Expression of Concern: Exploring Regional Variation in Roost Selection by Bats: Evidence from a Meta-Analysis
Year: 2024
Type: article
Abstract: Following publication of this article [1], concerns were raised about reporting of the methodology and discrepancies in the datasets.After follow up with the corresponding author and consultation with an expert member of the Editorial Board and a statistical reviewer, the PLOS ONE Editors received input that the issues do not affect the main conclusions; however, some items could not be fully resolved. Dataset errorsThe corresponding author stated that the originally published datasets used in the meta-analyses contained errors which arose when a cleaned dataset was accidentally overwritten with an older version.They noted that reanalysis of the corrected data produced minor changes to the results, with the main conclusions remaining supported.After reanalysis, one additional factor reaches statistical significance (distance to water).The revised meta-analyses are provided here in S1 File, along with detailed descriptions of the corrections made to the datasets in S2 File.Corrected versions of Table 1, Fig 3 and Fig 4 are also provided in S1 File.A consulted member of the Editorial Board and statistical reviewer advised that the revised results are consistent with the original conclusions and the numerical changes to the results are minor.However, one discrepancy was noted that was not addressed in the revised dataset.Specifically:• In the corrected meta-analysis for tree diameter (S1 File), sample sizes for the Perry and Thill ( 2007) study are reported as 40 and 42, while the source article reports sample sizes of 43 and 49.The statistical reviewer noted that this remaining discrepancy is unlikely to affect the results.Readers are advised there appear to be discrepancies in some descriptions of the revisions in S2 File, as follows:• For the analysis of Canopy Closure, the corrected mean value from Lacki et al. 2009 used in the revised analyses (S1 File) is 47% (revised from 45%), but these values are reversed in the description in S5_Table_correction report (S2 File).• For the analysis of Canopy Closure, the revised Table 1 in S1 File reports an updated P value of P < 0.0001 for the random effects model, while the S5_Table_correction report (S2 File) reports a revised P value of P = 0.0027 for the random effects model and P <0.0001 for the fixed effect model.• For the analysis of Tree Density, the mean (SD) values from the Vonhof and Gwilliam studies used in the revised analyses (S1 File and S7_Table_correction report in S2 File) are 38 (28.4), and 38 (30.9) for big brown bats (EPFU) and California bat (MYCA), respectively,
Source: PLoS ONE
Author: The PLOS ONE Editors
Language: English
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,290
OpenAlex ID: W4405523706