Table 1 Mean ratings of editors’ perceptions of the severity and frequency of various ethical issues at their journals, and their confidence in handling these issues

Issue	Severity†	Confidence‡	Frequency§	Trend¶
Redundant publication	1.19	0.70	1.39	3.43
Plagiarism	0.86	0.70	0.96	3.46
Duplicate submission	0.79	0.79	1.01	3.28
Undisclosed author conflicts of interest	0.73	0.73	0.90	3.28
Undisclosed reviewer conflicts of interest	0.69	0.71	0.94	3.08**
Gift authorship	0.67	0.51	1.08*	3.17
Disputed authorship	0.58*	0.90*	0.81	3.00**
Falsified or fabricated data	0.56	0.62	0.58	3.12**
Reviewer misconduct	0.56*	0.62	0.80	3.00**
Unethical research design or conduct	0.55	0.83*	0.70	2.98
Undisclosed commercial involvement	0.52	0.66	0.62	3.24
Ghost authorship	0.37	0.61	0.48	3.32
Image manipulation	0.30	0.69	0.47	3.34
Concerns over supplements	0.24**	1.08*	0.30**	2.97**
Concerns over advertising	0.13**	1.01*	0.20**	3.00**
Editorial interference by journal owner	0.05	1.51*	0.09	2.95**

Respondents who stated that they did not know were excluded from calculations of the mean. Editors were asked to base their replies on their perceptions of the situation at their own journal rather than their perceptions of problems in the literature in general.
†Severity was graded on a 4-point scale ranging from 0, “not a problem”, to 3, “a very serious problem”.
‡Editors’ confidence in handling issues was graded on a 4-point scale ranging from 0, “not at all confident”, to 3, “highly confident”.
§Frequency was graded on a 4-point scale ranging from 0, “never”, to 3, “very often (at least once a month)”.
¶Editors who stated that an issue occurred “sometimes (more than once a year)” or “very often (at least once a month)” were asked whether they thought the problem was decreasing a lot (1), decreasing slightly (2), occurring to the same extent as before (3), increasing slightly (4) or increasing a lot (5).
*Kurtosis less than −1.0, indicating a wide spread of opinions (curve flatter than normal distribution). **Kurtosis more than +3.0, indicating a high degree of consensus (curve sharper than normal distribution). Kurtosis of 0 indicates that responses followed a normal distribution.