Integration Techniques: How to Choose a Method and Use It

A practical guide to substitution, integration by parts, partial fractions, and trigonometric methods.

The central skill in integration is recognizing a useful transformation, not simply memorizing more formulas. Look for a simple substitution first, then use the integrand’s structure to consider integration by parts, partial fractions, or trigonometric methods. These choices are flexible: several methods may work, and a complete solution can require more than one. [2]

In this article

1. Start with structure, not a favorite formula
2. Substitution: simplify a function and its accompanying derivative
3. Match the technique to the transformation you need
4. Integration by parts versus partial fractions
5. Separate trigonometric integration from trigonometric substitution
6. Know whether you need an exact value, an approximation, or convergence

1. Start with structure, not a favorite formula

Check simple substitution first. An advanced-looking expression may become straightforward after a change of variable. [2]
Compare plausible approaches. Finding one method that works does not mean it is the easiest route. [2]
Expect methods to connect. Substitution can produce a new integral that calls for integration by parts or partial fractions. [2]

A useful first question is: “What kind of expression am I integrating?” A ratio of polynomials suggests one set of possibilities; a polynomial multiplied by an exponential suggests another. However, these categories overlap. Lamar University’s integration strategy explicitly treats them as guidelines rather than a rigid decision tree. The practical goal is to identify transformations that simplify the expression, while ruling out approaches that do not fit. Before committing to a lengthy calculation, inspect the entire integrand rather than reacting to its most noticeable feature. [2]

2. Substitution: simplify a function and its accompanying derivative

Simple substitution deserves an early check because it can avoid more elaborate calculations. As an original illustration, consider ∫ 2x/(x² − 1) dx. Choose u = x² − 1; then du = 2x dx. The numerator supplies exactly the differential needed, so the transformed integral is ∫ 1/u du = ln|u| + C. Substituting back gives ln|x² − 1| + C, valid on any interval that does not cross x = −1 or x = 1. This calculation illustrates the reference’s recommendation to try a quick substitution before advanced techniques. [2], [3]

The important recognition step is the relationship between x² − 1 and 2x, not merely the presence of a fraction. You can check the result directly: differentiating ln|x² − 1| gives 2x/(x² − 1). In this example, every part of the original integral has a role in the substitution; no unexplained x remains after changing variables. That makes it a particularly clean route. A rational-function technique is also possible here, but the substitution reveals the answer with less algebra. [2], [3]

3. Match the technique to the transformation you need

The following table is a selection guide, not a ranking or a guarantee. Its distinctions reflect the integration categories described in Lamar’s calculus materials and MIT’s course sequence. Notice that each technique changes the problem in a different way: partial fractions rewrites a rational expression, trigonometric substitution changes the variable, and quadratic manipulation prepares an expression for another method. Choosing well therefore means asking both what the integrand looks like and what the proposed transformation would accomplish. [2], [3], [1]

Pattern or obstacle	Technique to consider	Practical distinction
A simple change of variable is available	Substitution	Check this before more elaborate techniques. [2]
Polynomial × exponential, logarithm, or trigonometric function	Integration by parts	A useful candidate, not an automatic requirement. [2]
Polynomial divided by polynomial	Partial fractions	Rewrite suitable rational expressions into integrable pieces. [2], [3]
Products or powers of trigonometric functions	Trigonometric integration methods	Distinguish these from substituting a trigonometric expression for a variable. [1], [3]
A quadratic or root obstructs progress	Algebraic preparation or a suitable substitution	Manipulation may be needed before another technique becomes useful. [3], [1]

4. Integration by parts versus partial fractions

Integration by parts is worth considering for products such as a polynomial times an exponential, logarithm, or trigonometric function. Partial fractions, by contrast, targets suitable ratios of polynomials by rewriting them. Their distinction is therefore structural: spotting multiplication alone is not enough to choose parts, and spotting division alone does not make partial fractions the simplest option. Lamar’s strategy deliberately says these methods “may” work, leaving room for a simpler substitution or another transformation. [2], [3]

