An exponential model multiplies by the same factor over equal input intervals. A linear model adds the same amount. Recognizing this distinction lets you choose a model, interpret its parameters, and predict how a quantity changes.
Read the starting amount and repeated factor
In a is the value at because The base is the multiplier for each one-unit increase in . For a positive starting amount, gives growth, while gives decay.
For the initial amount is 200 and each period multiplies it by 1.3, an increase of 30%. For each period retains 70%, a decrease of 30%. The base records the amount kept, not just the percentage gained or lost.
The input unit matters. If is measured in years, the base is an annual multiplier. If counts three-month periods, it is a three-month multiplier. Always attach the period to the rate.
Example: growth repeats on the new amount
A culture starts with 80 cells and doubles each hour. The model is After three hours, cells.
The sequence is 80, 160, 320, 640. Its increases are 80, 160, and 320, not a constant amount. Its successive ratios are all 2. This is why adding to the starting value would produce the wrong model.
At the formula gives 40. That is an algebraic extrapolation to one hour earlier. Whether it is meaningful depends on whether the growth assumptions applied before the stated start. A formula can allow more inputs than the real situation justifies.
Cells
Doubling gives equal ratios, not equal increases: 80, 160, 320, 640.
Doubling gives equal ratios, not equal increases: 80, 160, 320, 640.
Describe this figure
Doubling gives equal ratios, not equal increases: 80, 160, 320, 640.
Curves: Cells.
Marked points: 80: filled marker at (0, 80); 160: filled marker at (1, 160); 320: filled marker at (2, 320); 640: filled marker at (3, 640).
Example: build a model from two values
Suppose and with a positive exponential multiplier. Write The second value gives so and
The model is representing a 20% decrease per period. The 36% total decrease over two periods is not a 36% decrease each period. Repeated change compounds:
If the inputs are equally spaced in a table, compare ratios. Unequal input intervals require adjusting the ratio to the length of the interval before inferring a one-unit multiplier.
Understand the graph
For and is always positive. The graph crosses the vertical axis at a and has horizontal asymptote In decay, the values approach zero but never equal zero at a finite input. In growth, the graph approaches zero in the negative-input direction.
A vertical shift changes the asymptote. For the long-term level is 10, not zero, and the initial value is 60, not 50. The coefficient before the exponential is the initial distance from the shifted baseline.
Model limits are part of the answer
Exponential growth usually cannot continue indefinitely in a real population because resources are limited. A model may still predict well within the observed interval. Likewise, a negative time or fractional count may require interpretation. Distinguish the mathematical graph from the model's practical domain and avoid claiming certainty outside the supplied assumptions.
Try it yourself
A device is worth 900 dollars initially and retains 80% of its value each year. Write a model and find its value after two years. Is the loss the same number of dollars each year?
Answer: so dollars. The first loss is 180 dollars and the second is 144 dollars. The percentage loss is constant, but the dollar loss decreases because it acts on a smaller value.
What to remember
Look for a constant ratio, identify the time unit, and interpret the base as the multiplier retained each period. The starting amount, asymptote, and practical domain complete the model.