To calculate rate of change from a table, subtract the output values and divide by the difference in input values for the same pair of rows: rate of change = (y₂ − y₁) ÷ (x₂ − x₁). Pick two rows that are easy to read, subtract in the same order on top and bottom, and keep the units attached to the answer. A table of paired values only supports a rate of change if it describes a linear relationship, so the same division run on a nonlinear table gives you an average over an interval rather than a single constant slope.

This method works identically for tables printed in a textbook, copied from a spreadsheet, or generated by a graphing calculator. The arithmetic takes under a minute; the judgment calls about which rows to use and what the units mean take longer. Below are the steps in the order you'd actually perform them, a worked example, the failure cases that produce wrong answers, and the unit conversion that trips up students working with time-based tables.

Identify the Input and Output Columns Before Subtracting Anything

The first risk in reading a table is swapping the columns, so label them before touching a calculator. In almost every textbook table, the left column holds the input (usually called x, or the independent variable) and the right column holds the output (usually named f(x) or y). Confirm that pattern by checking the table's header note, because some physics and business tables reverse the arrangement and put time down the page while measurements run across it.

Write the relationship as a verbal ratio first: "output change per input change." That phrasing makes the arithmetic self-checking, since a rate of change of distance over time does not give speed. If the table's column headings carry units — seconds, dollars, centimeters, degrees Celsius — copy them onto your scratch paper before you start so the final answer inherits the right unit.

Watch for tables where one column is not actually a variable. A t-interval column, a row-number column, or a cumulative total column changes what the subtraction means. A cumulative column, in particular, yields a rate of change for the running total rather than the underlying quantity.

If the table has three or more columns, you can compute several rates of change from the same rows — one per output column. Do them one at a time and label which pair of columns each rate belongs to.

Apply the Rate of Change Formula to Any Two Rows

Take two rows from the table and assign them as "first" and "second" in a consistent direction, usually left to right. The formula is (y₂ − y₁) ÷ (x₂ − x₁), where y is the output column and x is the input column. Subtract the first row's output from the second row's output, then subtract the first row's input from the second row's input, then divide.

The order matters. If you subtract top-minus-bottom on the output side, you must also subtract top-minus-bottom on the input side, or the sign of your answer flips. A rate of change of −3 and a rate of change of +3 from the same table usually means somebody reversed one subtraction and not the other.

This subtraction is identical to the slope formula you'll see in algebra class, often written as m = (y₂ − y₁) ÷ (x₂ − x₁). That's not a coincidence: the rate of change between two points is the slope of the line joining them. The phrase "rate of change" is more general and fits contexts where calling something a slope sounds odd, like a table of temperature readings or a spreadsheet of subscription revenue.

For tables with more than two rows, you don't have to use adjacent rows. Any two rows work, and choosing rows with clean numbers (whole inputs, round outputs) reduces arithmetic errors. The worked example below uses rows that are two apart specifically to show that.

Work Through a Two-Point Example With Whole Numbers

Here is a table of a small business's subscription revenue by month. The input is months, and the output is revenue in dollars.

Month (x) Revenue in dollars (y)
1 1,200
3 1,800
5 2,400
7 3,000

Using month 1 and month 7: the output change is 3,000 − 1,200 = 1,800 dollars, and the input change is 7 − 1 = 6 months. Divide 1,800 by 6 and the rate of change is 300 dollars per month. Using month 3 and month 5 instead gives 2,400 − 1,800 = 600 dollars over 5 − 3 = 2 months, which is also 300 dollars per month. Identical answer regardless of which pair you choose, which is the test that the table is linear.

Now the same arithmetic with a different table: a car's distance from a starting point measured every two hours.

Hours (x) Distance in miles (y)
0 40
2 140
4 240
6 340

From hour 2 to hour 6: (340 − 140) ÷ (6 − 2) = 200 ÷ 4 = 50 miles per hour. That's the speed, and the units confirm it, because miles divided by hours is miles per hour.

Run the check on both tables: compute the rate between every consecutive pair of rows. If all pairs return the same value, the relationship is linear and you can state the rate as a single number. If they don't agree, read the next section before interpreting your result.

Confirm the Table Is Linear, or Treat the Result as an Average

A table of values doesn't come with a linearity guarantee. Run the rate-of-change calculation on each consecutive pair of rows. If every pair gives the same number, the data is linear and that number is the constant rate of change. If the numbers differ, the table describes a curve, and any single calculation from two rows is an average rate of change over that interval, not a single characteristic rate.

Here is a nonlinear table, the height of a dropped ball measured over four seconds.

