To find a linear function from a table, check that the output changes by a constant amount each time the input rises by 1, then use the formula y = mx + b where m is that constant change and b is the output when the input equals 0. If the input isn't consecutive integers, divide the change in output by the change in input. If that ratio isn't constant, the table doesn't represent a linear function, and no amount of rearranging will change that.

First Confirm the Table Is Actually Linear

A table represents a linear function when the difference in y-values divided by the difference in x-values gives the same number for every pair of points. That number is the slope, usually written m. Take a table where x runs 1, 2, 3, 4 and y runs 5, 8, 11, 14. Each step in x is +1 and each step in y is +3, so the ratio is 3 every time. That constant ratio is the proof.

The failure case is more common than students expect. Consider x = 1, 2, 3, 4 with y = 2, 4, 8, 16. The changes in y are 2, 4, and 8 β€” not constant, so the ratio shifts and the table is exponential, not linear. A table with x = 1, 2, 4, 8 and y = 3, 6, 9, 12 looks decoy-linear because y rises by 3 each row, but the x-value doubles each row. Divide 3 by 1, then by 2, then by 4, and you get 3, 1.5, 0.75 β€” no constant slope.

A shortcut some teachers skip: compare the second differences. For a true linear function, the first differences are constant and the second differences are all zero. That check costs ten seconds and catches the decoy tables that trip people on tests.

One more caution. A table with only two rows can always be fit by some line, so it tells you almost nothing β€” infinite lines pass through two points. You need at least three rows at different x-values before you can claim the table is linear with confidence. The College Board's SAT math reference materials treat linear functions as having a constant rate of change, which is exactly the test you're running here.

Compute the Slope From Any Two Rows You Trust

Pick two rows, subtract the y-values, subtract the x-values in the same order, and divide. The order matters only in that both subtractions must go the same direction. Using (2, 8) and (5, 17): y goes 8 to 17, a change of 9. x goes 2 to 5, a change of 3. Slope is 9 Γ· 3 = 3.

Always subtract in the same direction. If you compute yβ‚‚ βˆ’ y₁, you must compute xβ‚‚ βˆ’ x₁ with the same labeling. Reversing one but not the other flips the sign and produces a slope of βˆ’3, which will fail when you plug it back in.

Pair of points Change in y Change in x Slope
(2, 8) and (5, 17) +9 +3 3
(5, 17) and (9, 29) +12 +4 3
(2, 8) and (9, 29) +21 +7 3

The third row is the cross-check. If the slope from the first two rows matches the slope from the outer two rows, you've verified linearity without checking every adjacent pair.

When the table has non-integer or negative inputs β€” say x = βˆ’3, βˆ’1, 1 with y = 11, 5, βˆ’1 β€” the arithmetic still works. From βˆ’3 to βˆ’1, y drops 6 and x rises 2, giving βˆ’3. From βˆ’1 to 1, y drops 6 and x rises 2, again βˆ’3. A negative slope just means the line falls as you read left to right.

Find the Y-Intercept When x = 0 Appears in the Table

If the table includes a row where x = 0, the y-value in that row is your b. Done. A table reading x = 0, 1, 2, 3 with y = 4, 7, 10, 13 gives slope 3 and intercept 4, so the function is y = 3x + 4. That's the fastest case you'll see.

Verify by substitution rather than trusting your eyes. Plug x = 2 into y = 3x + 4 and you get 10; the table says 10. Plug x = 3 and you get 13; the table says 13. If either fails, either the slope or the intercept is wrong, and the substitution test tells you within a few seconds.

Tables in textbooks and on Khan Academy's algebra practice sets often omit x = 0 deliberately, precisely so that you have to extend the pattern. If the table shows x = 2, 3, 4 with y = 7, 10, 13, the slope is 3 and the y-intercept sits two steps back: 7 βˆ’ 3 βˆ’ 3 = 1. So b = 1 and the function is y = 3x + 1.

Negative x-values appear more often than students expect on standardized tests. If the table starts at x = βˆ’2 and the slope is 3, subtract 3 twice more to work backward to x = 0. That's 6 less, so b shifts down accordingly.

Back-Calculate b When the Table Skips x = 0

Extending the pattern is the method to lean on whenever x = 0 is absent. Say the table reads x = 4, 5, 6 with y = 19, 23, 27. Slope is 4 per row. To reach x = 0 from x = 4, subtract 4 four times, which means subtracting 16 from 19. That gives b = 3, so y = 4x + 3.

