with no weighting. There is no penalty for wrong answers, so answer all 44 questions. Because the section is adaptive, the difficulty of what you answered correctly matters, not just the count.
How much does PSAT Math matter for National Merit?
Math counts once in the National Merit Selection Index, while Reading and Writing counts twice: the index is (RW section score × 2 + Math section score) ÷ 10, on a 48–228 range. Every 10-point Math gain adds 1 index point versus 2 for RW — but a weak Math score still caps you, since the Class of 2027 Commended cutoff is 208 and state Semifinalist cutoffs run roughly 208–223.
How much time should I spend per PSAT Math question?
Budget about 95 seconds per question — 35 minutes for 22 questions per module. In practice, bank time on the algebra items you can dispatch in under a minute and spend it on multi-step word problems and data questions. Checkpoint: at minute 18 you should be at or past question 11, or you are behind pace.
Where to go from here
You now understand the PSAT/NMSQT better than most juniors ever will — the adaptive structure, the 320–1520 scale, and the Selection Index that turns one October morning into a National Merit shot. The points come next.
| Do this next | Why |
|---|---|
| Take the official PSAT/NMSQT practice in Bluebook | Convert format knowledge into reflexes under the real timer. |
| Prioritize Reading & Writing accuracy | RW is double-weighted in the Selection Index — each RW point is worth two. |
| Re-read the National Merit chapter | Know your state’s Semifinalist band and what a realistic target looks like. |
| Drill traps in the AskSia app | Per-distractor coaching on why you miss — the part a static guide can’t give. |
