Limits...
Open evaluation: a vision for entirely transparent post-publication peer review and rating for science.

Kriegeskorte N - Front Comput Neurosci (2012)

Bottom Line: Complex PEFs will use advanced statistical techniques to infer the quality of a paper.The continual refinement of PEFs in response to attempts by individuals to influence evaluations in their own favor will make the system ungameable.OA and OE together have the power to revolutionize scientific publishing and usher in a new culture of transparency, constructive criticism, and collaboration.

View Article: PubMed Central - PubMed

Affiliation: Medical Research Council, Cognition and Brain Sciences Unit Cambridge, UK.

ABSTRACT
The two major functions of a scientific publishing system are to provide access to and evaluation of scientific papers. While open access (OA) is becoming a reality, open evaluation (OE), the other side of the coin, has received less attention. Evaluation steers the attention of the scientific community and thus the very course of science. It also influences the use of scientific findings in public policy. The current system of scientific publishing provides only journal prestige as an indication of the quality of new papers and relies on a non-transparent and noisy pre-publication peer-review process, which delays publication by many months on average. Here I propose an OE system, in which papers are evaluated post-publication in an ongoing fashion by means of open peer review and rating. Through signed ratings and reviews, scientists steer the attention of their field and build their reputation. Reviewers are motivated to be objective, because low-quality or self-serving signed evaluations will negatively impact their reputation. A core feature of this proposal is a division of powers between the accumulation of evaluative evidence and the analysis of this evidence by paper evaluation functions (PEFs). PEFs can be freely defined by individuals or groups (e.g., scientific societies) and provide a plurality of perspectives on the scientific literature. Simple PEFs will use averages of ratings, weighting reviewers (e.g., by H-index), and rating scales (e.g., by relevance to a decision process) in different ways. Complex PEFs will use advanced statistical techniques to infer the quality of a paper. Papers with initially promising ratings will be more deeply evaluated. The continual refinement of PEFs in response to attempts by individuals to influence evaluations in their own favor will make the system ungameable. OA and OE together have the power to revolutionize scientific publishing and usher in a new culture of transparency, constructive criticism, and collaboration.

No MeSH data available.


Related in: MedlinePlus

A plurality of paper evaluation functions (PEFs) provides multiple lenses onto the literature. Organizations and individuals can define PEFs according to their own priorities and make the resulting paper rankings publicly available. Competing PEFs provide multiple perspectives. Moreover, the OE system becomes “ungameable” as PEFs respond to any attempts by individual scientists or groups to take advantage of weaknesses of current PEFs. With constantly evolving PEFs, each scientist and organization is motivated to aim for truth and objectivity. Red and blue pointers correspond to “excitatory” and “inhibitory” evaluative links, which could be represented by positive and negative numerical ratings. Beyond simple averaging of ratings, PEFs could employ sophisticated inference algorithms to jointly estimate the probabilities of all papers’ title claims.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3473231&req=5

Figure 5: A plurality of paper evaluation functions (PEFs) provides multiple lenses onto the literature. Organizations and individuals can define PEFs according to their own priorities and make the resulting paper rankings publicly available. Competing PEFs provide multiple perspectives. Moreover, the OE system becomes “ungameable” as PEFs respond to any attempts by individual scientists or groups to take advantage of weaknesses of current PEFs. With constantly evolving PEFs, each scientist and organization is motivated to aim for truth and objectivity. Red and blue pointers correspond to “excitatory” and “inhibitory” evaluative links, which could be represented by positive and negative numerical ratings. Beyond simple averaging of ratings, PEFs could employ sophisticated inference algorithms to jointly estimate the probabilities of all papers’ title claims.

Mentions: The necessary selection of papers for reading can be based on the reviews and their associated numerical judgments. Any group or individual can define a PEF based on content and quality criteria. A PEF could for example, rely only on signed ratings from post-PhD scientists, and weight different rating scales in a particular way. A PEF could also employ social-network information, e.g., to downweight ratings from reviewers that are associated with the authors. Social networks could also contribute evaluative information on papers to PEFs, including usage and sharing statistics as well as ratings (Lee, 2012; Priem and Hemminger, 2012; Walther and van den Bosch, 2012; Zimmermann et al., 2012; all in this collection). Beyond weighted averaging, PEFs could use complex recurrent inference algorithms, e.g., to infer probabilities for the title claims of papers. Social web and collaborative filtering algorithms (Goldberg, 1992; Breese et al., 1998; Schafer et al., 2007) will be applied to this problem. However, evaluating the scientific literature poses unique challenges and requires greater transparency and justification than product recommendation systems. The development of PEFs will build on and extend the existing literature on collaborative filtering systems. There will be a plurality of PEFs prioritizing the literature from multiple perspectives (Figure 5). When reviewers start using a new rating scale in their evaluations, PEFs may utilize the ratings on the new scale if the evaluative evidence the scale provides is thought to justify its inclusion.


