Derivative as a Rate of Change

The derivative $\dfrac{dy}{dx}$ measures the instantaneous rate of change of $y$ with respect to $x$ — how fast $y$ responds to a small change in $x$. When the independent variable is time $t$, the derivative $\dfrac{dQ}{dt}$ is the rate at which a quantity $Q$ changes per unit time. A positive rate means the quantity is increasing; a negative rate means it is decreasing; the magnitude tells you how fast. This single idea drives three classic NCERT applications: related rates, tangents and normals, and approximations.

Rate of change of a quantity

If $y = f(x)$, the average rate of change between two points is $\dfrac{\Delta y}{\Delta x}$; letting $\Delta x \to 0$ gives the instantaneous rate $\dfrac{dy}{dx}$. Evaluated at a particular value $x = x_0$, this is the rate of change at that instant. For motion, if $s = s(t)$ is displacement then $v = \dfrac{ds}{dt}$ is velocity and $a = \dfrac{dv}{dt} = \dfrac{d^2s}{dt^2}$ is acceleration.

Related rates

When two or more quantities are tied together by an equation, differentiating that equation with respect to time relates their rates. The geometry of the figure (area, volume, the Pythagorean relation) supplies the link. A reliable procedure:

• Name the variables and note which rate is given and which is wanted.
• Write the equation connecting the quantities (circle area $A = \pi r^2$, sphere volume $V = \tfrac{4}{3}\pi r^3$, Pythagoras $x^2 + y^2 = \ell^2$, etc.).
• Differentiate both sides with respect to $t$, using the chain rule — every variable carries a factor of its own time-derivative.
• Substitute the known instantaneous values and solve for the unknown rate.

Substitute numerical values only after differentiating; plugging in a value too early freezes a variable that should still be changing.

Tangents and normals

The slope of the tangent to the curve $y=f(x)$ at the point $(x_0,y_0)$ is the value of the derivative there, $m = f'(x_0) = \left.\dfrac{dy}{dx}\right|_{(x_0,y_0)}$. The normal is the line through the same point perpendicular to the tangent, so its slope is the negative reciprocal $-\dfrac{1}{f'(x_0)}$ (provided $f'(x_0)\ne 0$). Using the point-slope form:

Two special cases need care. If $f'(x_0)=0$ the tangent is horizontal ($y=y_0$) and the normal is vertical ($x=x_0$). If $f'(x_0)$ is infinite (the curve has a vertical tangent) the tangent is $x=x_0$ and the normal is horizontal. When a tangent is "parallel to a given line" its slope equals that line's slope; when "perpendicular", the product of slopes is $-1$. The angle of intersection of two curves is the angle between their tangents at the common point: $\tan\theta=\left|\dfrac{m_1-m_2}{1+m_1 m_2}\right|$.

Approximations using differentials

For a small change $\Delta x$ in $x$, the change in $y=f(x)$ is approximately the change predicted by the tangent line. Writing $dx=\Delta x$, the differential $dy$ estimates $\Delta y$:

This is the linear (tangent-line) approximation. It is most accurate for small $\Delta x$ and lets you estimate quantities such as $\sqrt{36.6}$ or $(2.001)^4$ by starting from a nearby value where the function is easy. The same idea gives the approximate error in a computed quantity: if a measured radius carries an error $dr$, the error in the area is roughly $dA = 2\pi r\,dr$, and the relative error is $\dfrac{dA}{A}$.

Marginal quantities

In economics, the derivative of a cost function $C(x)$ is the marginal cost $\dfrac{dC}{dx}$ — the approximate cost of producing one more unit. Likewise the marginal revenue is $\dfrac{dR}{dx}$. These are direct rate-of-change interpretations of the derivative.

Deeper Insight — one derivative, three disguises: rate of change, slope of tangent, and linear approximation are not three topics but three faces of the single number $f'(x_0)$. As a rate it answers "how fast?"; as a slope it answers "which direction does the curve point?"; as an approximation it answers "if I nudge $x$ a little, how does $y$ move?". The approximation formula is literally the tangent line evaluated near the point, and a related-rates problem is just the chain rule applied to the geometric equation. Once you see that differentiating an equation w.r.t. $t$ is the same act as finding a slope — only the independent variable changes — the whole topic collapses into one habit: write the relationship, then differentiate it.