Statistics: PYQ Walkthrough (2)

medium 4 min read

Question

(CBSE 2023 style.) The mean and variance of a set of 5 observations {x1,x2,x3,x4,x5}\{x_1, x_2, x_3, x_4, x_5\} are 4 and 5 respectively. If three of the observations are 1, 3, 8, find the other two and the standard deviation.

Solution — Step by Step

Mean =x1+x2+x3+x4+x55=4= \dfrac{x_1 + x_2 + x_3 + x_4 + x_5}{5} = 4, so the sum is 20.

Three given: 1+3+8=121 + 3 + 8 = 12. Let the other two be aa and bb. Then a+b=2012=8a + b = 20 - 12 = 8.

Variance =1n(xixˉ)2=5= \dfrac{1}{n}\sum(x_i - \bar{x})^2 = 5, so (xi4)2=25\sum(x_i - 4)^2 = 25.

For the three given values: (14)2+(34)2+(84)2=9+1+16=26(1-4)^2 + (3-4)^2 + (8-4)^2 = 9 + 1 + 16 = 26.

Wait — this already exceeds 25. So the unknown two contribute negatively? No — squares can’t be negative. Let me re-read the problem.

The problem must have variance \geq 26/5 just from the three known values. So variance 5 might be impossible. Let me re-examine: perhaps the data given is in a different setup.

Re-checking: (xi4)2=5×5=25\sum(x_i - 4)^2 = 5 \times 5 = 25. From the three given: 2626. So the constraint is 26+(a4)2+(b4)2=2526 + (a-4)^2 + (b-4)^2 = 25, which is impossible since squares are non-negative.

This means the problem as stated has no solution — which suggests we should use the formula Var=1nxi2xˉ2\text{Var} = \dfrac{1}{n}\sum x_i^2 - \bar{x}^2 instead, and check using xi2\sum x_i^2.

Var=1nxi2xˉ25=15xi216xi2=105\text{Var} = \dfrac{1}{n}\sum x_i^2 - \bar{x}^2 \Rightarrow 5 = \dfrac{1}{5}\sum x_i^2 - 16 \Rightarrow \sum x_i^2 = 105.

From given: 12+32+82=1+9+64=741^2 + 3^2 + 8^2 = 1 + 9 + 64 = 74.

So a2+b2=10574=31a^2 + b^2 = 105 - 74 = 31. Combined with a+b=8a + b = 8: (a+b)2=64=a2+2ab+b2=31+2ab(a+b)^2 = 64 = a^2 + 2ab + b^2 = 31 + 2ab, so ab=16.5ab = 16.5.

The pair (a,b)(a, b) satisfies a+b=8a + b = 8 and ab=16.5ab = 16.5. The quadratic t28t+16.5=0t^2 - 8t + 16.5 = 0 has discriminant 6466=2<064 - 66 = -2 < 0 — no real solutions.

This particular numerical setup yields no real a,ba, b — illustrating that not every combination of mean, variance, and partial data is feasible. In standard CBSE problems with valid numbers, you’d get real answers. For example, if the variance were 9.2 instead of 5, xi2=5(9.2+16)=126\sum x_i^2 = 5(9.2 + 16) = 126, giving a2+b2=52a^2 + b^2 = 52, ab=6ab = 6, a+b=8a+b=8, so quadratic t28t+6=0t^2 - 8t + 6 = 0 giving t=4±10t = 4 \pm \sqrt{10}.

Standard deviation = variance=52.24\sqrt{\text{variance}} = \sqrt{5} \approx 2.24.

Why This Works

The two-equation system (sum and sum-of-squares) uniquely determines two unknowns when the data is consistent. The trick is using the variance formula in its computational form Var=x2xˉ2\text{Var} = \overline{x^2} - \bar{x}^2, which only needs x\sum x and x2\sum x^2.

When the given data is inconsistent (as in our specific numbers), the quadratic for a,ba, b has no real roots — a sign that something is off in the problem setup.

Always use Var=x2xˉ2\text{Var} = \overline{x^2} - \bar{x}^2 rather than (xxˉ)2\overline{(x-\bar{x})^2} in calculations. The first form needs only xi2\sum x_i^2, the second needs each individual deviation. Saves time.

Alternative Method

If asked for the SD directly: SD=Var=52.236\text{SD} = \sqrt{\text{Var}} = \sqrt{5} \approx 2.236. No need to find aa and bb first if SD is the only ask.

Students compute variance from “sum of deviations” without squaring — that always equals zero! The variance involves squared deviations: (xixˉ)2(x_i - \bar{x})^2.

Final answer: SD =52.236= \sqrt{5} \approx 2.236. The two unknowns satisfy a+b=8a + b = 8 and a2+b2=31a^2 + b^2 = 31 (this specific numerical example has no real solution; in valid problems, solve the quadratic).

Want to master this topic?

Read the complete guide with more examples and exam tips.

Go to full topic guide →

Try These Next