Seconds (x) Height in meters (y)
0 44.1
1 39.2
2 24.5
3 0

From second 0 to second 1: (39.2 − 44.1) ÷ 1 = −4.9 meters per second. From second 2 to second 3: (0 − 24.5) ÷ 1 = −24.5 meters per second. The falling speed increases, so the average change per second more than quintuples across equal intervals. Reporting "−4.9 meters per second" as the ball's rate of change would be accurate only for that first second.

For a nonlinear table, the honest phrasing mentions the interval every time: "the average rate of change between x = 2 and x = 3 is −24.5 meters per second." That phrasing tells the reader exactly which two rows were subtracted.

You can also use the pattern of differences to identify the type of relationship. If the outputs change by a constant second difference (the differences of the differences), the table is quadratic, which is what the ball-height table shows — each successive difference of −4.9, −14.7, −24.5 changes by −9.8. If the outputs multiply by a constant factor, the table is exponential and subtraction-based rate of change is the wrong tool entirely.

Pick Non-Adjacent Rows When the Table Has Wide or Awkward Intervals

Textbook tables sometimes skip inputs: a table might show minutes 0, 5, 10, and 20 rather than every minute. In those cases, use the rows that give you the most readable calculation — usually the two with the largest separation and the cleanest output values. The answer is the same for a linear table, and for a nonlinear table the wider interval captures a longer average, which is often what you want anyway.

If a table gives intervals as labels rather than single points — for instance, a frequency table where the input is a range like "10–20" and the output is a count — the rate of change calculation needs a fixed representative value per interval. Use the midpoint of each interval. Broadcasting each range's midpoint before subtracting keeps the result comparable across intervals of different widths.

Rounding is the other judgment call. Tables rounded to one or two decimal places will produce rates that differ slightly depending on which rows you pick, so chain your calculation to the rawest numbers the table gives and round only at the end. A rate that comes out as 2.997 across one pair and 3.004 across another is a rounding artefact, not evidence of a nonlinear relationship.

For tables with more than two rows in a linear sequence, computing the rate across adjacent pairs and averaging them gives the same result as a single calculation across the endpoints. Do the two-step version only when you want to check the linearity. Otherwise one calculation across the endpoints is faster and just as accurate.

Keep the Units Attached and Avoid the Most Common Mistakes

Most wrong answers from table calculations are not arithmetic errors. They are order and unit mistakes. Subtract in the same direction on both sides of the fraction. Every rate of change carries a unit of the form "output unit per one input unit." Write that unit down. If it comes out as "miles per mile" or "dollars per dollar", you've swapped the columns.

The single most common error is reversing the subtraction order. A rate that should be −12 comes out as +12 because the outputs were subtracted first-row-minus-second-row while the inputs were subtracted second-row-minus-first-row. Cross-multiply the signs mentally before you write the answer down.

The second common error is treating a nonlinear table as if it were linear. The check from the earlier section — computing the rate for every pair of rows — takes under a minute for a table of five or fewer rows, and it's the only reliable way to catch this.

Watch out for tables where the input column is not evenly spaced. If the labels are 0, 2, 6, 10 (gaps of 2, 4, 4), using the gap size from the leftmost column as a constant divisor is wrong. Subtract the actual input values. The formula handles the unevenness automatically.

Finally, confirm what the input column represents when the table was generated by an instrument or a spreadsheet. A spreadsheet column of dates might be stored as text, in which case subtracting dates produces no meaningful number until the cells are converted to actual dates. Check the cell format on the column before trusting any subtraction.

Run the Arithmetic in a Spreadsheet Without Breaking the Sign

For large tables, a spreadsheet formula such as =(B3−B2)/(A3−A2) placed in a new column next to row 2 and copied downward computes the rate for every adjacent pair at once. Put the output column in column B and the input column in column A, then the formula in any other column gives the rate of change between each pair of consecutive rows.

The order inside the formula matters more than it does on paper because there is no human eye watching the copy direction. Anchor the first-row references correctly, or fill the formula downward and confirm that row 3's formula subtracts row 2's values rather than a mix of columns. Auditing one computed cell by hand catches almost every copy error.

If the input column contains dates, convert them to a serial number first. Excel and Google Sheets store a date as a serial number, and subtracting two serial numbers returns the number of days between them. Divide by 30, 365, or whichever unit the rate should be expressed in. Sheets also has a handy function for the specific case of averaging multiple rate pairs, but a simple sum of all differences divided by the sum of all input differences gets you the same result without extra functions.

Reading the output column of a spreadsheet table in a formula bar shows the values, but dates and times may not be visible as their serial representation. Confirm the format in the cell, not the display. Similarly, if any outputs are percentages formatted to display as "12%", the stored value is 0.12, and subtracting stored values gives a rate in decimal form.

