NYSTCE Math CST common mistakes
NYSTCE Math CST · Understand the traps and how to avoid them.
When a word problem gives quantities in different units or asks for a rate, the candidate plugs numbers into a formula without converting to consistent units first, then reports the answer in whichever unit the formula happened to produce.
Under time pressure the candidate pattern-matches the problem to a memorized formula (d = rt, density = mass/volume, unit price) and treats units as decoration rather than part of the quantity. Mixed units such as minutes and hours or centimeters and meters are easy to skim past because the numbers look compatible.
What to do instead
Before computing, write every given quantity with its unit on a scratch line and circle any unit that appears only once. Convert to a single system (usually SI or the unit the answer requests) before substituting. After computing, attach the unit to the result and check it against what the question asked for; if the requested unit differs, do one final conversion. Practice by labeling units in every step of five rate problems per study session until it becomes automatic.
The candidate answers the quantity the problem's numbers make easy to compute rather than the quantity actually requested, for example finding the discount amount when the question asks for the final price, or finding the total when it asks for the difference.
Multi-step word problems bury the final request in the last sentence, and the candidate's working memory is consumed by the arithmetic for the intermediate step. The first computed number feels like an answer, so the candidate stops and selects it from the choices, where it is usually present as a distractor.
What to do instead
Read the last sentence of the problem first and underline the exact quantity requested, including qualifiers like 'per person', 'after tax', or 'how much more'. Restate the goal in your own words before computing. When finished, reread the underlined request and ask whether the number on your paper answers it; if not, do the missing operation. Deliberately check the answer choices for the intermediate value and reject it.
When simplifying expressions with negative exponents or radicals, the candidate moves a factor across the fraction bar but mishandles the sign, writing something like 1/x^-2 = x^2 as 1/x^2, or distributing an exponent over a sum as (a+b)^2 = a^2 + b^2.
Exponent rules are memorized as symbol patterns without the underlying meaning, so the candidate applies the rule that 'looks right' for the shape of the expression. The invalid distribution over addition is a persistent overgeneralization from the distributive property of multiplication.
What to do instead
Rebuild each rule from a small numeric example before using it: verify that x^-2 means 1/x^2 by testing x = 3, and verify that (3+4)^2 is 49, not 25. Keep a one-page rule sheet where each rule has a numeric check beside it. On the exam, when an exponent or radical step feels uncertain, substitute a small number into the original and the transformed expression; if they disagree, the step is wrong.
The candidate confuses rational and irrational numbers or misjudges whether a simplified radical or a decimal expansion is rational, and answers questions about closure or classification by intuition about how the number is written rather than its value.
Textbook definitions are learned as vocabulary lists separate from examples, and decimals like 0.333... or 1.414 look 'messy' so they get labeled irrational. The candidate does not connect 'rational' to 'can be written as a ratio of integers'.
What to do instead
Drill classification with a decision procedure: can the number be written as p/q with integers p and q, q nonzero? Apply it to repeating decimals by converting them algebraically (let x = 0.333..., then 10x - x = 3, so x = 1/3). For radicals, check whether the radicand is a perfect square or has a square factor. Build a table of ten numbers of varied forms and classify each with written justification.
Given a quadratic or rational function, the candidate reads intercepts, asymptotes, or the vertex directly off the written form without transforming it, for example treating the vertex of y = (x-3)^2 + 5 as (3, 5) when the sign is misread as (-3, 5), or missing a vertical asymptote at a canceled factor.
The candidate has seen vertex form and factored form but has not internalized how each parameter shifts the graph, so sign conventions are guessed. Cancellation in rational expressions feels like 'simplifying away' the problem, hiding the removable discontinuity.
What to do instead
For every function, rewrite it into the form that exposes the feature asked about before answering: complete the square for vertex and axis of symmetry, factor for zeros, and put over a common denominator for asymptotes. Sketch a rough graph with three labeled points. Remember that (x - h) shifts right by h, so test with x = 0 to confirm the direction. For rational functions, note canceled factors as holes, not asymptotes.
