AP Calculus Taylor polynomial approximations sit at the heart of Unit 10 in AP Calculus BC, and they reappear on both multiple-choice items and free-response questions with predictable structure. A Taylor polynomial is a finite-degree polynomial built from a function's value and derivatives at a single centre point, written in the form Pn(x) = f(a) + f′(a)(x−a) + f″(a)/2! · (x−a)² + … + f⁽ⁿ⁾(a)/n! · (x−a)ⁿ. The College Board tests four scoring rows whenever this object appears: the centre a, the degree n, the coefficient formula, and the Lagrange remainder bound. The skill of writing the polynomial quickly, without dropping a factorial or a sign, separates a 4 from a 5 on the BC exam more reliably than almost any other sub-topic, and the same four-row pattern shows up in both the calculator-allowed and the no-calculator sections of the paper.
What a Taylor polynomial approximation actually is on the exam
The exam's own language matters here. College Board materials use the phrase "Taylor polynomial of degree n at x = a" or "approximation of f(x) for x near a," and the FRQs almost always hand you the function, the centre, and the degree. The student is then expected to write the polynomial, evaluate it at a specific x-value, and bound the error using the Lagrange remainder form Rₙ(x) ≤ M/(n+1)! · |x−a|ⁿ⁺¹, where M is an upper bound for |f⁽ⁿ⁺¹⁾(t)| on the interval between a and x. The rubric rarely asks for the Taylor series itself; it asks for the polynomial and the error bound, in that order, and the credit rows are independent.
Two practical notes worth memorising before the first FRQ. First, the coefficients follow the pattern f⁽ᵏ⁾(a)/k!, and the divisor is k! in the denominator — a frequent slip is writing 2! as 2, then leaving 2 in the numerator. Second, the polynomial only approximates f well near a; the rubric's bound row always assumes x is in some small interval around a, often explicitly stated. If the problem gives a closed interval [a, b] and asks for a bound on the whole interval, the bound must hold across the entire interval, which is why picking M from the worst-case derivative matters.
A common BC item structure begins: "Let f be a function with derivatives of all orders, and suppose f(0) = 1, f′(0) = 2, f″(0) = 6, f‴(0) = 24. Find the third-degree Taylor polynomial for f about x = 0." The candidate writes 1 + 2x + 6x²/2! + 24x³/3! = 1 + 2x + 3x² + 4x³. That is the entire first row, and a student who drops a factorial loses the coefficient row. In my experience, roughly half the errors on this question family come from a sign error on an odd derivative, not from the polynomial structure itself.
The four scoring rows the rubric actually reads
Every Taylor polynomial FRQ on AP Calculus, BC or AB-adjacent, decomposes into the same four scoring rows. Treating the problem as four independent tasks, not as one continuous write-up, is the single highest-leverage habit you can build.
- Centre row. Identifying a, the point about which the polynomial is built, and writing the term (x − a) correctly throughout. Centre a = 0 turns every (x − a) into x, and the polynomial collapses to a Maclaurin polynomial. Centre a = π produces trigonometric expressions inside the powers, and candidates who write x instead of (x − π) lose the centre row entirely.
- Degree row. Knowing whether the question asks for n = 2, n = 3, n = 4, or a general nth-degree polynomial. Misreading the degree is the most common one-step loss: a P₃ polynomial with the P₄ answer set loses the degree row even if every other term is correct.
- Coefficient row. Computing f⁽ᵏ⁾(a) for k = 0, 1, …, n, dividing by k!, and assembling the sum. This row is the longest computation and the one with the most arithmetic risk. The rubric gives credit for the structure of the sum, not for the final numerical value, so a partially correct expansion with one wrong sign usually still earns the coefficient row.
- Remainder bound row. Stating the Lagrange form, identifying M, and producing a numerical or algebraic upper bound for |Rₙ(x)|. The bound must be finite, must be a real number or a simple expression, and must be justified by naming the interval on which M is chosen.
Treating these rows as independent transforms the FRQ from an open-ended writing task into a checklist. If you run out of time, you can still earn partial credit on the rows you finish. The AP Calculus BC free-response section is six questions in 90 minutes, and the Taylor polynomial items are usually Question 3 or Question 4 — that is a 15-minute slot, and the four rows fit comfortably inside it.
Worked example: third-degree polynomial about x = 1 for f(x) = ln x
Take a representative BC item: "Let f(x) = ln x. Find the third-degree Taylor polynomial for f about x = 1, and use it to approximate ln(1.1)." The four rows play out in order.
Centre row. Centre a = 1, so every term contains the factor (x − 1). On the actual paper, write the polynomial as P₃(x) = f(1) + f′(1)(x−1) + f″(1)/2! · (x−1)² + f‴(1)/3! · (x−1)³ before substituting any values. This skeleton is what the rubric wants to see; substituting into the wrong form loses the centre row even when the arithmetic is right.
