SAT Problem Solving and Data Analysis: Essential Formula Chart for Success

The digital SAT Problem-Solving and Data Analysis domain tests how well students use numerical information in realistic situations. It covers ratios, units, percentages, distributions, scatterplots, probability, sampling and statistical claims. Good performance depends less on memorising lengthy statistics formulas than on selecting the correct relationship and interpreting the result.

College Board assigns approximately 5–7 of the 44 SAT Math questions to this domain—about 15% of Math. Questions appear across the two 35-minute adaptive Math modules, and a calculator is available throughout. The official on-screen formula sheet contains mainly geometry references; it does not supply a dedicated list of interest, probability or standard-deviation formulas. The chart below is a study guide created for this article, not a reproduction of the Bluebook reference sheet.

Problem-Solving and Data Analysis formula and concept chart

Skill Useful relationship Critical condition
Unit rate Rate = quantity ÷ associated unit count Keep numerator and denominator units consistent
Percent change (New − original) ÷ original × 100% The original value must be nonzero
Increase/decrease factor New = original × (1 ± decimal rate) Use a plus for increase and minus for decrease
Mean Sum of values ÷ number of values Use all values, including repetitions
Median Middle ordered value, or average of the two middle values Order data before finding the centre
Range Maximum − minimum Measures only extreme-value spread
Probability Favourable equally likely outcomes ÷ all equally likely outcomes Define the possible outcomes correctly
Conditional probability P(A given B) = P(A and B) ÷ P(B) P(B) must be greater than zero
Slope (y₂ − y₁) ÷ (x₂ − x₁) Only when the x-values differ

These relationships are tools for interpreting a question, not automatic templates. The correct denominator, units, reference group and level of precision depend on the problem.

Ratios, rates and dimensional analysis

If a vehicle travels 150 miles in 3 hours at a constant average rate, the average speed is 50 miles per hour. The quotient’s units are miles divided by hours. If a question instead asks for travel time at 50 miles per hour over 200 miles, calculate 200 ÷ 50 = 4 hours.

A proportional relationship maintains a constant multiplier. If 3 notebooks cost $12 at a fixed unit price, one costs $4 and 7 cost $28. But not every word problem describes proportional quantities: a taxi ride with a fixed pickup fee has a nonzero intercept and requires a different linear model.

For conversions, multiply by a fraction equal to one. Converting 2.5 hours to minutes gives 2.5 h × (60 min/1 h) = 150 min. Write units as part of the calculation so that incorrect numerator/denominator choices become visible.

Percentages: know the base

When a price rises from $40 to $50, the increase is $10. Relative to the original $40, that is 25%. The result is not 10% merely because the numerical difference is ten. If a $100 item is discounted by 20%, its new price is 100 × 0.80 = $80.

Successive changes compound. A $100 price discounted by 20% and then increased by 20% becomes 100 × 0.80 × 1.20 = $96, not $100. For an 8% tax on a pretax price P, the final price is 1.08P; if the final price is $108, the original is 108/1.08 = $100.

For percentages shown in graphs, distinguish a change in percentage points from a relative percent change. A share rising from 20% to 30% increases by 10 percentage points, or 50% relative to its original share.

One-variable data: mean, median and spread

For the ordered sample 2, 4, 4, 10, the mean is (2+4+4+10)/4 = 5, the median is (4+4)/2 = 4, and the range is 10−2 = 8. The value 10 pulls the mean above the median; the difference is meaningful when comparing skewed data sets.

College Board can ask students to compare data distributions with the same or different standard deviations. Standard deviation measures variation around the mean. The original article printed a malformed standard-deviation and variance formula and treated specialised coefficients of variation as central SAT requirements. For current SAT preparation, first practise interpreting what greater versus smaller spread means and how data changes affect centre and variation.

A data set where every value is identical has standard deviation zero. Changing one extreme value can increase the spread even if most observations stay the same. The exact numerical standard-deviation calculation is not an ordinary replacement for understanding the displayed distribution.

Two-variable data: scatterplots and models

Always read axis labels and units. A scatterplot may suggest a positive or negative association; a line of best fit can be used for a reasonable prediction within the range and context supported by the data. Extrapolating far beyond observed values can be unreliable.

For a linear model, slope represents the estimated change in the vertical variable for a one-unit increase in the horizontal variable. A slope of 2.5 dollars per hour means an estimated $2.50 rise for each additional hour if the model fits the situation. A positive correlation does not by itself prove that one variable caused another.

The SAT may also compare linear and exponential growth. Linear models add approximately the same amount per equal interval; exponential models multiply by approximately the same factor. For an original quantity of 200 increasing by 5% per year, the value after two years is 200 × 1.05² = 220.5. Treat financial compound interest as an example of this relationship, not as evidence that College Board supplies a special compound-interest formula sheet.

Probability and conditional probability

Suppose a school activity has 40 participants. Of 25 attendees in a particular group, 10 pass an assessment. The probability of passing given membership in that group, when selecting a student uniformly from the group, is 10/25 = 40%. The fraction 10/40 = 25% instead describes those 10 students as a fraction of all participants. The reference population changes the answer.

Read “among,” “given that” and “of those who” as clues to a conditional denominator. If a two-way table is presented, identify the relevant row, column or intersection before calculating a proportion. Avoid assuming independence unless the passage states or supports it.

Samples, margin of error and evidence quality

A sample can estimate a larger population quantity, but its design and uncertainty matter. A stated margin of error qualifies the precision of a reported estimate; for example, 52% with a margin of error of ±3 percentage points describes an approximate 49%–55% interval under the poll’s stated method. This is not proof that the true value must fall there for every imaginable sample.

Distinguish an observational study from a properly controlled experiment. Observational data can show association; random assignment in an appropriately conducted experiment provides stronger evidence about cause and effect. A self-selected sample may not represent the claimed population even when its arithmetic is correct. Check the question’s scope before generalising a statistic.

Interpret charts without inventing conclusions

Chart or data display Check first Common error
Bar chart Categories, units, baseline and scale Misreading a truncated axis as a larger difference
Line graph Time intervals and vertical units Confusing a changing level with a changing rate
Scatterplot Direction, pattern, outliers and model Claiming causation from correlation
Two-way table Row or column used as denominator Computing a joint rather than conditional proportion
Distribution Centre, spread and outliers Using mean and median interchangeably

When a graphic includes a trendline, understand what its prediction means and whether the problem asks for an exact displayed point or an approximate modeled value. Estimate the result first to catch a calculator or axis-reading mistake.

How to practise with current digital SAT tools

  1. Take a scored official Bluebook SAT practice test with the two 35-minute Math modules.
  2. Review the Problem-Solving and Data Analysis mistakes in My Practice, including uncertain correct choices.
  3. Classify each error as wrong base, wrong units, data interpretation, probability denominator, statistical inference or time pressure.
  4. Use the College Board Student Question Bank for unfamiliar questions addressing that specific skill.
  5. Check reasoning and units before relying on calculator output; then retest under the current module timing.

The most useful SAT formula chart is a short set of relationships you can apply correctly. The PSDA domain is fundamentally about reasoning from numerical information. Learn when a formula fits, which population or original value belongs in the denominator, and whether the conclusion is justified by the data—not merely how to copy an equation.

  • img