Skip to main content

Table 3 Frequency of reporting for individual questions. Only the questions with the largest differences in each comparison are shown; complete data is available in Table S6 (for the independent samples) and Table S7 (for the paired samples). As none of these items had articles scoring “Partially”, only the number and percentage of “Yes” answers are shown. Results are from Fisher’s exact tests (in the independent samples comparisons) and McNemar’s exact tests (in the paired comparisons). 95% C.I.; 95% confidence interval

From: Comparing quality of reporting between preprints and peer-reviewed articles in the biomedical literature

  Preprint Peer-reviewed Odds ratio [95% C.I.] p value
Independent samples
 Unit-level data 22 3 0.1 [0.02, 0.4] 4.5 × 10−5
29% 4.2%
 Animal source/supplier (Vertebrate) 12 22 7.6 [1.6, 49.9] 0.005
48% 88%
 Eligibility criteria (Human) 19 29 6.4 [1.5, 39.8] 0.008
59.4% 90.6%
 Randomization (Vertebrate) 0 8 ∞ [2.0, ∞] 0.003
0% 47%
 Symbol meaning 27 45 4.9 [1.3, 23.0] 0.01
69.2% 91.8%
 Variation/precision measure 49 60 2.8 [1.2, 6.9] 0.01
66.2% 84.5%
 Conflict of interest statement 34 50 2.4 [1.2, 4.8] 0.01
44.7% 65.8%
Paired sample
 Conflict of interest statement 26 47 11.5 [2.8, 100.6] 1.9 × 10−5
46.4% 83.9%
 Funding source 50 55 ∞ [0.9, ] 0.06
89.3% 98.2%
 Unit-level data 14 19 ∞ [0.9, ∞] 0.06
25% 33.9%