Discrete Uniform Distribution

• Discrete uniform distribution is a discrete probability model for a finite set of distinct values where every value is equally likely. This means the probability mass function is constant across all allowed outcomes, so no value is favored. It applies when a random mechanism selects from a fixed list with perfect fairness.

• Support size is the number of possible outcomes, often denoted by $n$, and each probability is $\frac{1}{n}$. Because all mass values are identical, the distribution is fully determined once the outcome set is known. This makes setup simple but only when the equal-likelihood assumption is truly justified.

• Canonical form uses outcomes $1,2,\dots,n$, which gives standard closed-form formulas for moments. When outcomes are different numbers but still equally likely, the distribution is still uniform but may need transformation methods for quick moment calculations. This distinction prevents misuse of formulas outside their valid form.

• Normalization principle requires total probability to equal 1, so constant mass over $n$ values must be $\frac{1}{n}$. This is why equal likelihood and finiteness jointly force a unique pmf height. If either condition fails, the model is not discrete uniform.

• Symmetry principle explains why central tendency is stable: values are balanced around the midpoint of the support. In the canonical case, this gives mean and median at the center, while no single value has higher frequency than others. The lack of a unique highest probability is why there is typically no mode.

Core formulas (canonical support $1,2,\dots,n$)

$P(X=x)=\frac{1}{n}$ for $x\in\{1,2,\dots,n\}$, else $0$.

$E(X)=\frac{n+1}{2}$ and $\mathrm{Var}(X)=\frac{n^2-1}{12}$.

These formulas rely on equally spaced integer support from 1 to $n$, not just any equally likely labels.

Identification workflow

• Step 1: Verify finiteness by listing or characterizing all possible outcomes and confirming the set is finite. A model with infinitely many possible values cannot be discrete uniform in this syllabus context. This first check prevents incorrect formula use from the start.
• Step 2: Verify equal likelihood by reasoning from the mechanism, not by assumption. If any outcome is favored by design or repetition, the model is not uniform. This step is essential because equal probability is the defining structural condition.

Computation workflow

• Step 3: Compute probabilities by counting outcomes, using $P(X\in A)=\frac{|A|}{n}$ when all $n$ outcomes are equally likely. This avoids summing many identical terms and is usually the fastest method in exams. It works for inequalities, intervals, and set-based events.
• Step 4: Use moment formulas appropriately: for $X=1,2,\dots,n$, apply $E(X)=\frac{n+1}{2}$ and $\mathrm{Var}(X)=\frac{n^2-1}{12}$. For outcomes in an arithmetic sequence, map to canonical form using $Y=aX+b$ and then use $E(Y)=aE(X)+b$, $\mathrm{Var}(Y)=a^2\mathrm{Var}(X)$. This method is robust and reduces algebraic error.

Compare similar ideas carefully

• Discrete uniform vs general discrete distribution differs in pmf shape: uniform has constant mass, while general discrete models can assign different masses to different outcomes. This changes both method choice and interpretation of center and spread. Uniform models prioritize counting symmetry; non-uniform models require weighted sums.

• Canonical support vs transformed support is another critical distinction. Formulas $\frac{n+1}{2}$ and $\frac{n^2-1}{12}$ directly apply only to support $1,2,\dots,n$. For supports like arithmetic sequences, transform first or compute from definition to avoid structural mistakes.

• Misconception: equal spacing implies equal probability is false unless the random mechanism is also fair across outcomes. Spacing concerns values, while uniformity concerns probabilities assigned to those values. Always justify probability equality from the selection process.

• Pitfall: applying canonical formulas to arbitrary labels causes wrong means and variances. The formulas for $1,2,\dots,n$ are not plug-and-play for every equally likely set. Use transformation or direct expectation definitions when support is shifted or scaled.

• Inequality boundary errors such as confusing $P(X<k)$ with $P(X\le k)$ are common and costly. In discrete settings, a single boundary value can change the result by exactly one mass unit $\frac{1}{n}$. Always list included outcomes explicitly before counting.

• Connection to expectation and variance theory: discrete uniform is a clean case showing how distribution shape determines moments. Because masses are equal, averages reflect geometric center and spread grows systematically with support width. This makes it a useful benchmark for understanding more complex discrete models.

• Connection to linear transformations: if a variable is uniform on $1,\dots,n$, then any affine transform $Y=aX+b$ preserves uniformity over transformed values. The mean shifts and scales linearly, while variance scales by $a^2$, showing how location and spread respond differently. This links model setup directly to parameter interpretation.

• Connection to modeling practice: the distribution is appropriate for idealized random selectors over finite sets, such as index sampling or fair randomized assignment. It is inappropriate when outcomes have unequal exposure frequencies or structural duplication. Good modeling depends more on mechanism assumptions than on convenient formulas.