Table 1.
Sets of variables analyzed in this article.
Some identifiers are shared across these sets of variables and noted when appropriate.
Table 2.
Breakdown of reviews and manuscripts per type of reviewer.
Percentages are calculated as the fraction of total number of review suggestions/review invitations/reviews, and manuscripts where such information appears per type of reviewer. For example, the first cell (i.e., 5.8% / 23.9%) means author-opposed reviewers were suggested in 5.8% of all review suggestions involving 23.9% of all manuscripts. Similarly, the second cell (i.e., 0.2% / 1.0%) means that author-opposed reviewers were invited to 0.2% of all invitations involving 1.0% of all manuscripts.
Table 3.
Statistics per manuscript: Number of reviewers suggested by the author, invited, and who provided reviews across revisions.
IQR = Interquartile Range.
Fig 1.
Author-suggested reviewers are less likely to accept a review invitation than editor-suggested reviewers.
The point is the mean and the error bars are the standard errors of the mean (SEM). For comparison across means, the asterisk indicates the statistical significance of a two-sample t test where "***", "**", and "*" represent p < 0.01, p < 0.05 and p < 0.1, respectively.
Fig 2.
Author-suggested reviewers are more likely to give favorable scores.
The point is the mean and the error bars are the standard errors of the mean (SEM). For comparison across means, the asterisk indicates the statistical significance of a two-sample t test where "***", "**", and "*" represent p < 0.01, p < 0.05 and p < 0.1, respectively.
Fig 3.
An all-author-suggested panel of reviewers produces a large increase in the acceptance rate of a manuscript.
The point is the mean and the error bars are the standard errors of the mean (SEM). For comparison across means, the asterisk indicates the statistical significance of a two-sample t test where "***", "**", and "*" represent p < 0.01, p < 0.05 and p < 0.1, respectively.
Fig 4.
Author-suggested reviewers produce reviewers that are judged as lower quality by editors.
The point is the mean and the error bars are the standard errors of the mean (SEM). For comparison across means, the asterisk indicates the statistical significance of a two-sample t test where "***", "**", and "*" represent p < 0.01, p < 0.05 and p < 0.1, respectively.