8 topics. Each one opens with its key idea, then the detail you need to use it.
1. Riemann sums, the trapezoidal rule and the limit definition
Key idea: Find each piece's width and sample point before you add anything, and judge the error from monotonicity (left and right) or concavity (trapezoid and midpoint).
To estimate $\int_a^b f(x)\,dx$, split $[a,b]$ into pieces and build a rectangle on each one. The width of each rectangle is the length of its piece. The only choice is where you measure the height. A left sum uses the left endpoint, a right sum uses the right endpoint, and a midpoint sum uses the center. With $n$ rectangles you use $n$ heights, and the trapezoidal rule is the one method that uses all $n+1$ boundary values.
Exam tables often have unequal widths, so compute each piece separately. Suppose a rate $r(t)$ is $10, 14, 16, 12$ at $t=0, 2, 5, 9$. The widths are $2, 3, 4$. The left sum is $2(10)+3(14)+4(16)=126$, and the right sum is $2(14)+3(16)+4(12)=124$. The trapezoidal rule averages the two heights on each piece: $2\cdot\frac{10+14}{2}+3\cdot\frac{14+16}{2}+4\cdot\frac{16+12}{2}=125$.
The direction of the error follows from the graph. If $f$ is increasing, left sums underestimate and right sums overestimate, and the reverse holds when $f$ is decreasing. If $f$ is concave up, the straight tops of the trapezoids sit above the curve, so the trapezoidal sum overestimates. If $f$ is concave down, it underestimates. The midpoint sum goes the other way: it underestimates when $f$ is concave up and overestimates when $f$ is concave down.
The exact integral is the limit as the pieces shrink. You may need to read this backward. In $\lim_{n\to\infty}\sum_{k=1}^{n}\frac{2}{n}\left(1+\frac{2k}{n}\right)^3$, the width is $\frac{2}{n}$ and the sample point $1+\frac{2k}{n}$ runs from 1 to 3. So the limit equals $\int_1^3 x^3\,dx = 20$.
2. The definite integral and its properties
Key idea: Swap limits and the sign flips, adjacent intervals add, and a constant term $c$ contributes $c$ times the width of the interval.
The definite integral is a number, the signed area between $f$ and the $x$-axis from $a$ to $b$. When the graph is made of lines and circles, you can compute it with geometry and no antiderivative. For example, $\int_{-2}^{2}\sqrt{4-x^2}\,dx$ is the area of a semicircle of radius 2, which is $2\pi$. Region areas below the axis enter with a minus sign.
The properties of definite integrals let you build new integrals from given ones:
- $\int_a^a f(x)\,dx = 0$, since there is no width.
- $\int_b^a f(x)\,dx = -\int_a^b f(x)\,dx$, so swapping the limits flips the sign.
- $\int_a^b f(x)\,dx + \int_b^c f(x)\,dx = \int_a^c f(x)\,dx$, so adjacent intervals add.
- $\int_a^b [k f(x) + g(x)]\,dx = k\int_a^b f(x)\,dx + \int_a^b g(x)\,dx$, so constants and sums split apart.
Here is how they combine. Suppose $\int_0^5 f(x)\,dx = 8$ and $\int_3^5 f(x)\,dx = 3$. Then $\int_0^3 f(x)\,dx = 5$ by additivity. Next, $\int_0^3 [2f(x)+1]\,dx = 2(5)+3 = 13$, because the integral of the constant 1 over a width of 3 is 3. Finally, $\int_3^0 [2f(x)+1]\,dx = -13$. One trap: there is no product rule. $\int f(x)g(x)\,dx$ is not $\int f(x)\,dx \cdot \int g(x)\,dx$.
3. The Fundamental Theorem (Part 1): accumulation functions
Key idea: For $g(x)=\int_a^x f(t)\,dt$, the graph of $f$ is the graph of $g'$, so heights of $f$ are slopes of $g$ and areas under $f$ are values of $g$.
An accumulation function has the form $g(x)=\int_a^x f(t)\,dt$. Its input is the upper limit, and its output is the signed area built up from $a$ to $x$. Part 1 of the Fundamental Theorem says $g'(x)=f(x)$. The reason is simple. Push $x$ forward a tiny bit and the area grows by a thin strip about $f(x)$ tall, so the rate of growth is $f(x)$. You never need an antiderivative for this. For example, $\frac{d}{dx}\int_2^x \ln(t^2+3)\,dt = \ln(x^2+3)$.
