How to fit an exponential model to data with Desmos
When each y-value changes by about the same factor, fit an exponential model in Desmos. Enter the x- and y-values as lists, plot the points, then type `Y ~ ab^X`. Desmos estimates the starting value and growth factor. If the differences are equal instead, use a linear model.
Updated

Check whether the data grows by a factor
When data is not a straight line, it can be tempting to try several models until one looks close. That can hide the key question: do the values change by a steady amount, or by a steady factor? Choose the wrong pattern and your model may describe the data poorly, even if Desmos gives you numbers.
For an exponential model, each step in x multiplies y by about the same factor. For example, values that go from 2 to 3 to 4.5 to 6.75 grow by about 1.5 each step. The model for that pattern is y = 2(1.5)^x.
For a linear model, the difference between each pair of y-values stays the same. Look at the values in order before fitting anything. Real data may not match a pattern exactly, so look for a consistent trend rather than demanding perfect matches.
Enter the data as lists in Desmos
Put the x-values in one list and the matching y-values in another. Keep the order consistent: the first x-value must belong with the first y-value, the second with the second, and so on. A mismatched pair can make the plotted points and fitted model misleading.
- Type `X = [0, 1, 2, 3, 4]`.
- On the next line, type `Y = [2, 3, 4.5, 6.75, 10.13]`.
- Type `(X, Y)` to plot the paired values.
Check that the points appear in the order you expect. If one value was copied into the wrong place, correct the list before fitting the model. The graph helps you see the overall shape, but it does not replace checking that the data was entered correctly.
Fit an exponential model with the regression symbol
Type `Y ~ ab^X` on a new line. Use the tilde for the fit, and press the right arrow after the exponent when entering the expression. Desmos then estimates a and b from the lists.
In the example, the estimates are close to a = 2 and b = 1.5. The fitted model is therefore y = 2(1.5)^x. Here, a is the starting value, and b is the factor by which the model grows for each one-step increase in x.
Do not confuse the fitted model with a rule that every future value must follow exactly. The example's last value is 10.13, while the repeated-factor pattern gives a value close to that. Small differences can occur in the data, so describe the fit as a model of the pattern.
Compare the model with the plotted points
After Desmos returns a and b, ask whether the values make sense for the data. A starting value near 2 fits the example because its first y-value is 2 when x is 0. A factor near 1.5 fits because each following value is about one and a half times the one before it.
A fitted curve is useful only if it reflects the pattern you noticed. Check whether the curve follows the plotted points rather than relying only on the displayed estimates. If the points do not show a steady factor, reconsider the model instead of accepting the first result.
Use a linear model when the differences are equal
If the y-values change by the same amount each step, use a linear model instead. In Desmos, type `Y ~ mX + b` to fit a line. For example, values that rise by the same amount from one x-value to the next have a constant difference, not a constant factor.
This check prevents a common mistake: seeing increasing values and assuming they must be exponential. Both linear and exponential data can rise. The difference is how they rise: an equal amount suggests a line, while an equal factor suggests an exponential model.
If the fitted exponential factor b is between 0 and 1, the model describes shrinking values, not growth. That may still match the data, but it is not evidence of exponential growth. Check the direction of the pattern as well as its shape.
Practise reading the pattern before fitting
On SAT maths questions, a calculator is allowed on every question, and Bluebook has a built-in Desmos calculator with graphing and scientific modes. A formula reference sheet is also available throughout the section. Learn to recognise the factor pattern first, then use Desmos to fit the model and check your reasoning.
One good practice routine is to explain the pattern in words before entering the lists: say whether the data changes by a steady amount or a steady factor, and identify the values that support your choice. Then fit the model and check what a and b mean. This makes it less likely that you will treat a calculator result as an answer without understanding it.
Official practice questions can help you apply the method in exam-style problems. When you finish one, do not only check whether your final answer matches. Go back to the data and identify the clue that made the exponential model suitable. That reasoning is what helps you choose the right method on a new question.
Also asked
- How can I tell if data is exponential or linear?
- Compare the changes between neighbouring y-values. If the values change by about the same factor each step, an exponential model may fit. If they change by the same amount each step, the data is linear. Check the pattern in the data before choosing a model.
- What does the b mean in an exponential model?
- In the model y = ab^x, b is the factor applied when x increases by one step. If b is greater than 1, the model grows. If b is between 0 and 1, the values shrink. The value a is the model's starting value.
- How do I enter an exponential regression in Desmos?
- Enter your x-values as a list called X and your y-values as a list called Y. Type (X, Y) to plot the points, then enter Y ~ ab^X. Desmos estimates a and b from your data. Use the right arrow after the exponent when entering the expression.