When composing or inverting functions, the candidate computes f(g(x)) as f(x)·g(x), or finds an inverse by swapping the variables and then fails to solve for y, leaving an expression that is not a function of x.
Notation f(g(x)) visually resembles multiplication, and the candidate has practiced function arithmetic more than composition. Inversion is often reduced to 'swap x and y', which is only the first step.
What to do instead
Write composition explicitly as substitution: f(g(x)) means replace every x in f's rule with the entire expression g(x), using parentheses. Practice with linear and simple nonlinear pairs and verify numerically, for example check that f(g(2)) equals the value obtained by evaluating g(2) then f. For inverses, always finish with 'solve for y' and then verify by composing: f(f^-1(x)) should return x.
The candidate applies a similarity or congruence criterion incorrectly, for example claiming two triangles are similar from angle-side-side data, or scaling only one dimension when enlarging a figure and reporting the area or volume as if it scaled linearly.
Similarity criteria (AA, SAS, SSS) are memorized as three-letter labels without the ordering constraint, and the area and volume scaling rules are confused with the linear scale factor because all three are taught together.
What to do instead
State the criterion with its condition when you use it: SAS similarity requires the included angle, so angle-side-side does not qualify. For scaling, write the factor k, then apply area factor k^2 and volume factor k^3 explicitly, and sanity-check with a square or cube of side 1 doubled to side 2. Practice by computing the new area and volume for three enlargements and confirming the ratios.
On circle and polygon problems, the candidate uses the diameter where the radius belongs (or vice versa) in area and circumference formulas, or uses degree mode when the problem requires radians, producing answers off by a factor of 2, 4, or pi.
Formulas are recalled as strings of symbols without the geometric meaning of each variable, and calculators default to degree mode while radian measures appear in trigonometry items. The candidate does not pause to identify which length the problem actually supplied.
What to do instead
Before substituting, write 'r = ___' and 'd = ___' from the problem statement and convert d to r explicitly. Label the formula with words: area = pi times radius squared. For angle problems, decide the mode first and set the calculator, then check the reasonableness of the angle (a radian measure near 1.57 is about 90 degrees). Do a reasonableness check on the result: an area should be smaller than the square of the diameter and larger than the square of the radius.
The candidate computes a mean, median, or standard deviation for a data set and then answers a question about the sampling distribution or the effect of an outlier using the sample statistic as if it were the population parameter, or ignores that adding a constant shifts the mean but not the spread.
Descriptive and inferential statistics are studied in separate chapters, so the candidate does not track which quantity a question is about. Transformations of data are learned as rules to memorize rather than consequences of the definitions.
What to do instead
Classify each statistics question before computing: is it asking about the data (descriptive) or about a statistic's distribution across samples (inferential)? For transformation questions, reason from the definition: adding c to every value adds c to the mean and median, leaves the standard deviation and range unchanged, and multiplying by k multiplies the mean and standard deviation by |k|. Verify with a tiny data set such as {1, 2, 3} transformed by adding 10 and by doubling.
The candidate treats events as independent or mutually exclusive interchangeably, computing P(A and B) = P(A)P(B) for events that are not independent, or adding probabilities for events that can occur together without subtracting the overlap.
The addition and multiplication rules look symmetric and the words 'and' and 'or' are mapped to operations without checking the relationship between events. The candidate skips the conditional-probability setup because it requires an extra reasoning step.
What to do instead
Before computing, draw a two-way table or Venn diagram and ask two questions: can both events happen at once (mutually exclusive?), and does knowing one changes the other's probability (independent?). Use P(A or B) = P(A) + P(B) - P(A and B) always, and P(A and B) = P(A)P(B|A) for dependent events, which reduces to the product rule when independent. Check with a concrete contingency table where the counts make dependence obvious.