The lower limit $a$ does not appear in the answer. Changing $a$ only adds a constant to $g$, and constants differentiate to zero. Three variations show up often:
- If the upper limit is a function $u(x)$, use the chain rule: $\frac{d}{dx}\int_1^{x^2}\cos t\,dt = \cos(x^2)\cdot 2x$.
- If $x$ is in the lower limit, flip the limits and negate: $\frac{d}{dx}\int_x^5 f(t)\,dt = -f(x)$.
- If you are given only the graph of $f$, read $g$ from it. $g(a)=0$. $g$ increases where $f>0$ and has a relative maximum where $f$ changes from positive to negative. $g$ is concave up where $f$ is increasing.
That last case is a free-response favorite. You find values of $g$ by adding signed areas on the graph of $f$. You find slopes of $g$ by reading heights of $f$. You find concavity of $g$ by reading the slope of $f$.
4. The Fundamental Theorem (Part 2): evaluating the definite integral
Key idea: Evaluate the antiderivative at both limits and subtract, and check that the integrand is continuous on the whole interval first.
Part 2 of the Fundamental Theorem turns area into algebra: $\int_a^b f(x)\,dx = F(b)-F(a)$, where $F$ is any antiderivative of $f$. The $+C$ cancels in the subtraction, so you can drop it here. For example, $\int_1^3 (x^2+1)\,dx = \left[\frac{x^3}{3}+x\right]_1^3 = (9+3)-\left(\frac13+1\right) = \frac{32}{3}$.
The subtraction is where points disappear. Students evaluate at the top limit and stop, or they assume the bottom value is zero. At $x=0$, $e^x$ equals 1 and $\cos x$ equals 1, so neither term vanishes. Always write both values in brackets before you subtract.
The theorem also needs $f$ to be continuous on $[a,b]$. If $f$ blows up inside the interval, as $\frac{1}{x^2}$ does at $x=0$ on $[-1,1]$, then plugging in the endpoints gives a wrong answer of $-2$. That integral is improper, and when you work it out with limits it diverges, so there is no finite value to replace the $-2$.
On the calculator section, you often cannot find an antiderivative at all. You then use the theorem in its net-change form and let the calculator compute the integral numerically. If $f(2)=7$ and $f'(x)=\sin(x^2)$, then $f(4)=7+\int_2^4 \sin(x^2)\,dx$. No elementary antiderivative exists, but the calculator gives the number directly.
5. Accumulation & net change: displacement versus total distance
Key idea: Displacement is $\int v\,dt$, total distance is $\int |v|\,dt$, and they differ whenever the velocity changes sign.
Accumulation & net change problems give you a rate and ask for an amount. The template is always: amount at the end equals amount at the start plus the integral of the rate. If a tank holds 50 gallons at $t=0$ and water flows in at $r(t)$ gallons per minute, it holds $50+\int_0^4 r(t)\,dt$ gallons at $t=4$. When water also leaves at $d(t)$, integrate the net rate, $r(t)-d(t)$.
Motion problems add one twist. The integral $\int_a^b v(t)\,dt$ is displacement, the change in position, and backward motion cancels forward motion. Total distance is $\int_a^b |v(t)|\,dt$. To compute it, find where $v$ changes sign, split the interval there, and add the absolute value of each leg.
Take $v(t)=4-2t$ on $[0,5]$. It changes sign at $t=2$. The first leg is $\int_0^2 (4-2t)\,dt = 4$, so the particle moves 4 units forward. The second leg is $\int_2^5 (4-2t)\,dt = -9$, so it moves 9 units back. The displacement is $4-9=-5$, but the total distance is $4+9=13$.
Check the question's wording before you answer. "Position at time 5" means initial position plus displacement. "Total distance traveled" means the sum of the legs. "How far from its starting point" means the absolute value of the displacement. Each is a different number, and the wrong ones always appear among the choices.
6. Antiderivatives and initial value problems
Key idea: Integrate, add $+C$, then use the given point to solve for $C$, and differentiate your answer to check it.
Antiderivatives are derivative rules read backward, so you can check every answer by differentiating it. The core list:
- $\int x^n\,dx = \frac{x^{n+1}}{n+1}+C$ for $n\neq -1$, and $\int \frac{1}{x}\,dx = \ln|x|+C$.