The earlier example makes that overlap concrete. Algebra gives 2x/(x² − 1) = 1/(x − 1) + 1/(x + 1). Integrating the two pieces produces ln|x − 1| + ln|x + 1| + C, which combines to ln|x² − 1| + C. Thus partial fractions and substitution reach the same result. This original calculation illustrates why classification should generate candidates rather than dictate a single method: the rational structure suggests partial fractions, but the derivative relationship makes substitution shorter. [2], [3]

5. Separate trigonometric integration from trigonometric substitution

Trigonometric integration begins with trigonometric functions already present—for example, products of sine and cosine or secant and tangent. MIT’s course distinguishes odd and even exponents in its treatment of sine–cosine powers. Trigonometric substitution is a different operation: it introduces a substitution involving trigonometric functions to simplify an integral. These topics belong in the same toolkit, but they are not interchangeable labels. Recognizing which situation you have prevents the vague instruction “use trig” from replacing an actual plan. [1], [3]

Algebra can also be the missing preparation step. The references explicitly include completing the square and discuss quadratics that require manipulation before earlier techniques become useful. Consequently, failure to recognize an immediate integration pattern is not necessarily a reason to abandon the problem: first ask whether rewriting the quadratic would expose a suitable form. For definite integrals, Lamar additionally flags subtleties in trigonometric substitution, so an indefinite-integral procedure should not be transferred mechanically without checking the definite-integral setup. [1], [3]

6. Know whether you need an exact value, an approximation, or convergence

Not every integration problem asks for the same kind of result. An exact evaluation seeks a precise expression; numerical approximation seeks a usable approximate value of a definite integral. Lamar’s materials explicitly include approximation because exact evaluation is not always available. An improper integral raises another question: does a finite value exist at all? Infinite intervals and discontinuous integrands can lead to improper integrals, which may converge or diverge. Comparison tests can sometimes settle that question without evaluating the integral explicitly. [3]

A practical workflow is therefore to identify the requested outcome, check for a quick substitution, classify the remaining structure, and compare promising techniques. If one transformation leaves another nontrivial integral, repeat the selection process rather than treating that as failure. Keep approximation and convergence methods available when exact evaluation is not the appropriate endpoint. This approach turns the techniques into a connected toolkit: the objective is a valid, manageable solution, not using the most advanced-looking method. [2], [3]

Frequently asked questions

Should every rational function be handled with partial fractions?

No. A rational expression makes partial fractions a candidate, not a requirement. A simple substitution may require much less work. Inspect the numerator and denominator together before committing to decomposition. [2]

Is there always one correct integration method?

No. Several methods can be valid, and their relative convenience can differ. Compare plausible approaches rather than assuming that the first recognizable technique is the only acceptable one. [2]

What if a substitution does not finish the problem?

It may still be a useful first step. A substitution can lead to an integral requiring integration by parts or partial fractions. Judge it by whether it makes progress toward a manageable expression, not whether it immediately produces an answer. [2]

Does an improper integral necessarily have an infinite value?

No. Improper integrals may or may not have finite values. Determining convergence is a separate task, and comparison tests can establish convergence or divergence even when direct evaluation is unavailable. [3]

When is numerical approximation appropriate?

Approximation is useful when a definite integral’s value is needed but exact evaluation is unavailable. It supplies an approximate value rather than an exact antiderivative; those are different outcomes, not competing versions of the same calculation. [3]

Conclusion

A reliable integration strategy starts with simple transformations and structural recognition. Try substitution early, distinguish products from rational expressions, and keep trigonometric integration separate from trigonometric substitution. Compare methods when several are plausible, and combine them when necessary. Finally, identify whether the task calls for exact evaluation, approximation, or a convergence decision. That flexibility is more useful than a rigid rule assigning one technique to every integral. [2], [3]

Explore next

Sources

Open a source or expand its cited passages.

Relevance scores are model estimates of how well each source addresses this question, not probabilities or fact-check ratings. Reasoned guidance is identified on hover.