The formula form of this is b = y βˆ’ mx, using any single row. Take (6, 27) and the slope 4: b = 27 βˆ’ 4(6) = 27 βˆ’ 24 = 3. Same answer. This version is faster when the table starts at large x-values, like x = 20, because extending backward 20 times invites arithmetic slips.

Do the arithmetic twice, using two different rows. With (4, 19): b = 19 βˆ’ 4(4) = 3. With (5, 23): b = 23 βˆ’ 4(5) = 3. Two matching answers is your confirmation.

Fractional slopes make this messier, and that's where the failure case lives. Suppose the table reads x = 3, 6, 9 with y = 5, 9, 13. Slope is 4 Γ· 3, or 4/3. Using (3, 5): b = 5 βˆ’ (4/3)(3) = 5 βˆ’ 4 = 1. The function is y = (4/3)x + 1. But if you round 4/3 to 1.33 and multiply by 3, you get 3.99 and b = 1.01 β€” close enough to pass a casual glance, wrong enough to fail an exact-answer check. Keep fractions as fractions until the final step.

Handle Tables Where x Steps Aren't 1

When consecutive rows show x jumping by 3, 5, or 0.5, the slope formula does all the work and the "add one per row" shortcut becomes false. Take x = 2, 5, 8 with y = 10, 16, 22. From row to row, y rises 6 while x rises 3, so the slope is 2. But some students see "+6 per row" and write y = 6x + b, which is badly wrong.

Extend backward by the slope, not by the row change. With slope 2 and x starting at 2, step back once to x = 0: y drops by 2 Γ— 2 = 4, giving b = 6. Verify with (5, 16): 16 βˆ’ 2(5) = 6. Verify with (8, 22): 22 βˆ’ 2(8) = 6. The function is y = 2x + 6.

Non-integer x-steps behave the same way. A table with x = 0.5, 1.5, 2.5 and y = 4, 7, 10 has slope 3 and b = 4 βˆ’ 3(0.5) = 2.5, giving y = 3x + 2.5. Half-steps are typical in conversion problems β€” the kind that map cups to fluid ounces, or kilometers to miles.

One safeguard: compute the slope using the two rows farthest apart. Rows far apart amplify any rounding in the table's values, which exposes a table that only looks linear because its rows are close together.

Extend a Table to Predict Values It Doesn't Show

Once you have y = mx + b, any x gives you a y. With y = 3x + 4, x = 10 predicts 34, and x = 100 predicts 304. That extrapolation is what makes this skill worth learning β€” the table is a sample, and the function is the rule behind the sample.

Interpolation is the safer direction. If the table shows x = 2 giving y = 10 and x = 4 giving y = 16, the function behind it is y = 3x + 4, and x = 3 predicts 13. That value falls between two known rows, so a wrong slope usually shows up as an answer that clashes with its neighbors. Extrapolation has no such guardrail, which is why textbooks treat predictions well outside the table with caution.

Worked prediction with a real-world flavor: a table tracking a phone plan's monthly cost shows 2 GB used costing $42 and 6 GB used costing $58. Slope is 16 Γ· 4 = $4 per GB. Then b = 42 βˆ’ 4(2) = $34. That's likely a $34 base fee, with $4 per GB added. The 10 GB prediction is $34 + $40 = $74, and it's honest arithmetic from the table, not a guess.

Check the answer against the table's last row before predicting anything far away. If your derived function reproduces every row exactly, predictions close to the data are trustworthy. If it misses even one row, stop and recheck the slope.

Write the Equation and Verify Every Row

Writing the answer means naming both constants clearly: slope m and y-intercept b, arranged as y = mx + b. Students lose points for writing "y = 3 + 4x" when the answer format asks for y = mx + b, even though the mathematics is identical. Match the format the question requests.

Table row y = 3x + 4 computed Table value Match
x = 1 7 7 yes
x = 2 10 10 yes
x = 3 13 13 yes
x = 4 16 16 yes

Run all four rows. A single mismatch means slope or intercept is off, and the mismatch pattern points to which. All rows off by the same amount means b is wrong. Rows diverging as x grows means m is wrong.

Two failure patterns worth memorizing. If y = 3x + 5 passes three of four rows but the last row is off by 2, the table wasn't perfectly linear to begin with, and the honest answer is "approximately linear" with a note. If your function matches every row except the first, you likely made a subtraction error in the first row's arithmetic, not a modeling error.

Also name the domain the table supports. A linear function fitted to a table of hours 1 through 6 has no business predicting year 6 of something. Stating where the table ends keeps the answer intellectually honest.

Avoid the Errors That Break Most Student Answers

The four errors below account for nearly every wrong answer on linear-function tables in classroom and test settings. Spot them before they cost you.

Error What it looks like The fix
Treating row change as slope y rises 6 per row, answer says m = 6, but x rises 3 per row Divide Ξ”y by Ξ”x, not by rows
Reversing one subtraction Calculating yβ‚‚ βˆ’ y₁ over x₁ βˆ’ xβ‚‚ Both subtractions must run the same direction
Rounding fractional slopes early 4/3 becomes 1.33, b drifts by 0.01+ Keep fractions until final answer
Skipping the verification step Function works for two rows, fails the third Substitute every row

The rounding issue sounds trivial and isn't. In a table with slope 4/3 and b = 1, rounding to 1.33 produces a predicted value at x = 30 of 40.9 instead of 41 β€” a visible mismatch in a word problem about dollars or miles.

Error two, the reversed subtraction, produces a sign flip that shows up immediately if you substitute. A function that matches at x = 1 and fails at x = 5 with the opposite sign drilled into it is almost always a sign error in slope calculation.

Silent mistakes happen because slope and intercept are two unknowns and one wrong answer can produce a convincing half-right equation. Verifying every row is not optional courtesy β€” it's the only way to catch which unknown is wrong. In that sense a linear-function table is an underdetermined puzzle dressed up as arithmetic.

FAQ

What if the table doesn't start at x = 0 and has no obvious pattern?

Extend backward using the slope. Compute slope from two rows, then subtract m Γ— (step to zero) from any y-value. If the slope is 3 and the table starts at x = 5 with y = 22, then b = 22 βˆ’ 3(5) = 7. No x = 0 row is required β€” the formula b = y βˆ’ mx works for any row.

Can a table represent a linear function if the y-values aren't evenly spaced?

Yes, as long as the ratio of change in y to change in x is constant. A table with x = 1, 3, 7 and y = 5, 11, 23 has y rising by 6 then 12 while x rises by 2 then 4 β€” both ratios equal 3, so the function is y = 3x + 2. Even spacing of y-values is a red herring; the constant ratio is what matters.

Is there any quick way to check for linearity without computing slope repeatedly?

Compute second differences. Take the successive changes in y and find the changes between those changes. If every second difference is 0, the table is linear. For x = 1, 2, 3, 4 with y = 7, 10, 13, 16, the first differences are 3, 3, 3, and the second differences are 0, 0. One arithmetic pass catches decoy tables fast.

What if a table gives four points and only three of them fit a line?

Then the table does not represent a linear function as written. Either the outlier is a typo in the source table, or the relationship is piecewise or nonlinear. Report the failure honestly instead of forcing a fit. In a classroom setting, raise the discrepancy β€” a teacher may confirm the outlier is a misprint.

Can two different tables produce the same linear function?

Yes, and often do. The table x = 1, 2, 3 with y = 7, 10, 13 and the table x = 10, 20, 30 with y = 34, 64, 94 both describe y = 3x + 4. That's the whole point of a function: the rule survives being sampled at different inputs. If two tables give different functions, at least one table is nonlinear.

What does a negative slope look like in a table?

As x increases, y decreases at a constant rate. A table with x = 1, 2, 3 and y = 9, 5, 1 has slope βˆ’4 and b = 13, yielding y = βˆ’4x + 13. Watch the order of subtraction: (5 βˆ’ 9) Γ· (2 βˆ’ 1) = βˆ’4 is correct, but (9 βˆ’ 5) Γ· (1 βˆ’ 2) also equals βˆ’4 because both differences flipped direction consistently.

How is finding a linear function from a table different from finding it from two points?

It's essentially the same procedure, with one advantage: a table gives you extra rows to verify against. Two points always define a line, so they can't tell you whether a relationship is truly linear. A table with three or more rows lets you test for constant slope β€” the actual proof of linearity β€” before committing to the equation.