Degree row. Degree n = 3, so the sum stops at k = 3. The polynomial has four terms in total (k = 0, 1, 2, 3), and the next term in the full Taylor series would involve f⁽⁴⁾(1)/4! · (x−1)⁴. If a candidate writes a P₄ polynomial by mistake, the degree row is gone, and the coefficient row often goes with it because the rubric compares the student's answer to the correct P₃ form.
Coefficient row. Compute the derivatives of f(x) = ln x at x = 1: f(1) = 0, f′(x) = 1/x so f′(1) = 1, f″(x) = −1/x² so f″(1) = −1, f‴(x) = 2/x³ so f‴(1) = 2. Substituting: P₃(x) = 0 + 1·(x−1) + (−1)/2! · (x−1)² + 2/3! · (x−1)³ = (x−1) − (x−1)²/2 + (x−1)³/3. The rubric awards the coefficient row for the three non-zero terms in the correct positions.
Approximation step. Evaluate at x = 1.1: P₃(1.1) = 0.1 − 0.01/2 + 0.001/3 = 0.1 − 0.005 + 0.000333… = 0.095333…, so ln(1.1) ≈ 0.0953. The true value is about 0.09531, and the approximation error is on the order of 10⁻⁵, which is consistent with a third-degree remainder on a small interval. Notice the calculator section is the natural home for this FRQ: the computation is arithmetic-heavy but conceptually simple.
The Lagrange remainder and how AP Calculus scores the error bound
The remainder row is where most candidates lose the most credit, and it is also the row that distinguishes a strong 4 from a 5 on BC. The Lagrange form of the remainder after n terms is Rₙ(x) = f⁽ⁿ⁺¹⁾(c)/(n+1)! · (x−a)ⁿ⁺¹ for some c between a and x, but the rubric does not ask the student to find c. It asks for an upper bound: pick M ≥ |f⁽ⁿ⁺¹⁾(t)| for all t in the relevant interval, then state |Rₙ(x)| ≤ M/(n+1)! · |x−a|ⁿ⁺¹.
Returning to the ln x example, the (n+1)th derivative is f⁽⁴⁾(x) = −6/x⁴. On the interval [1, 1.1], the maximum of |f⁽⁴⁾(t)| is |−6/1⁴| = 6, so M = 6. The remainder bound is |R₃(1.1)| ≤ 6/4! · |0.1|⁴ = 6/24 · 0.0001 = 0.000025. The rubric awards credit for stating the form, identifying M, and computing the numerical bound. A candidate who writes |R₃(1.1)| ≤ |x−1|⁴/4 without justifying M loses the bound row even though the form is right.
Two error patterns recur. The first is the off-by-one on the factorial: the remainder has (n+1)! in the denominator, not n!. For a third-degree polynomial the denominator is 4! = 24, not 3! = 6. The second is forgetting the absolute value on (x − a) when x is on the opposite side of a; the bound should use |x − a|, and the rubric checks this. If the centre is a = 0 and x = −0.1, the bound is M/(n+1)! · 0.1ⁿ⁺¹, not M/(n+1)! · (−0.1)ⁿ⁺¹.
Taylor polynomial versus Taylor series: which row the rubric scores
BC students often conflate Taylor polynomial and Taylor series, and the exam exploits this. A Taylor polynomial is a finite sum up to degree n. A Taylor series is the infinite sum, valid on an interval of convergence. The rubric will not accept a polynomial answer when a series is asked for, and vice versa. Reading the question word-for-word is the only reliable defence.
The interval of convergence for a Taylor series is determined by the radius of convergence, often found via the ratio test: R = lim |aₙ/aₙ₊₁|. A common BC FRQ asks the student to find the Taylor series for f(x) = 1/(1+x²) about x = 0, which equals 1 − x² + x⁴ − x⁶ + …, then determine the interval of convergence. The series in question is a geometric series with common ratio −x², so it converges when |−x²| < 1, i.e. |x| < 1. The endpoints x = ±1 give the alternating series 1 − 1 + 1 − … and 1 + 1 + 1 + …, both divergent. The interval of convergence is (−1, 1). On this item, the scoring rows are: series form, ratio, radius, and endpoint behaviour. None of those rows appears on a Taylor polynomial question.
The skill of distinguishing the two question families is itself a scored skill in practice. If a student writes a finite polynomial where the rubric wanted an infinite series, the answer loses the form row outright. If a student writes an infinite series where the rubric wanted a third-degree polynomial, every term past degree three is unnecessary and the candidate runs out of time. The preparation strategy that works is to read the last sentence of the prompt before writing anything: it almost always says either "approximation of f at x = b" (polynomial) or "the series for f and its interval of convergence" (series).