Open evaluation: a vision for entirely transparent post-publication peer review and rating for science.

Kriegeskorte N - Front Comput Neurosci (2012)

A plurality of paper evaluation functions (PEFs) provides multiple lenses onto the literature. Organizations and individuals can define PEFs according to their own priorities and make the resulting paper rankings publicly available. Competing PEFs provide multiple perspectives. Moreover, the OE system becomes “ungameable” as PEFs respond to any attempts by individual scientists or groups to take advantage of weaknesses of current PEFs. With constantly evolving PEFs, each scientist and organization is motivated to aim for truth and objectivity. Red and blue pointers correspond to “excitatory” and “inhibitory” evaluative links, which could be represented by positive and negative numerical ratings. Beyond simple averaging of ratings, PEFs could employ sophisticated inference algorithms to jointly estimate the probabilities of all papers’ title claims.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3473231&req=5

Figure 5: A plurality of paper evaluation functions (PEFs) provides multiple lenses onto the literature. Organizations and individuals can define PEFs according to their own priorities and make the resulting paper rankings publicly available. Competing PEFs provide multiple perspectives. Moreover, the OE system becomes “ungameable” as PEFs respond to any attempts by individual scientists or groups to take advantage of weaknesses of current PEFs. With constantly evolving PEFs, each scientist and organization is motivated to aim for truth and objectivity. Red and blue pointers correspond to “excitatory” and “inhibitory” evaluative links, which could be represented by positive and negative numerical ratings. Beyond simple averaging of ratings, PEFs could employ sophisticated inference algorithms to jointly estimate the probabilities of all papers’ title claims.
Mentions: The necessary selection of papers for reading can be based on the reviews and their associated numerical judgments. Any group or individual can define a PEF based on content and quality criteria. A PEF could for example, rely only on signed ratings from post-PhD scientists, and weight different rating scales in a particular way. A PEF could also employ social-network information, e.g., to downweight ratings from reviewers that are associated with the authors. Social networks could also contribute evaluative information on papers to PEFs, including usage and sharing statistics as well as ratings (Lee, 2012; Priem and Hemminger, 2012; Walther and van den Bosch, 2012; Zimmermann et al., 2012; all in this collection). Beyond weighted averaging, PEFs could use complex recurrent inference algorithms, e.g., to infer probabilities for the title claims of papers. Social web and collaborative filtering algorithms (Goldberg, 1992; Breese et al., 1998; Schafer et al., 2007) will be applied to this problem. However, evaluating the scientific literature poses unique challenges and requires greater transparency and justification than product recommendation systems. The development of PEFs will build on and extend the existing literature on collaborative filtering systems. There will be a plurality of PEFs prioritizing the literature from multiple perspectives (Figure 5). When reviewers start using a new rating scale in their evaluations, PEFs may utilize the ratings on the new scale if the evaluative evidence the scale provides is thought to justify its inclusion.

Bottom Line: Complex PEFs will use advanced statistical techniques to infer the quality of a paper.The continual refinement of PEFs in response to attempts by individuals to influence evaluations in their own favor will make the system ungameable.OA and OE together have the power to revolutionize scientific publishing and usher in a new culture of transparency, constructive criticism, and collaboration.

View Article: PubMed Central - PubMed

Affiliation: Medical Research Council, Cognition and Brain Sciences Unit Cambridge, UK.

ABSTRACT
The two major functions of a scientific publishing system are to provide access to and evaluation of scientific papers. While open access (OA) is becoming a reality, open evaluation (OE), the other side of the coin, has received less attention. Evaluation steers the attention of the scientific community and thus the very course of science. It also influences the use of scientific findings in public policy. The current system of scientific publishing provides only journal prestige as an indication of the quality of new papers and relies on a non-transparent and noisy pre-publication peer-review process, which delays publication by many months on average. Here I propose an OE system, in which papers are evaluated post-publication in an ongoing fashion by means of open peer review and rating. Through signed ratings and reviews, scientists steer the attention of their field and build their reputation. Reviewers are motivated to be objective, because low-quality or self-serving signed evaluations will negatively impact their reputation. A core feature of this proposal is a division of powers between the accumulation of evaluative evidence and the analysis of this evidence by paper evaluation functions (PEFs). PEFs can be freely defined by individuals or groups (e.g., scientific societies) and provide a plurality of perspectives on the scientific literature. Simple PEFs will use averages of ratings, weighting reviewers (e.g., by H-index), and rating scales (e.g., by relevance to a decision process) in different ways. Complex PEFs will use advanced statistical techniques to infer the quality of a paper. Papers with initially promising ratings will be more deeply evaluated. The continual refinement of PEFs in response to attempts by individuals to influence evaluations in their own favor will make the system ungameable. OA and OE together have the power to revolutionize scientific publishing and usher in a new culture of transparency, constructive criticism, and collaboration.

No MeSH data available.


Related in: MedlinePlus