A balanced incomplete block design (BIBD) has parameters v = 4 (treatments), b = 6 (blocks), r = 3 (replicates per treatment), k = 2 (block size), λ = 1. The following adjusted treatment totals (after removing block effects) are given as Q₁ = 12, Q₂ = 8, Q₃ = 4 and Q₄ = 0. The MSE is given as 2·5.
(i) Verify if this is a valid BIBD.
(ii) Estimate the treatment effects and their variances.
(iii) Compute the efficiency factor of this BIBD relative to a randomized complete block design with the same number of replications.
15 marks · directive: Verify
The core of a good answer
(i) bk = vr = 12, λ(v − 1) = r(k − 1) = 3 and b ≥ v, so the parameters are valid (the design is all 6 pairs of 4 treatments). (ii) τ̂ᵢ = kQᵢ/(λv) = Qᵢ/2 = 6, 4, 2, 0 (3, 1, −1, −3 after centring, since the Qᵢ as given sum to 24 instead of 0); V(τ̂ᵢ) = k(v − 1)σ²/(λv²) = 0.9375 and V(τ̂ᵢ − τ̂ⱼ) = 2kσ²/(λv) = 2.5. (iii) E = λv/(rk) = 2/3 ≈ 0.667.
An outline written by upsc.help, not an official UPSC answer; every fact checked by a separate review. How we work
What the question asks
Check the two parameter identities and Fisher's inequality; use the intra-block estimator τ̂ = kQ/(λv) with its variance formulas, substituting the given MSE for σ²; and compute the efficiency factor as the ratio of the variance of a treatment difference in an RCBD with r replicates to that in the BIBD.
Answer outline
Introduction
v = 4, b = 6, r = 3, k = 2, λ = 1, σ̂² = MSE = 2.5. Intra-block analysis: Qᵢ = Tᵢ − (1/k)∑(block totals containing i), and ∑Qᵢ = 0 identically.
Part (i): Validity
- bk = 6 × 2 = 12 = vr = 4 × 3.
- λ(v − 1) = 1 × 3 = 3 = r(k − 1) = 3 × 1.
- Fisher's inequality b ≥ v: 6 ≥ 4.
- A design exists: the C(4, 2) = 6 pairs (12), (13), (14), (23), (24), (34) as blocks, each pair once.
Part (ii): Treatment effects
- τ̂ᵢ = kQᵢ/(λv) = 2Qᵢ/4 = Qᵢ/2: τ̂₁ = 6, τ̂₂ = 4, τ̂₃ = 2, τ̂₄ = 0.
- Remark: adjusted totals must satisfy ∑Qᵢ = 0; the printed values sum to 24. Contrasts are unaffected; centring under ∑τᵢ = 0 gives 3, 1, −1, −3.
Part (ii): Variances
- V(τ̂ᵢ − τ̂ⱼ) = 2kσ²/(λv) = 2 × 2 × 2.5/4 = 2.5.
- V(τ̂ᵢ) (under ∑τᵢ = 0) = k(v − 1)σ²/(λv²) = 2 × 3 × 2.5/16 = 0.9375.
- Standard error of a difference = √2.5 = 1.58.
Part (iii): Efficiency factor
- RCBD with r = 3 replicates: V(τ̂ᵢ − τ̂ⱼ) = 2σ²/r.
- BIBD: 2kσ²/(λv).
- E = (2σ²/r)/(2kσ²/(λv)) = λv/(rk) = 4/6 = 2/3 ≈ 0.667, assuming equal error variance in both designs.
Conclusion
The parameters form a valid (unreduced) BIBD; treatment effects are estimated as Qᵢ/2 with variance 0.9375 each and 2.5 for a difference; the efficiency factor is 2/3, so the BIBD loses one third of the information per comparison unless smaller blocks reduce σ² enough to compensate. Values computed in Python.
Examples and data to cite
- Here the information matrix C = rI − NN′/k = 2(I − J/4) has non-zero eigenvalue 2 against r = 3 for an RCBD, giving E = 2/3 (checked in Python).
Facts to remember
- BIBD conditions: bk = vr; λ(v − 1) = r(k − 1); b ≥ v (Fisher's inequality).
- Intra-block estimate: τ̂ᵢ = kQᵢ/(λv); ∑Qᵢ = 0.
- V(τ̂ᵢ − τ̂ⱼ) = 2kσ²/(λv).
- Efficiency factor E = λv/(rk) < 1.