You find a linear function from a table by testing whether the change in output is constant for every equal step in input, then writing y = mx + b using that constant change as the slope m and the output at x = 0 as the intercept b. The method only works when the input values are evenly spaced and every pair produces the same ratio of output change to input change β if either condition breaks, the table belongs to a different function family.
Step 1: Confirm the Table Is Linear Before Calculating Anything
Before touching formulas, verify that the input column moves by a fixed amount. Tables built for textbook problems almost always use steps of 1, but real data β sensor readings logged every 30 seconds, revenue reported for uneven fiscal quarters β often does not. If x jumps 0, 2, 3, 7, you cannot compare consecutive output changes directly; you must compute the change per single unit of x for each pair instead. Skipping this check is the most common reason students label a quadratic table as linear.
Next, compute the first difference in the y column for each consecutive row. Subtract each output from the one below it and write that number in a scratch column. If the input steps are all equal, these differences should all match. A table of (1, 5), (2, 8), (3, 11), (4, 14) gives differences of 3, 3, 3 β every row agrees, so the function is linear with slope 3.
Watch for the failure case where the y differences look nearly constant but are not. Accuracy matters here: a table of (1, 2), (2, 5), (3, 10) has differences of 3 and 5, which is not linear β that data fits y = xΒ² + 1. Rounding a difference of 3 and a difference of 3.1 to "about 3" will send you down the wrong path. If the original data carries decimals, keep every digit until the final answer.
A third check catches tables where y changes by a constant factor instead of a constant amount. If outputs go 2, 4, 8, 16, the differences (2, 4, 8) grow while the ratios (2, 2, 2) stay fixed β that is exponential growth, not a line. Linear functions have constant differences; exponential functions have constant ratios. Memorize that distinction and you will never misclassify the two.
Step 2: Calculate Slope From Any Two Rows Using the Change Formula
Slope is the change in y divided by the change in x between any two points on the table. Pick two rows with clean numbers β the first and last are usually easiest β and apply m = (yβ β yβ) Γ· (xβ β xβ). For (1, 5) and (4, 14), that is (14 β 5) Γ· (4 β 1) = 9 Γ· 3 = 3. Because the function is linear, every pair you choose must return the same result, which gives you a built-in accuracy check.
Two cases demand care. If xβ β xβ equals zero, the slope is undefined and the "table" describes a vertical line, which is not a function at all β it fails the vertical line test. If the slope comes out negative, keep the sign; a table of (0, 10), (2, 6), (4, 2) has m = β2, and dropping the negative sign is a frequent grading error.
Cross-check with a second pair before trusting your slope. Using rows (2, 8) and (3, 11) from the same table: (11 β 8) Γ· (3 β 2) = 3. Two independent pairs agreeing at 3 confirms the table is linear and the slope is correct. If the second pair gives a different number, the table is not linear despite what the first pair suggested β return to Step 1.
Here is the arithmetic laid out for three sample tables so you can compare the outcomes:
| Table | Points | Change in y per +1 x | Slope | Classification |
|---|---|---|---|---|
| A | (1, 5), (2, 8), (3, 11) | 3, 3 | 3 | Linear |
| B | (1, 2), (2, 5), (3, 10) | 3, 5 | Not constant | Nonlinear |
| C | (0, 10), (2, 6), (4, 2) | β2 per +1 | β2 | Linear, negative |
Step 3: Solve for the y-Intercept and Write the Equation
With slope in hand, substitute one row's x and y into y = mx + b and solve for b. Using m = 3 and the point (2, 8): 8 = 3(2) + b, so 8 = 6 + b, and b = 2. The function is y = 3x + 2. Substitute a different row to verify β (4, 14) gives 14 = 12 + 2, which checks out.
The shortcut: if the table includes x = 0, the y value on that row is the intercept directly, no algebra needed. Tables of (0, 10), (2, 6), (4, 2) show b = 10 immediately, so the equation is y = β2x + 10. Many textbook tables deliberately skip x = 0 to force the substitution step, so learn both routes.
Reverse-check by building the table from your equation. Plug x = 1, 2, 3, 4 into y = 3x + 2 and confirm you recover 5, 8, 11, 14. If any value disagrees with the original table, either you misread a cell or the table was not truly linear. This final rebuild takes about 30 seconds and catches nearly every sign error.
Handle fractional slopes the same way. A table of (2, 1), (4, 2), (6, 3) gives m = (2 β 1) Γ· (4 β 2) = 1/2. Substituting (4, 2): 2 = (1/2)(4) + b β 2 = 2 + b β b = 0, so y = x/2. Keep fractions exact rather than converting to 0.5 early; rounding too soon is how decimal answers drift off the answer key.
