Corrigendum to “Assessing energy efficiency in the built environment: A quantile regression analysis of CO2 emissions from buildings and manufacturing sector” [Energy Build. 338 (2025) 115733] (Energy & Buildings (2025) 338, (S0378778825004633), (10.1016/j.enbuild.2025.115733))
Awan A. Kocoglu M. Subhan M. Utepkaliyeva K. bte Mohamed Yusoff N.Y. Hossain M.E.
1 October 2025Elsevier Ltd
Energy and Buildings
2025#344
The authors regret any confusion caused by the presentation of the Bayesian estimation results in the original article. To enhance transparency and reproducibility, Fig. 4 and Table 7 have been updated based on a refined estimation procedure incorporating improved convergence diagnostics. Specifically, the Bayesian multilevel model was re-estimated using the bayes: prefix in Stata 17, employing a random-effects panel specification (xtreg, re) with six parallel Markov chains. Each chain was run for 40,000 iterations, with the first 10,000 discarded as burn-in and a thinning interval of 6 applied, resulting in 5,000 post–burn-in draws per chain (30,000 in total). Convergence was rigorously assessed using the Gelman–Rubin potential scale reduction factor (R̂), effective sample size (ESS), and trace plots. All R̂ values remained well below 1.05, and ESS values exceeded 400 for all parameters, indicating strong chain mixing and convergence. As a result of this refinement: Fig. 4 now displays well-behaved trace plots demonstrating robust convergence;[Figure presented] Fig. 4. Multilevel regression posterior distributions of random effects. Table 7 presents the re-estimated posterior results, which are fully consistent with the original findings; [Formula presented] [Formula presented] A new Bayesian Estimation Summary Table A has been added, reporting key diagnostics (ESS, autocorrelation time, efficiency, and R̂) to improve methodological transparency. In re-estimating the model, we used Statas default prior settings. Specifically, regression coefficients were assigned normal priors, N (0, 10,000), while variance parameters received inverse-gamma priors, InvGamma (0.01, 0.01). Thus, the updated results corroborate the original findings. Importantly, the re-estimation did not alter the direction, magnitude, or interpretation of the models substantive conclusions. All coefficient signs and credible intervals are consistent with the original version, reaffirming the validity of the studys results. These updates strengthen the analytical rigor of the study without modifying its conclusions. The authors apologize for any inconvenience this may have caused.
Text of the article Перейти на текст статьи
Kashmir Institute of Economics, University of Azad Jammu and Kashmir, Pakistan
Institute of Energy Policy and Research, Universiti Tenaga Nasional Malaysia, Malaysia
Erciyes University, Kayseri, 38280, Turkey
Prague University of Economics and Business, Faculty of Finance and Accounting. W, Churchill Sq. 4, Prague 3, 130 67, Czech Republic
Marwadi University Research Center, Faculty of Management Studies, Marwadi University, Gujarat, Rajkot, 360003, India
Department of Economics, Atyrau University Named After Kh. Dosmukhamedov, Atyrau, Kazakhstan
UNITEN Business School, Institute of Energy Policy and Research, Universiti Tenaga Nasional, Malaysia
Department of Agricultural Sciences, Texas State University, San Marcos, 78666, TX, United States
Applied Science Research Center, Applied Science Private University, Amman, Jordan
Kashmir Institute of Economics
Institute of Energy Policy and Research
Erciyes University
Prague University of Economics and Business
Marwadi University Research Center
Department of Economics
UNITEN Business School
Department of Agricultural Sciences
Applied Science Research Center
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026