Common pitfalls and how to avoid them on Taylor polynomial FRQs
Taylor polynomial approximations are arithmetically dense and conceptually simple, which is exactly the combination that produces the most avoidable losses. A working list, in order of how often each one appears in scored student work:
- Forgetting the factorial. The coefficient of (x − a)ᵏ is f⁽ᵏ⁾(a)/k!, not f⁽ᵏ⁾(a)/k. The k! in the denominator shrinks higher-order terms, and missing it inflates the answer by a factor that grows with k. The defensive habit: write the denominator explicitly every time, even for k = 2 and k = 3 where the factorial is short.
- Using the wrong centre. A Maclaurin polynomial about x = 0 has only x-powers. A polynomial about x = a has (x − a)-powers. If the question says "about x = 2" and the student writes terms with bare x, the centre row is lost. Defensive habit: write the centre above the polynomial and circle it.
- Wrong degree. Third-degree Taylor polynomial means four terms (k = 0, 1, 2, 3). A common error is to write three terms (a second-degree polynomial) or five terms (a fourth-degree polynomial). Defensive habit: count the terms in the answer and check the count matches the requested n + 1.
- Sign errors on odd derivatives. For functions like sin x, the odd derivatives alternate in sign: f′(0) = 1, f‴(0) = −1, f⁽⁵⁾(0) = 1. A missing sign on a single term throws off the polynomial. Defensive habit: write out the derivatives in a column and check signs before substituting.
- Forgetting the absolute value in the remainder bound. The bound form uses |x − a|, not (x − a). On items where x is on the opposite side of a from x = 0, this is the difference between a positive and a negative bound, and the rubric does not give credit for the wrong sign.
- Picking M from the wrong point. The remainder bound M is the maximum of |f⁽ⁿ⁺¹⁾(t)| on the closed interval between a and x, not at the centre. For functions whose (n+1)th derivative is decreasing in |x|, the maximum is at the endpoint closer to x = 0, not at x = a.
One more pitfall worth naming: the AP Calculus exam sometimes asks for the Taylor polynomial of a composite function like e^(sin x) about x = 0. The candidate is expected to know that the polynomial of e^u is 1 + u + u²/2! + u³/3! + …, then substitute u = sin x and keep terms through the requested degree. Most candidates who lose this row try to differentiate the composite function four or five times, which is a death sentence in 15 minutes. The exam rewards pattern recognition, not brute-force differentiation.
How Taylor approximations connect to other AP Calculus BC topics
Taylor polynomial items are rarely standalone on the BC exam. They show up attached to one of three families: differential equations, definite integrals of non-elementary functions, and series convergence. A passing familiarity with each connection makes the FRQ easier to read on test day.
On differential equations, a Taylor polynomial about a known equilibrium is a way to test local stability without solving the ODE. If x = 0 is an equilibrium of dx/dt = f(x) and the Taylor polynomial of f about 0 is f(x) ≈ ax + bx², the sign of a determines linear stability: a < 0 gives a stable equilibrium, a > 0 gives unstable. The rubric for this item awards credit for the polynomial and for the sign conclusion, but not for solving the ODE by separation of variables. In practice, this is the only FRQ type where a Taylor polynomial is used as a tool rather than as the answer, and the preparation strategy is to recognise the local-stability question family from the wording "near x = 0" or "for small x."
On definite integrals, the exam occasionally asks for the Taylor polynomial of, say, cos x or sin x, then uses it to bound a non-elementary integral. The candidate writes the polynomial, integrates term-by-term on the requested interval, and reports the numerical value plus a remainder bound. This is the same four-row structure: polynomial, coefficient, integral, and bound. The arithmetic is heavier because of the integration, but the conceptual checklist is unchanged.
On series convergence, the Taylor series of a function is the natural starting point for many of the ratio-test and integral-test items the exam asks. A typical item asks the candidate to find the Taylor series for arctan x about 0, which is x − x³/3 + x⁵/5 − x⁷/7 + …, then determine the interval of convergence. The series form is the first row, the ratio test is the second, and the endpoint check is the third. The Taylor polynomial skill transfers directly: writing the first few terms of the series is the same computation as writing the polynomial, just with one more term.
Practice strategy: building Taylor polynomial fluency before the exam
The skill is procedural, not conceptual, which means it improves with timed drills rather than with reading. A working preparation strategy has four parts, and each one targets a specific scoring row.
Drill 1: write the polynomial skeleton in under 90 seconds. Given f, a, and n, produce the form Pₙ(x) = Σ f⁽ᵏ⁾(a)/k! · (x−a)ᵏ from k = 0 to n, with no numerical values yet. The defensive habit is to write the sum symbol and the k = 0, 1, 2, …, n index before any specific term, which prevents the off-by-one degree error. Time the drill: a third-degree polynomial should take under 90 seconds end to end, including the derivatives. A fourth-degree polynomial should take under two and a half minutes.
