Error Estimation in Linear Approximation
You'll learn how to estimate how far off a linear approximation might be and when to trust it most.
The Core Idea
When you use the tangent line to approximate a function near a point, you're not getting the exact value—there's an error. Understanding this error helps you know whether your approximation is good enough for practical use.
The error in linear approximation is the difference between the actual function value and the approximation:
Error = |f(x) - L(x)|
where L(x) is your linear approximation.
When Are Approximations Most Accurate?
Linear approximations work best when:
-
You stay close to the base point (small Δx): The farther you move from where you drew the tangent line, the more the curve pulls away from the straight line.
-
The function is "nearly linear": Functions with small second derivatives don't curve much, so their tangent lines stay close longer.
-
The function doesn't curve sharply: Think of approximating a gentle hill versus a sharp peak—the gentle hill's tangent line stays useful longer.
Estimating the Error
While the exact error formula involves concepts beyond our current scope, you can estimate error by:
- Checking how much the second derivative f''(x) varies near your point (larger |f''| means more curvature and more error)
- Testing your approximation at nearby points where you know the exact value
- Recognizing that error grows roughly proportionally to (Δx)²
Analogy: Imagine driving on a curving road but steering straight for a few seconds. If the road curves gently, you stay close to the road. If it curves sharply, you quickly veer off.
Key Takeaway: The error in linear approximation grows as you move farther from your base point and increases with the function's curvature; approximations are most reliable when Δx is small and f''(x) is small.