Convert the Result When the Table's Units Don't Match the Answer You Need

The rate of change you calculate inherits the units of the table's rows, so if a table lists hours but the question asks for dollars per minute, convert after the division. Convert either units before dividing or the final rate after dividing; the answer is the same as long as you don't mix them.

Here is a small conversion table that covers most table problems. Use it when the table's input unit doesn't match the unit the question wants.

Table input unit Input unit requested Multiply the rate by
Hour Minute 1 ÷ 60
Minute Second 1 ÷ 60
Day Hour 24
Kilometer Meter 1,000
Pound Ounce 16
Dollar Cent 100

For example, a table of water flowing into a tank might list liters per minute while you need gallons per hour. Calculate the rate in the table's own units first — a clean liters-per-minute figure — then apply both conversions in a single multiplication. Mixing conversion and division in the same arithmetic is the source of most unit mistakes.

For currency or physical-unit conversions, use a verified reference for the conversion factor rather than a memorised figure; the U.S. National Institute of Standards and Technology publishes conversion factors for SI units and their accepted non-SI equivalents. For currency, the rate moves daily, so the factor used needs to be recorded alongside the calculation.

For the reverse direction — you have the rate and need to extend the table — multiply the rate by the input change and add it to a known output. A table with a rate of change of 300 dollars per month means that in three months revenue grows by 900 dollars.

Use the Table Method to Compare Two Data Sets or Verify a Claim

A single rate of change is a line of evidence; rates from two tables compared side by side are a verdict. If two tables share an input column and differ only in output, computing the rate from each gives a direct comparison — for example, $300 per month in subscription revenue versus $250 per month in advertising revenue tells you which channel pays faster.

The comparison table method also works for verifying a stated rate. If a table or article claims a rate of 12 units per hour but the table's endpoint values give 10, the table is the fuller evidence. The claim may have been made from a narrower interval within the table.

Comparing tables with different input intervals requires care. A table with monthly readings and a table with quarterly readings both yield rates, but the quarterly rate represents three months of change. Compare after converting to the same interval — the earlier section's conversion factors handle it.

For tables where a single cycle matters, such as seasonality in retail sales, rate-of-change comparisons across quarters can mislead. Compute the rate from the same quarter of consecutive years instead, and the seasonal pattern cancels out.

FAQ

What's the difference between rate of change and slope?

Nothing, when the table describes a linear relationship — the two names describe the same division. The phrase rate of change is more general and fits situations where the table describes a physical, financial, or biological quantity rather than a graph, such as temperature rise per hour or revenue growth per month. Slope is the more specific geometric term for the same calculation. Some textbooks use rate of change to include nonlinear cases where the calculation is an average, which slope does not.

Can I calculate a rate of change from a nonlinear table?

Yes, but only as an average rate of change over the specific interval between the rows you chose. Different pairs of rows give different answers on a nonlinear table, and no single number describes the whole table. Report the interval explicitly, as in "the average rate of change between x = 2 and x = 3." If you need an instantaneous rate of change, you need a graph or a formula, not a table.

What does a negative rate of change mean?

A negative rate means the output is decreasing as the input increases, so the calculation should come out negative when you subtract in the standard order. In the ball-height table earlier in this article, all rates are negative because the falling ball's height decreases over time. A negative sign is not automatically an error; it tells you the direction of change. If the sign surprises you, check the order of subtraction first.

What rate of change should a table produce if all inputs are the same?

No rate of change exists if the input values are equal, because the division denominator is (x₂ − x₁) = 0. A table with two rows at x = 5 and x = 5 cannot define a rate. This is the same restriction as slope on a vertical line, where the line is vertical and the slope is undefined. Check that the input column contains distinct values before calculating.

What's the average rate of change across every row of a table?

The average across every row of an evenly spaced table is the total output change divided by the total input change, using the first and last rows. That calculation weights each interval equally only if the input intervals are equal. For a table with uneven input intervals, the endpoint calculation reflects the whole table better than averaging the individual pair results, because averaging over uneven intervals overweights the short gaps.

Can I use a rate of change table to predict future values?

You can predict future values only if the table is linear and the pattern persists beyond the last row. Multiply the rate by the input change since the last known row, then add it to the last known output. That projection assumes the trend continues, which a table cannot verify. For nonlinear tables, extrapolation from a rate of change calculation drifts quickly and should not be trusted.

What are the units on a rate of change from a table?

The units are the output units divided by the input units, so a table of miles against hours gives a rate in miles per hour. Copy the units from the column headings before calculating and attach them to your answer. If the units come out weird — for example, "ounces per foot" from a table of weights and lengths — an interpretation may still exist, such as a material's linear density. Don't discard an unusual unit until you've checked what it describes.