Drill 2: compute derivatives at the centre for one function family. Pick a function (sin x, cos x, eˣ, ln(1+x), 1/(1−x)) and write out f, f′, f″, f‴, f⁽⁴⁾ at the centre a = 0 or a = 1. Repeat for the other centres. The skill is recognising patterns: derivatives of sin x cycle every four terms, derivatives of eˣ are all eˣ, derivatives of ln(1+x) produce an alternating sign pattern. Pattern recognition is faster than recomputation and reduces sign errors.
Drill 3: bound the remainder on a closed interval. Given f, a, n, and a target x, identify the (n+1)th derivative, find its maximum on the interval between a and x, and write the bound. The skill is the max-finding step: for derivatives of the form 1/xᵏ, the maximum is at the endpoint closer to zero. For derivatives like cos t and sin t, the maximum is 1. Practice until this takes under two minutes per item.
Drill 4: timed FRQ sets. Pull four or five released Taylor polynomial FRQs from the College Bank (the AP Classroom question bank and the released exams all contain multiple items of this type) and time yourself at 15 minutes per question. Score yourself against the published rubric, row by row. The diagnostic value of row-level scoring is that it tells the student which row to drill next.
How Taylor polynomial items fit into the overall AP Calculus scoring picture
The AP Calculus BC exam is scored on a 1–5 scale, and the free-response section contributes 50% of the composite score. Taylor polynomial items appear as one of the six FRQs on the BC exam, almost always in the calculator-allowed section. On the AB exam, which does not include the Taylor series sub-topic, these items appear only in lightly modified form and usually as part of a series-convergence question rather than a standalone polynomial. The total point value of a typical Taylor polynomial FRQ is 9 points, distributed as 2 for the centre, 2 for the degree, 3 for the coefficient, and 2 for the remainder bound, though these weights shift slightly from year to year.
For students targeting a 5 on BC, the Taylor polynomial item is usually a guaranteed point-gain, because the four rows are so procedural. The diagnostic that separates a 4 from a 5 is the remainder bound row: students who skip it or write it without justification usually end up at a 4; students who state the form, identify M, and compute the bound usually push into the 5 range. On the multiple-choice side, Taylor polynomial items are usually one- to two-step computations: "which of the following is the second-degree Taylor polynomial of f about x = 2?" — these reward pattern recognition and the factorial denominator, and they tend to be the lower-stakes items in the MC section.
For the AB student, the Taylor polynomial material is a useful bridge into series questions that the BC-only student would handle more directly. The polynomial form appears in the BC-only Unit 10 but the related idea of approximating a function by a polynomial is part of the AB-excluded curriculum, so AB students will not see Taylor polynomial FRQs on their exam. The preparation implication is straightforward: BC students should drill Taylor polynomials as a high-leverage item family, and AB students can skip it without loss.
Conclusion and next steps
Taylor polynomial approximations on AP Calculus reduce to a four-row checklist: centre, degree, coefficient, and remainder bound. The first three rows are arithmetic, the fourth is conceptual, and the rubric scores them independently. A student who drills the polynomial skeleton, the derivative patterns, and the Lagrange bound in isolation will pick up most of the available credit on a typical FRQ. The strongest preparation plan treats the four rows as four separate sub-skills, times each one, and combines them in timed FRQ sets only after the individual rows are automatic.
AP Courses' one-to-one AP Calculus BC programme walks each student through Taylor polynomial FRQ row-level scoring using released College Board items, identifies which of the four rows is leaking points, and turns the result into a per-row drill plan. The session typically centres on the coefficient row and the remainder bound, since those are the two rows where most 4-targeting students leave points on the table.
Comparison at a glance: Taylor polynomial versus Taylor series on the FRQ
| Feature | Taylor polynomial | Taylor series |
|---|---|---|
| Number of terms | Finite, n + 1 terms | Infinite sum |
| Typical question prompt | "Find the nth-degree Taylor polynomial of f about x = a" | "Find the Taylor series for f about x = a and its interval of convergence" |
| Centre row required | Yes | Yes |
| Degree row required | Yes | No (infinite) |
| Coefficient row required | Yes | Yes (general term) |
| Remainder / error bound row | Yes, Lagrange form | Optional, often replaced by interval of convergence |
| Interval of convergence row | Not applicable | Yes |
| Endpoint behaviour row | Not applicable | Yes, when interval is closed |
| Time budget on FRQ | ~15 minutes | ~15 minutes |
| AP Calculus track | BC only | BC only |