- $\int e^x\,dx = e^x+C$, and $\int \sin x\,dx = -\cos x+C$, while $\int \cos x\,dx = \sin x+C$.
- $\int \sec^2 x\,dx = \tan x+C$.
- $\int \frac{1}{1+x^2}\,dx = \arctan x+C$, and $\int \frac{1}{\sqrt{1-x^2}}\,dx = \arcsin x+C$.
The $+C$ is there because any constant differentiates to zero, so an indefinite integral is a whole family of functions. Rewrite before you integrate. Roots become powers, and a fraction with one term on the bottom splits apart. For example, $\int \frac{x^2+1}{\sqrt{x}}\,dx = \int (x^{3/2}+x^{-1/2})\,dx = \frac25 x^{5/2}+2x^{1/2}+C$.
Initial value problems pick one member of that family. Suppose $f'(x)=6x^2-4$ and $f(1)=5$. Integrating gives $f(x)=2x^3-4x+C$. Substituting $x=1$ gives $2-4+C=5$, so $C=7$ and $f(x)=2x^3-4x+7$. You can also skip $C$ by using net change: $f(x)=5+\int_1^x (6t^2-4)\,dt$. This second form is the one to use when $f'$ has no simple antiderivative.
7. u-Substitution (indefinite and definite)
Key idea: Every piece must change together: the integrand, the differential and, for a definite integral, both limits.
Substitution reverses the chain rule. Look for a function inside another function whose derivative also appears, up to a constant. In $\int x\cos(x^2)\,dx$, the inside is $x^2$ and its derivative $2x$ is present apart from the 2. Let $u=x^2$, so $du=2x\,dx$ and $x\,dx=\frac12\,du$. The integral becomes $\frac12\int\cos u\,du=\frac12\sin(x^2)+C$. For u-substitution (indefinite), always finish by writing the answer back in terms of $x$.
Linear insides come up so often that you should do them on sight. You divide by the inner coefficient: $\int e^{3x}\,dx = \frac13 e^{3x}+C$ and $\int (2x+1)^4\,dx = \frac{(2x+1)^5}{10}+C$.
For u-substitution (definite), change the limits along with the variable. Take $\int_1^2 \frac{x}{x^2+1}\,dx$ and let $u=x^2+1$. When $x=1$, $u=2$, and when $x=2$, $u=5$. The integral becomes $\frac12\int_2^5 \frac{1}{u}\,du = \frac12\ln\frac52$. You never go back to $x$ because the new limits already belong to $u$.
Three errors account for most wrong choices. Some students keep the $x$-limits on a $u$-integral. Some miss the constant, either dropping the $\frac12$ or counting the factor twice after $du$ has already absorbed it. Others forget that $e^0=1$ when they evaluate.
8. Selecting an antidifferentiation technique
Key idea: Try a basic form first, then substitution, then the algebra (division or completing the square), then parts or partial fractions.
Selecting an antidifferentiation technique is a decision you make in order. First, check whether the integral is a basic form or becomes one after rewriting. Second, look for an inside function whose derivative is present, which signals substitution. If neither works, reach for the techniques below.
Integration: long division & completing the square handles rational functions. When the top has degree at least as large as the bottom, divide first. Since $x^2+1=(x+1)(x-1)+2$, you get $\int\frac{x^2+1}{x+1}\,dx=\int\left(x-1+\frac{2}{x+1}\right)dx=\frac{x^2}{2}-x+2\ln|x+1|+C$. When an irreducible quadratic sits on the bottom, complete the square to reach the arctangent form: $\int\frac{dx}{x^2+4x+5}=\int\frac{dx}{(x+2)^2+1}=\arctan(x+2)+C$.
The BC course adds three more tools.
- Integration by parts, $\int u\,dv = uv-\int v\,du$, fits a product such as $\int x e^x\,dx = xe^x-e^x+C$. Choose $u$ as the factor that gets simpler when you differentiate it.
- Partial fractions split a bottom made of distinct linear factors: $\frac{1}{x^2-1}=\frac{1/2}{x-1}-\frac{1/2}{x+1}$, which integrates to two logarithms.
- Improper integrals replace an infinite limit or a blow-up point with a limit. $\int_1^\infty \frac{1}{x^2}\,dx = \lim_{b\to\infty}\left(1-\frac1b\right)=1$ converges, but $\int_1^\infty \frac1x\,dx$ diverges because $\ln b$ grows without bound.