How to Handle Uneven Input Values in Real Tables
Real tables rarely step by 1. Suppose a table shows x = 0, 3, 6, 9 with y = 4, 10, 16, 22. Consecutive differences are 6 each, but that is per 3 units of x, so the slope is 6 Γ· 3 = 2. Substituting (0, 4) gives b = 4, and the function is y = 2x + 4. Read the step size in the x column before dividing.
Generalize the formula to m = Ξy Γ· Ξx using the actual gap, not the raw difference. For x = 0, 3, 6, 9 with outputs 4, 10, 16, 22, any pair returns 2: (22 β 10) Γ· (9 β 3) = 12 Γ· 6 = 2. State the step in your work so the grader can follow the reasoning. Writing "difference = 6" without dividing by the x step of 3 is the most common error on uneven tables.
If the x steps themselves vary β say 0, 1, 4, 9 β you cannot use consecutive differences at all. Compute m for each pair individually and compare. When those per-pair slopes disagree, the function is nonlinear; when they agree, average nothing β the single agreement is your slope, and you can verify with b as usual.
This matters in applied settings. A water tank logged at 10-minute intervals with readings 30, 55, 80 gives 2.5 gallons per minute, because each 25-gallon jump happens across 10 minutes. Misreading the step as 1 minute gives 25 gallons per minute β a tenfold error that would overflow a system sized from the wrong number.
Checking Your Answer Against Every Row Instead of Just Two
A linear function must satisfy all rows simultaneously. After deriving y = 3x + 2, substitute every x in the table. Row (3, 11): 3(3) + 2 = 11 β. If even one row fails, your slope or intercept is wrong, or the table was never linear. Checking all four rows takes less time than redoing the whole problem.
Build a verification column when the table is long. For a ten-row table, tabulating the predicted value beside the actual value exposes the exact row where the pattern breaks. That row often reveals a transcription error in the original problem rather than a math mistake β a useful distinction when helping someone else debug their work.
There is a numeric tolerance question. Real measurements carry noise: a physics lab table of (0, 2.01), (1, 4.98), (2, 8.03) will not produce an identical slope from every pair. Round the slope to a sensible precision (about 3, here), state the assumption, and note that near-constant differences suggest a linear model rather than proving it. Pure algebra tables, by contrast, should match exactly, with no tolerance.
Fail loudly when rows conflict. Suppose rows (1, 5), (2, 8), (3, 12) appear β differences of 3 and 4. Writing y = 3x + 2 because it fits most points is misleading without a caveat. The honest answer is that no linear function passes through all three points, and the data is nonlinear or contains an error. State which.
Negative Slopes, Zero Slopes, and Special Table Shapes
A negative slope means y falls as x rises. Table (0, 6), (1, 4), (2, 2), (3, 0) has differences of β2 each, so m = β2 and b = 6, giving y = β2x + 6. The output reaches zero at x = 3; that point is the x-intercept and yields the root of the equation. Tables of declining values β temperature dropping overnight, a phone battery draining β almost always carry negative slopes, so the sign is meaningful, not decoration.
A zero slope produces a horizontal line. Table (1, 7), (2, 7), (3, 7) has differences of 0, so m = 0 and y = 7 regardless of x. Some students assume a "flat" table is not a function; it is a perfectly valid constant function. Identifying it correctly saves time on standardized tests that reward recognizing y = c.
Vertical patterns are not functions. Table (3, 1), (3, 4), (3, 9) has the same x for every row; the slope is undefined and you cannot write y = mx + b. This table describes a vertical line x = 3. State that conclusion rather than forcing it into slope-intercept form.
Mixed signs in y need careful subtraction. For (1, β2), (2, 1), (3, 4), differences are 1 β (β2) = 3 and 4 β 1 = 3, so m = 3. Substituting (1, β2): β2 = 3 + b, so b = β5, giving y = 3x β 5. The double negative inside the subtraction trips up many students; write the arithmetic out fully before simplifying.
Standardized-Test Shortcuts and Common Grading Traps
On exams like the SAT, PSAT, and ACT, most linear-function table questions reward the difference check over full algebra. The College Board's own scoring guidance for the SAT math section emphasizes function interpretation, and the fastest route is usually: compute one Ξy, divide by Ξx, then test answer choices by substituting x = 0. That cuts a 90-second problem to roughly 30 seconds.
Test the answer choices by substitution when the slope is awkward. If choices are y = 2x β 1, y = 2x + 3, y = 3x + 2, y = β2x + 5 and the table includes (0, 3), only y = 2x + 3 and y = β2x + 5 survive the intercept test. Then use one other row β (2, 7) eliminates β2x + 5 β and you have the answer without ever computing a slope. This strategy works whenever the table contains an x = 0 row, which many SAT tables do.
Traps that appear repeatedly on released practice tests include: answer choices that differ only by the sign of the slope; tables where one cell is a deliberate outlier to tempt an averaging approach; and answer choices written in slope-intercept when the question wants standard form Ax + By = C. Read the requested form before writing anything. A correct slope and intercept on the wrong form still loses the point.
