second derivative test multivariable

Second Derivative Test Multivariable: Simple Guide with Examples

By Laura Rodriguez
AD

If you are studying multivariable calculus, the second derivative test multivariable version is the tool that turns critical points from vague suspects into classified features of a surface. In single-variable calculus, the second derivative tells you whether a critical point is a peak or a valley. In higher dimensions, the story is richer because a function can curve upward in one direction and downward in another, creating a saddle point that has no true one-dimensional equivalent.

This article breaks down the test in practical language while keeping the mathematical precision you need for exams and real applications. You will see how the Hessian matrix organizes second partial derivatives, how the determinant and leading principal minors drive the classification, and how to avoid the mistakes that cause students to mislabel local maxima, local minima, and saddle points.

What the Multivariable Second Derivative Test Measures

The multivariable second derivative test inspects the local curvature of a function f(x, y) or f(x, y, z) near a critical point. A critical point occurs where the gradient is zero or where the gradient is undefined. At such a point, the first derivative test is inconclusive because the tangent plane is horizontal in the differentiable case, so the function's behavior depends on second-order terms.

AD
IGCSE 0606 — Stationary Points: The 2nd Derivative Test
IGCSE 0606 — Stationary Points: The 2nd Derivative Test
Second Derivative Test
Second Derivative Test
2nd Order Derivatives Cheat Sheet | Mathematics Free Notes PDFs | Second Derivative Formulas
2nd Order Derivatives Cheat Sheet | Mathematics Free Notes PDFs | Second Derivative Formulas

The test uses second partial derivatives to build a matrix called the Hessian. For a function of two variables, the Hessian collects fxx, fxy, fyx, and fyy. When the mixed partial derivatives are continuous, fxy equals fyx, so the Hessian is symmetric. That symmetry is not just a technical detail; it makes the eigenvalues real and allows the determinant and leading principal minors to summarize curvature efficiently.

From One Variable to Two Variables: The Hessian Matrix

In one-variable calculus, the second derivative f double prime at a critical point tells you whether the graph bends upward or downward. For two variables, a single number is not enough. You need a matrix to describe how the function bends in every direction. The Hessian matrix H at a point (a, b) is written with fxx, fxy, fyx, and fyy arranged in a 2 by 2 grid.

Use the Second Derivative Test to Find all Relative Extrema f(x) = x^3 - 3x^2 + 2
Use the Second Derivative Test to Find all Relative Extrema f(x) = x^3 - 3x^2 + 2
SECOND ORDER DERIVATIVES
SECOND ORDER DERIVATIVES
Application of Derivatives Formulae
Application of Derivatives Formulae

The Hessian acts like a multidimensional second derivative. Positive eigenvalues correspond to upward curvature in their eigenvector directions, while negative eigenvalues correspond to downward curvature. If all eigenvalues are positive, the surface opens upward locally in every direction. If all are negative, it opens downward locally. If the eigenvalues have mixed signs, the surface is saddle-shaped.

The Discriminant and Its Role in Classification

For functions of two variables, the discriminant D is often defined as fxx times fyy minus fxy squared. This is exactly the determinant of the Hessian matrix. The sign of D, together with the sign of fxx, gives a quick classification rule at a critical point. If D is positive and fxx is positive, the point is a local minimum. If D is positive and fxx is negative, the point is a local maximum.

AD

If D is negative, the critical point is a saddle point, regardless of fxx. If D equals zero, the test is inconclusive, and you need another method such as examining higher-order terms, using a parameterization, or applying a more advanced theorem. This inconclusive case is common in textbook problems because it forces you to understand the geometry rather than just plug numbers into a formula.

Graph of the first and second derivative
Graph of the first and second derivative
Derivatives Formula Sheet | Calculus Derivatives, Rules & Applications | JEE, NEET & MHT CET Maths
Derivatives Formula Sheet | Calculus Derivatives, Rules & Applications | JEE, NEET & MHT CET Maths
Ex 5.4 Q2(ii) Second Order Derivative | Prove d²y/dx² = 49y | Class 12 Applied Maths
Ex 5.4 Q2(ii) Second Order Derivative | Prove d²y/dx² = 49y | Class 12 Applied Maths
Second Derivative Problems Explained (Step-by-Step) | Calculus Exam Revision Guide
Second Derivative Problems Explained (Step-by-Step) | Calculus Exam Revision Guide
Finding The Derivative Color Worksheet #1
Finding The Derivative Color Worksheet #1
Applications of Differentiation Lesson 4 of 7
Applications of Differentiation Lesson 4 of 7

How to Apply the Second Derivative Test Step by Step

Applying the multivariable second derivative test follows a predictable workflow. First, find all critical points by solving the system where the first partial derivatives are zero. Second, compute the second partial derivatives. Third, evaluate the Hessian at each critical point. Fourth, classify the point using the determinant and appropriate minor conditions. The order matters because a single arithmetic error in the first derivatives can make every later conclusion wrong.

For functions of more than two variables, the classification uses the eigenvalues of the Hessian or the signs of its leading principal minors. A local minimum requires all eigenvalues to be positive, which means the Hessian is positive definite. A local maximum requires all eigenvalues to be negative, meaning the Hessian is negative definite. Mixed signs indicate a saddle point, while zero eigenvalues leave the test inconclusive.

AD

Worked Example: A Simple Quadratic Surface

Consider f(x, y) = x squared + y squared. The first partial derivatives are 2x and 2y, so the only critical point is (0, 0). The second partial derivatives are fxx = 2, fyy = 2, and fxy = 0. The discriminant is D = 2 times 2 minus 0 squared, which equals 4. Since D is positive and fxx is positive, the point (0, 0) is a local minimum. The surface is a bowl opening upward.

Standard Derivatives Formula Blueprint | Calculus Cheat Sheet for JEE, MHT CET & Boards
Standard Derivatives Formula Blueprint | Calculus Cheat Sheet for JEE, MHT CET & Boards
Derivatives Formula Sheet!!!💗💗
Derivatives Formula Sheet!!!💗💗
Limit Definition of the Derivative- Multiple Representation Reference Sheet
Limit Definition of the Derivative- Multiple Representation Reference Sheet
Prepare for board exams-Definite integral problems cover 5-15 marks
Prepare for board exams-Definite integral problems cover 5-15 marks

Now consider f(x, y) = x squared minus y squared. The critical point is again (0, 0), with fxx = 2, fyy = -2, and fxy = 0. The discriminant is D = 2 times (-2) minus 0, which equals -4. Because D is negative, the point is a saddle point. Moving along the x-axis increases the function, while moving along the y-axis decreases it. This contrast is the signature of a saddle.

Common Pitfalls and Special Cases

One common mistake is to assume that D positive always means a local extremum without checking fxx. A positive determinant alone tells you that the curvature has the same sign in the principal directions, but it does not tell you whether that sign is positive or negative. You must pair D with fxx for a two-variable function, or with eigenvalue signs for higher dimensions.

AD

Another trap is forgetting that the second derivative test only applies to critical points. If the gradient is not zero at a point, the Hessian may still be interesting, but it does not classify that point as a local extremum. Also, functions with discontinuous second partial derivatives can violate the expected symmetry of the Hessian. In most standard calculus settings, however, the mixed partials are continuous, so the test behaves exactly as described.

When you next work through a multivariable optimization problem, resist the urge to memorize only the final signs. Sketch the surface, identify the candidate points, and ask what the Hessian says about curvature in different directions. That habit will make the second derivative test multivariable feel less like a formula and more like a reliable lens for seeing the shape hidden inside an equation.

AD