Why These Errors Matter
Average value and the Mean Value Theorem connect a function’s behavior across an interval to its behavior at particular inputs. The ideas are related, but they do not say the same thing. A calculation can go wrong by omitting the interval length, by substituting endpoint values for an average, or by overlooking a theorem’s hypotheses.
There is also a frequent logical overreach: a theorem may guarantee that at least one suitable input exists without guaranteeing that there is exactly one. Careful use begins by identifying which quantity is being averaged and which theorem applies.
Average Value: Include the Interval Length
The integral gives the signed accumulated value across the interval. Dividing by \(b-a\), the interval’s length, converts that accumulation into an average value. In a physical application, if \(f\) is measured in units such as metres per second and \(x\) in seconds, the integral has units of metres; dividing by seconds gives metres per second, the same units as \(f\).
A useful interpretation is that a constant function with value \(f_{\mathrm{avg}}\) has the same integral over \([a,b]\) as \(f\). The interval length matters: an accumulated total and an average are not interchangeable.
Worked Example: Avoid Omitting the Interval Length
Find the average value of \(f(x)=x^2\) on \([0,2]\), and compare it with the endpoint values.
First calculate the integral:
The interval length is \(2-0=2\), so the average value is
The endpoint values are \(f(0)=0^2=0\) and \(f(2)=2^2=4\). Their arithmetic mean is \((0+4)/2=2\), which is not the function’s average value. Omitting the interval-length division would instead give \(8/3\), also incorrect. As a check, the constant function with value \(4/3\) has integral \((4/3)(2)=8/3\) on this interval, matching the original integral.
Do Not Replace an Average with Endpoint Values
The average value depends on the function throughout the interval, not just on its values at the two endpoints. In particular, averaging \(f(a)\) and \(f(b)\) is not the definition of average value. That endpoint calculation can happen to agree with the integral average for some functions, but it is not a generally valid method.
For a linear function, the average value over an interval does equal the arithmetic mean of its endpoint values. This special case can make the mistaken method seem convincing. For a nonlinear function, as the previous example shows, the results can differ.
Two Mean Value Theorems, Two Different Conclusions
Two results are especially easy to confuse. The Mean Value Theorem for Integrals concerns a function value equal to the average value. The (differential) Mean Value Theorem concerns a derivative equal to a secant slope.
| Result | Conclusion | Required hypotheses |
|---|---|---|
| Mean Value Theorem for Integrals | There is a \(c\in[a,b]\) such that \(f(c)=\frac{1}{b-a}\int_a^b f(x)\,dx\). | \(f\) is continuous on \([a,b]\), with \(a<b\). |
| Mean Value Theorem | There is a \(c\in(a,b)\) such that \(f'(c)=\frac{f(b)-f(a)}{b-a}\). | \(f\) is continuous on \([a,b]\) and differentiable on \((a,b)\), with \(a<b\). |
The interval locations in these conclusions matter. The integral theorem allows \(c\) at an endpoint; the differential theorem guarantees an input strictly between the endpoints. Neither conclusion says that \(c\) is unique.
Why Continuity Guarantees an Input Attaining the Average
Here is a proof of the Mean Value Theorem for Integrals. Assume that \(f\) is continuous on \([a,b]\) and \(a<b\). By the Extreme Value Theorem, \(f\) has a minimum \(m\) and maximum \(M\) on this interval. Thus \(m\leq f(x)\leq M\) for every \(x\in[a,b]\).
Since \(b-a>0\), dividing the inequalities by \(b-a\) preserves their direction. Therefore \(m\leq f_{\mathrm{avg}}\leq M\). The Intermediate Value Theorem now applies: because \(f\) is continuous, it takes every value between its minimum and maximum. In particular, there is some \(c\in[a,b]\) such that \(f(c)=f_{\mathrm{avg}}\).
This argument explains why continuity is important: it prevents the function from skipping a value between its minimum and maximum. Integrability alone does not provide that guarantee.
