Continuous Uniform Distribution

• Continuous uniform distribution is a continuous model on an interval $[a,b]$ where the density is constant and zero outside that interval. Because total probability must be 1, the constant height is fixed by area balance rather than chosen freely. This gives a rectangular density shape that represents equal likelihood per unit length, not equal point probabilities.

Core PDF: $f(x)=\begin{cases}\frac{1}{b-a},&a\le x\le b\\0,&\text{otherwise}\end{cases}$

• Equal density does not mean point probabilities are nonzero in a continuous setting. Any single value still has probability 0, while intervals have nonzero probability based on width. This distinction explains why the model is called uniform over an interval even though exact outcomes are not assigned positive mass.

• Normalization principle determines the density height from geometry: rectangle area equals 1. Since width is $(b-a)$, the height must be $\frac{1}{b-a}$ so that width times height is exactly 1. This is why every valid continuous uniform distribution is fully determined once $a$ and $b$ are known.

• Probability as area gives a direct rule for any sub-interval. For $a\le c<d\le b$, $P(c\le X\le d)=\int_c^d \frac{1}{b-a}\,dx=\frac{d-c}{b-a}.$ The result depends only on interval length, which is the defining invariance property of uniformity.

• Symmetry about midpoint explains central tendency results. The interval midpoint $\frac{a+b}{2}$ balances equal lengths to the left and right, so mean and median coincide there. This symmetry also supports quick plausibility checks when solving problems.

• Step 1: identify support correctly by writing the interval where density is nonzero. If a value lies outside $[a,b]$, its probability is immediately 0, which prevents unnecessary calculations. This support-first habit is the fastest way to avoid setup errors.

• Step 2: choose rectangle-length method or integration based on question form. For plain interval probabilities, use length ratio $\frac{d-c}{b-a}$ because it is faster and less error-prone. Use integration form when deriving formulas or when transformations make direct length reasoning less obvious.

• Step 3: compute summary measures with formulas and meaning. Use $E(X)=\frac{a+b}{2},\quad \mathrm{Var}(X)=\frac{(b-a)^2}{12},\quad \sigma=\frac{b-a}{\sqrt{12}}.$ The mean locates the center, while variance and standard deviation scale with interval width, so wider intervals imply greater spread.

• Uniform continuous vs discrete uniform differ in what is equally likely. Continuous uniform has equal density per unit interval and $P(X=x)=0$ for each single point, while discrete uniform assigns equal positive probability to each listed value. This distinction changes how probabilities are computed: area/integration versus summation.

• Interval probability vs point probability is a core exam distinction. In continuous models, including or excluding endpoints does not change interval probability, so $P(c\le X\le d)=P(c<X<d)$. Treating point events as nonzero is a conceptual mismatch from discrete intuition.

• Comparison table helps method selection under time pressure. The table highlights what stays the same and what changes across common continuous models.

• Confusing density value with probability is a frequent error. The value $f(x)=\frac{1}{b-a}$ is a height, not the probability at a point, so it can be greater than 1 when the interval is narrow. Probability comes from area over an interval, not from reading a single y-value.

• Forgetting to clip intervals to support leads to impossible results above 1 or below 0. If asked for $P(c\le X\le d)$ with bounds outside $[a,b]$, replace them with overlap bounds before computing. This enforces valid probabilities and keeps geometry consistent.

• Assuming a unique mode exists is incorrect for a flat density. Every point in $[a,b]$ attains the same maximum density, so there is no single most likely value. Recognizing this avoids forcing a non-existent peak via differentiation.

• Continuous uniform is the baseline model of complete interval uncertainty and often appears as a first approximation when only bounds are known. It is mathematically simple but conceptually rich because it trains interval-based probabilistic thinking. This makes it a bridge from deterministic ranges to full stochastic modeling.

• It connects directly to linear transformations: if $Y=mX+c$, then $Y$ is also uniform on $[ma+c,mb+c]$ (with endpoints ordered). This shows how measurement-unit changes preserve distribution family while shifting and scaling location/spread. The rule supports rapid conversion between equivalent problem statements.

• It also links to simulation and Monte Carlo methods because many random generators start from uniform draws on $[0,1]$. More complex distributions are often built by transforming uniform variables. Understanding uniform structure is therefore foundational for computational probability.