top of page

ISS 2017 Statistics Paper-1 Solution: Question 10 (Variance of Sum of Uniform Variables)

1 day ago
5 min read

Continuing our question-by-question walk through the ISS Statistics Paper-1 archive, we now reach 2017, Question 10 — a short but important problem on the variance of a sum of independent uniform random variables. If you have been following this series from the start, you already know the drill: no shortcuts, every step shown.


Quick Summary


  • Topic: Variance of a continuous Uniform distribution and variance of a sum of independent random variables

  • Reference: ISS Statistics Paper-1, 2017, Question 10

  • Correct Answer: Option (a) — 3/2


The Question


Let X and Y be independent random variables, each having the uniform distribution on [-1, 2]. Then Var(X + Y) is equal to


  • (a) 3/2

  • (b) 5/2

  • (c) 4

  • (d) 6

Step 1: Write Down the Density of X

X is uniform on the interval [-1, 2]. For a Uniform(a, b) distribution, the probability density function is constant over the interval and equal to 1/(b - a) everywhere inside it, and zero outside. Here a = -1 and b = 2, so


f(x) = 1/(2 - (-1)) = 1/3, for -1 ≤ x ≤ 2, and f(x) = 0 otherwise.


Y has exactly the same density, since it is described as having "the uniform distribution on [-1, 2]" as well.

Step 2: Compute E(X) by Direct Integration

Rather than quoting the mean formula from memory, let's derive it. For a continuous random variable, E(X) = ∫ x f(x) dx over the support.


E(X) = ∫-12 x · (1/3) dx = (1/3) ∫-12 x dx


Now integrate x with respect to x: ∫ x dx = x2/2. Apply the limits -1 to 2:


∫-12 x dx = (22/2) - ((-1)2/2) = (4/2) - (1/2) = 2 - 0.5 = 1.5 = 3/2


So E(X) = (1/3) × (3/2) = 1/2.

Step 3: Compute E(X²) by Direct Integration

To get the variance we also need the second moment, E(X²) = ∫ x² f(x) dx.


E(X²) = (1/3) ∫-12 x2 dx


Integrate x² with respect to x: ∫ x² dx = x3/3. Apply the limits -1 to 2:


∫-12 x2 dx = (23/3) - ((-1)3/3) = (8/3) - (-1/3) = 8/3 + 1/3 = 9/3 = 3


So E(X²) = (1/3) × 3 = 1.

Step 4: Find Var(X) Using the Standard Identity

For any random variable, Var(X) = E(X²) - [E(X)]². Substituting the values found above:


Var(X) = 1 - (1/2)2 = 1 - 1/4 = 3/4


Since Y has the identical distribution to X, by exactly the same working, Var(Y) = 3/4 as well. (In fact, there is a direct formula for a Uniform(a, b) variable: Var(X) = (b - a)2 / 12. Here (b - a) = 2 - (-1) = 3, so Var(X) = 9/12 = 3/4 — matching what we derived from first principles, which is a good consistency check.)


Pro Tip: The step students most often fumble here is computing (b - a) for an interval that starts at a negative number. It is tempting to write b - a as "2 - 1" out of habit, giving the wrong width of 1 instead of the correct width of 3. Always subtract the lower limit as written, sign included: 2 - (-1) = 3, not 2 - 1. This is exactly the kind of small sign slip that costs marks in a timed exam, and it is usually the first thing drilled into students until it becomes automatic — working a few of these shifted-interval uniform problems by hand, the way it's typically taught, is enough to make the habit stick for good.

Step 5: Combine Using Independence

The question asks for Var(X + Y), not Var(X) or Var(Y) individually. In general, for two random variables,


Var(X + Y) = Var(X) + Var(Y) + 2·Cov(X, Y)


We are told X and Y are independent, and independence implies Cov(X, Y) = 0. So the covariance term vanishes entirely, leaving simply:


Var(X + Y) = Var(X) + Var(Y) = 3/4 + 3/4 = 6/4 = 3/2

Step 6: Match Against the Options and the Official Key

Our derived value is 3/2, which corresponds to option (a). Checking against the official ISS 2017 Paper-1 answer key (Series A), item 10 is indeed marked as (a). Our independent derivation and the official key are in full agreement — there is no discrepancy to flag for this question.

Why This Question Matters

This problem packs three separate ideas into one short question: recalling (or deriving) the moments of a Uniform distribution, applying the variance identity Var(X) = E(X²) - [E(X)]², and correctly using the independence condition to simplify Var(X + Y). ISS Paper-1 regularly tests exactly this kind of compound application rather than a single isolated formula, so being comfortable chaining these three steps together under time pressure is a genuinely useful exam skill, not just a one-off trick for this particular question.

Frequently Asked Questions

Why is Var(X + Y) not simply 2 × Var(X) when X and Y have the same distribution?

It happens to equal 2 × Var(X) in this specific case only because Cov(X, Y) = 0, which follows from independence. If X and Y were identically distributed but correlated, the cross term 2·Cov(X, Y) would not vanish, and the simple doubling would no longer hold.

What is the general formula for the variance of a Uniform(a, b) distribution?

Var(X) = (b - a)² / 12. It is worth memorizing this after deriving it once by hand, as we did above, so that you understand why it works rather than just reciting it.

Does independence of X and Y also mean E(XY) = E(X)E(Y)?

Yes. Independence is actually a stronger condition than zero covariance — it implies E(XY) = E(X)E(Y), which in turn implies Cov(X, Y) = E(XY) - E(X)E(Y) = 0. The reverse is not generally true: zero covariance does not always imply independence.

Could this question have been solved without computing E(X) and E(X²) separately?

Yes, by directly quoting Var(X) = (b - a)² / 12 for a Uniform(a, b) variable. We solved it from first principles here so the formula isn't a mystery, but in the actual exam, using the direct formula is faster and perfectly acceptable once you trust it.

What is a common mistake in identifying the interval [-1, 2] for this kind of problem?

Misreading the lower limit's sign and computing the width as 2 - 1 = 1 instead of 2 - (-1) = 3. Since the width is squared in the variance formula, this single sign error changes the final answer by a factor of 9, so it is worth double-checking every time.

If the question had asked for Var(X - Y) instead, would the answer change?

No. Since Var(X - Y) = Var(X) + Var(Y) - 2·Cov(X, Y), and Cov(X, Y) = 0 by independence, we would still get Var(X - Y) = Var(X) + Var(Y) = 3/4 + 3/4 = 3/2 — the same numerical answer as Var(X + Y) in this particular case.


That wraps up Question 10 of the ISS 2017 Statistics Paper-1. If any step above wasn't clear, or you'd like to see a particular part worked out differently, drop a comment below — and if this helped, do share it with a fellow aspirant working through the same paper.

Recent Posts

See All

Comments


bottom of page