When you can, show the substitution step explicitly in written work. Teachers and graders looking for method credit want to see b computed, not guessed, because the intercept is the step most students skip. Keeping a scratch column for differences and one line of substitution is usually enough evidence. Practice on a handful of tables from released SAT practice materials until the three-step flow is automatic.
Which Method to Use: Differences, Slope Formula, or Graphing
Three methods solve the same problem, and each suits a different situation rather than being universally better. Choose by table size, presence of x = 0, and whether you have graph paper nearby.
| Method | Best when | Weakness | Time for a 4-row table |
|---|---|---|---|
| Difference check | Inputs step by 1 | Useless on uneven x | Under 1 minute |
| Slope formula | Any spacing, any size | Requires substitution for b | 2β3 minutes |
| Graphing | Visual learners, checking work | Slow, imprecise on paper | 5β8 minutes |
For a table with x steps of 1 and an x = 0 row, scan the y column and write the answer in your head. For uneven or fractional inputs, the slope formula is the only reliable path. Use graphing as a verification step, not the primary method β plots catch an outlier row quickly but rarely give an exact fractional intercept by eye.
Reasoning about a table of values also builds the skill tested in function-notation questions, where you may be asked for f(5) given a table. Once y = 3x + 2 is established, f(5) = 17 without extending the table to a fifth row. That extension skill β generalizing from the table to the rule β is exactly what most curricula are measuring with these problems.
Where This Shows Up Beyond the Math Worksheet
The same three steps apply whenever a table of paired numbers appears in applied work. Engineering calibration sheets, chemistry lab data, and business revenue logs all pair an input with an output, and a constant-difference table signals a linear relationship you can model directly. U.S. high school algebra standards, reflected in the Common Core's functions domain (8.F.B.4 for constructing linear functions), require exactly this process, which is why it appears on every state assessment from grade 8 onward.
There is a safety-adjacent edge case. In engineering and pharmacy contexts, assuming linearity between table points can be dangerous when the underlying relationship is actually nonlinear β drug clearance curves and stress-strain plots frequently curve, and the table may sample evenly spaced points that disguise it. Always check at least three differences, not two, before trusting a linear model in real applications.
The practical habit that carries over: every time you build y = mx + b from a table, plug the x values back in and confirm you reproduce the y column. That single check catches sign errors, misread cells, and nonlinear tables that slipped past the difference test. It is the quiet step most people skip and the one that saves marks.
The slope is also the interpretation of a real-world rate β miles per gallon, dollars per hour, degrees per minute β so identifying it correctly matters for the units, not just the graph. Carrying units in the slope calculation (14 miles β 5 miles Γ· 4 gallons... no, mpg is miles Γ· gallons) keeps the answer meaningful. Label m and b with their units once and the model becomes readable to anyone, not just the person who built it.
FAQ
What if the table is linear but has a decimal step in the x column?
Compute slope with the actual decimal step: a table of x = 0, 0.5, 1.0 with y = 1, 3, 5 gives m = 2 Γ· 0.5 = 4. Round only at the final step, and label the slope with units per one whole unit of x. Decimals in the input are the highest-error case because students subtract the outputs correctly but divide by 1 instead of 0.5.
Can every linear table be written as y = mx + b?
No β a table with a repeated x value, such as (3, 1), (3, 4), (3, 9), describes a vertical line whose slope is undefined, and it does not qualify as a function. Likewise, if the y differences are constant on a table of decreasing y, the slope is simply negative but the form still works; only repeated x values break the form.
How many rows do I need before I can be sure the table is linear?
Three rows give two differences and a real check; two rows technically determine a line but cannot reveal whether the pattern continues. With exactly two points, any line through them fits, so you should say "this table is consistent with a linear model" rather than claiming proof. A fourth row is the strongest confirmation in a typical worksheet.
Do nonlinear tables ever look linear at first?
Often β a quadratic table like (1, 2), (2, 5), (3, 10) shows growing differences of 3 and 5, but a table sampled so widely that the curve flattens could show differences of, say, 3.0, 3.0, 3.2 that mask curvature. Three differences rather than two is the cheapest corrective, and if any adjacent differences disagree beyond rounding, classify the function as nonlinear.
What is the fastest check on a timed test?
Substitute the x = 0 row into each answer choice to eliminate candidates by intercept, then test one nonzero row. On the SAT this usually resolves the question in under 30 seconds, per the substitution approach described in College Board practice materials. Skip full algebra unless the table contains no x = 0 row.
Is it ever better to graph a table than to do the algebra?
Graphing beats algebra when the table is short, the points are clean integers, and the slope is small, because the intercept is visible where the line meets the y-axis. It loses to algebra with fractional slopes, large x values, or answer choices that differ by 0.5 in the intercept, where an eyeballed plot cannot distinguish options.