Answers are grounded in this paper and cite their sources.
Reinforced concrete (RC) roof slabs in institutional buildings deteriorate through cracking, leakage, corrosion, and surface degradation, yet maintenance still relies largely on visual inspection and reactive repair. This study developed and cross-validated an inspection-based model for estimating the current condition of RC roof slabs at the Dar es Salaam Institute of Technology and the Tanzania Institute of Accountancy. All 74 accessible panels were inspected with a structured form recording nine defect indices and a condition outcome (CO). Seventy panels in 17 buildings formed the primary dataset; four panels from newly constructed buildings were reserved for sensitivity analysis. Multiple linear regression with cluster-robust inference was fitted and validated by holding out one building at a time. A seven-factor model explained 97.2% of the variance in CO (adjusted R² = 0.969). A conventional, non-cluster-adjusted test did not reject equality of its coefficients, F (6, 62) = 1.16, p = .341, although this does not demonstrate that they are equal. An equal-weight index of the seven factors predicted held-out buildings at least as well (MAE = 1.46 and RMSE = 1.88 percentage points, prediction R² = 0.966, condition class correct for 64 of 70 panels under proposed classes). Maintenance deficiency and waterproofing defects accounted for the largest shares of explained variance, but most other rankings were uncertain. Because predictors and outcome were derived from the same inspections and may partly overlap, these figures should be regarded as upper-bound estimates; including the four panels from new buildings did not change the conclusions. The model estimates current rather than future condition and provides facility managers with a simple screening tool, pending independent verification.
No takes yet. Share an insight, caveat, or question.
Kilukibi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: