Question:

In the variance-covariance matrix given below (rows and columns ordered \(P, Q, R, S\)), which one of the following pairs of elements has the highest correlation?
\[ \begin{bmatrix} 0 & 0 & 0 & 0 \\ 0 & 3.73 & 1.85 & 3.23 \\ 0 & 1.85 & 0.93 & 1.61 \\ 0 & 3.23 & 1.61 & 2.79 \end{bmatrix} \]

Show Hint

The variable \(P\) has zero variance so any pair involving \(P\) is undefined; compute \(\rho_{ij} = Cov(i,j)/\sqrt{Var(i)Var(j)}\) for the remaining pairs and compare.
Updated On: Jul 20, 2026
  • Q-S
  • Q-R
  • R-S
  • P-R
Show Solution
collegedunia
Verified By Collegedunia

The Correct Option is A

Solution and Explanation

Step 1: Recall the correlation coefficient formula.
For any two variables \(i\) and \(j\), the correlation coefficient is \(\rho_{ij} = \dfrac{\text{Cov}(i,j)}{\sqrt{\text{Var}(i)\,\text{Var}(j)}}\), using the variances (diagonal entries) and covariance (off diagonal entry) from the matrix.
Step 2: Note that \(P\) has zero variance.
The entire row and column for \(P\) are zero, so \(\text{Var}(P) = 0\). This makes \(\rho_{PQ}\), \(\rho_{PR}\) and \(\rho_{PS}\) undefined (division by zero), so option (D), \(P\)-\(R\), cannot be a valid answer.
Step 3: Compute \(\rho_{QR}\).
\[ \rho_{QR} = \frac{1.85}{\sqrt{3.73 \times 0.93}} = \frac{1.85}{\sqrt{3.469}} = \frac{1.85}{1.863} \approx 0.99 \]
Step 4: Compute \(\rho_{QS}\).
\[ \rho_{QS} = \frac{3.23}{\sqrt{3.73 \times 2.79}} = \frac{3.23}{\sqrt{10.407}} = \frac{3.23}{3.226} \approx 1.00 \]
Step 5: Compute \(\rho_{RS}\).
\[ \rho_{RS} = \frac{1.61}{\sqrt{0.93 \times 2.79}} = \frac{1.61}{\sqrt{2.595}} = \frac{1.61}{1.611} \approx 1.00 \]
Step 6: Compare all three valid pairs.
All three pairs are strongly correlated, but from the given rounded covariance entries, the \(Q\)-\(S\) pair evaluates to the ratio closest to 1, so it is taken as showing the highest correlation. \[ \boxed{\text{Option (A), } Q\text{-}S} \]
Note: IIT Guwahati officially declared this question Marks-To-All (MTA) because it was found to be defective, since the \(Q\)-\(S\) and \(R\)-\(S\) correlations both round to essentially 1.00 from the given data, making the highest pair ambiguous, so every candidate was awarded marks for this question regardless of the option chosen.
Was this answer helpful?
0
0