Continuity Is a Hypothesis, Not a Decoration
A function can be integrable and still fail to attain its average value. The next example shows why the continuity condition in the Mean Value Theorem for Integrals cannot simply be ignored.
Worked Example: An Integrable Function That Skips Its Average
On \([0,1]\), define \(f(x)=0\) for \(0\leq x<1/2\), and \(f(x)=1\) for \(1/2\leq x\leq1\). This bounded step function is integrable, though it is discontinuous at \(x=1/2\).
Its integral is the area of the part of the interval where \(f=1\):
The interval length is \(1-0=1\), so \(f_{\mathrm{avg}}=(1/1)(1/2)=1/2\). But the only values of \(f\) are 0 and 1. Thus there is no \(c\in[0,1]\) for which \(f(c)=1/2\). The average is correctly calculated, but the theorem guaranteeing an input that attains it does not apply because \(f\) is not continuous on the interval.
The Mean Value Theorem Gives Existence, Not Uniqueness
For the differential Mean Value Theorem, the required conditions are that \(f\) be continuous on the closed interval \([a,b]\) and differentiable on the open interval \((a,b)\), with \(a<b\). It then guarantees at least one \(c\in(a,b)\) for which the instantaneous rate \(f'(c)\) equals the secant slope across the full interval.
A brief derivation shows where the conclusion comes from. Define the line \(L\) joining \((a,f(a))\) to \((b,f(b))\), and set \(g(x)=f(x)-L(x)\). The function \(g\) is continuous on \([a,b]\), differentiable on \((a,b)\), and satisfies \(g(a)=g(b)=0\). Rolle’s Theorem states that a function with those properties has some \(c\in(a,b)\) where \(g'(c)=0\). Since \(L'(x)=(f(b)-f(a))/(b-a)\), the equation \(g'(c)=0\) gives the Mean Value Theorem’s conclusion. Rolle’s Theorem guarantees at least one such input, not exactly one.
Worked Example: Find and Check Both Mean Value Theorem Inputs
Apply the Mean Value Theorem to \(f(x)=x^3-3x\) on \([-2,2]\). This polynomial is continuous on \([-2,2]\) and differentiable on \((-2,2)\), so the theorem applies.
First calculate the endpoint values and secant slope:
Differentiate and set the derivative equal to this slope:
Thus \(c=2/\sqrt{3}\) or \(c=-2/\sqrt{3}\). Both lie in \((-2,2)\), since \(2/\sqrt{3}<2\). Checking each value,
Both inputs give the secant slope. This verifies the theorem and shows concretely why its existence guarantee should not be mistaken for uniqueness.
A Reliable Error-Checking Routine
Before accepting an average-value or Mean Value Theorem calculation, check what the result is supposed to represent and which assumptions support it.
An average value uses an integral divided by interval length; a Mean Value Theorem for derivatives uses a secant slope.
Confirm \(a<b\), use the interval length \(b-a\), and note whether the theorem’s input lies in \([a,b]\) or \((a,b)\).
For the integral theorem, require continuity on \([a,b]\). For the differential theorem, require continuity on \([a,b]\) and differentiability on \((a,b)\).
“There exists” means at least one. Claim uniqueness only if you establish it separately.
These checks catch errors that arithmetic alone will not. A correct integral can still be mistaken for an average, and a correctly applied theorem can still be overstated if its conclusion is treated as unique.
Check Your Understanding
For each question, identify the calculation or theorem condition that matters.
- For \(f(x)=x\) on \([1,3]\), what is the average value? Show where the interval length enters.
- Why is \((f(a)+f(b))/2\) not a general formula for average value?
- Which continuity condition is needed to guarantee that a function attains its average value?
- State both hypotheses of the differential Mean Value Theorem and where its input \(c\) lies.
- Does the Mean Value Theorem guarantee exactly one input \(c\)? Explain briefly.
