ISS 2017 Statistics Paper-1 Solution: Question 11 (Sum of Independent Normal Variables)
Continuing our question-by-question walk through the ISS Statistics Paper-1 archive, today we land on a question from the 2017 paper that tests one of the most heavily used results in the entire syllabus: what happens to the normal distribution when you add several independent normal variables together. If you have been following this series from the earlier 2017 questions, this one will feel like a nice break after the probability-heavy run we just finished.
Quick Summary
Topic: Sum of independent normal random variables
Question reference: ISS 2017, Statistics Paper-1, Question 11
Correct answer: Option (c), N(150, 625)
The Question
In an examination, marks obtained by the students in Physics, Mathematics and Statistics, denoted by X, Y and Z, are normally distributed with mean 50, 52 and 48 respectively and with standard deviation 15, 12 and 16 respectively. The distribution of (X + Y + Z) is
(a) N(350, 53)
(b) N(55, 350)
(c) N(150, 625)
(d) N(625, 150)
Step 1: Write Down What You Are Given
We are told three random variables:
X, the Physics marks, follow a normal distribution with mean μX = 50 and standard deviation σX = 15, so X ∼ N(50, 15²) = N(50, 225).
Y, the Mathematics marks, follow a normal distribution with mean μY = 52 and standard deviation σY = 12, so Y ∼ N(52, 12²) = N(52, 144).
Z, the Statistics marks, follow a normal distribution with mean μZ = 48 and standard deviation σZ = 16, so Z ∼ N(48, 16²) = N(48, 256).
A question like this always implicitly assumes X, Y and Z are independent of each other, since no joint information or correlation is given, and the problem is only solvable under that assumption. We will use that throughout.
Step 2: Recall the Additive Property of the Normal Distribution
This is the single fact the entire question hinges on, so let's build it up carefully rather than just quoting it.
If X1, X2, ..., Xn are independent normal random variables with Xi ∼ N(μi, σi²), then their sum S = X1 + X2 + ... + Xn is also normally distributed, with:
Mean of S = μ1 + μ2 + ... + μn (means simply add, always, independence or not)
Variance of S = σ1² + σ2² + ... + σn² (variances add only when the variables are independent)
Why does the mean always add, with or without independence? Because expectation is a linear operator: E(X + Y + Z) = E(X) + E(Y) + E(Z), and this identity needs no independence assumption at all, it holds for any random variables as long as their expectations exist.
Why does the variance need independence? Because in general, for any two random variables:
Var(X + Y) = Var(X) + Var(Y) + 2Cov(X, Y)
If X and Y are independent, Cov(X, Y) = 0, and the cross term vanishes, leaving Var(X + Y) = Var(X) + Var(Y). Extend this to three variables and the same logic gives:
Var(X + Y + Z) = Var(X) + Var(Y) + Var(Z) + 2Cov(X,Y) + 2Cov(Y,Z) + 2Cov(X,Z)
and with independence all three covariance terms are zero, so we are simply left with the sum of the individual variances.
Finally, why is the sum itself normal and not some other shape? This follows from the moment generating function (MGF) of the normal distribution. The MGF of a normal variable X ∼ N(μ, σ²) is MX(t) = exp(μt + σ²t²/2). For independent variables, the MGF of the sum is the product of the individual MGFs:
MX+Y+Z(t) = MX(t) · MY(t) · MZ(t) = exp[(μX+μY+μZ)t + (σX²+σY²+σZ²)t²/2]
This is exactly the MGF form of a normal distribution with mean (μX+μY+μZ) and variance (σX²+σY²+σZ²). Since the MGF uniquely determines a distribution, the sum must itself be normal with those exact parameters.
Pro Tip: The single most common slip on this type of question is forgetting that it is
Step 3: Compute the Mean of X + Y + Z
Using the additive property:
Mean = 50 + 52 + 48 = 150
Step 4: Compute the Variance of X + Y + Z
First square each standard deviation to get the individual variances:
15² = 225
12² = 144
16² = 256
Now add them, since X, Y, Z are independent:
Variance = 225 + 144 + 256
Let's add this carefully in two steps: 225 + 144 = 369, and then 369 + 256 = 625.
So Variance = 625.
Step 5: State the Final Distribution
Putting the mean and variance together:
X + Y + Z ∼ N(150, 625)
This is exactly option (c). Notice how options (a), (b) and (d) are all built from the same three numbers (150, 625, and a distractor like 350 or 53) just arranged in the wrong order or paired incorrectly, which is a very common distractor design in ISS papers: the numbers are all familiar, but only one arrangement is actually correct.
This matches the official ISS 2017 answer key, which lists option (c) as the correct response for this question, so our independently derived answer is confirmed.
Why This Question Matters
This question looks almost too simple for a competitive exam like ISS, but that is precisely the point: papers routinely include a handful of questions built entirely around a single clean theorem, specifically to reward candidates who have internalized the result instead of needing to re-derive it from first principles under time pressure. The additive property of independent normal variables underlies far more than just adding some exam marks: it is the backbone of results like the distribution of sample means, the Central Limit Theorem's exact (not just asymptotic) special case when the parent distribution is already normal, and plenty of multivariate normal theory you will meet later in the syllabus. Getting fluent with the rule that means always add, variances add only under independence, and the sum of independent normals is again normal, pays off across dozens of future questions, not just this one.
Frequently Asked Questions
Do the means of X, Y and Z need to be independent for their sum's mean to add up directly?
No. The mean of a sum always equals the sum of the means, regardless of whether the variables are independent or correlated in any way. This is simply the linearity property of expectation, and it holds unconditionally for any random variables with finite expectations.
What would change if X, Y and Z were not independent?
The mean of the sum would still be 150, since linearity of expectation never requires independence. However, the variance would need extra covariance terms: Var(X+Y+Z) = 225 + 144 + 256 + 2Cov(X,Y) + 2Cov(Y,Z) + 2Cov(X,Z), and without being told those covariances, the question would become unsolvable, which is exactly why independence is implicitly assumed here.
Why is the sum of independent normal variables again normal, and not some other distribution?
This follows from moment generating functions. The MGF of a sum of independent variables is the product of their individual MGFs, and when you multiply the MGFs of normal variables together, the result is algebraically identical in form to the MGF of another normal distribution with the summed mean and summed variance. Since an MGF uniquely identifies its distribution, the sum must be normal.
Is this the same as the Central Limit Theorem?
Not exactly. The Central Limit Theorem is an asymptotic result that says sums of many independent variables approach a normal distribution as the sample size grows, regardless of the parent distribution. Here, X, Y and Z are already individually normal, so their sum is exactly normal for any sample size, with no approximation involved.
How can I avoid mixing up standard deviation and variance under exam pressure?
Make it a fixed habit to immediately square every standard deviation you are given before doing any further arithmetic, and only write SD or take a square root again at the very final step if the question specifically asks for it. Writing N(mean, variance) instead of N(mean, SD) as your default notation while solving also helps avoid this error.
Would the answer change if the question asked for the distribution of the average of X, Y and Z instead of their sum?
Yes. For the average (X+Y+Z)/3, the mean would be 150/3 = 50, but the variance would be divided by 3² = 9 (since Var(aX) = a²Var(X)), giving 625/9, not 625/3. This scaling rule for variance under a constant multiplier is another frequently tested detail.
If anything in this derivation was unclear, or you spotted a quicker way to reach the same answer, drop a comment below, we would love to hear it, and if you found this useful, consider sharing it with a fellow ISS aspirant who might be stuck on the same question.

Comments