Three dimensions of scientific impact

Area of Science:

• Bibliometrics
• Scientometrics
• Scholarly communication analysis

Background:

• Bibliometric indexes, such as the h index, are increasingly popular for evaluating scientific impact.
• Criticism exists that scientific impact cannot be accurately represented by a single numerical value.
• Some argue that complex scientific realities resist quantitative descriptions.

Purpose of the Study:

• To challenge the extremes of scientific impact evaluation, either oversimplification or complete rejection of quantification.
• To develop a nuanced model for summarizing citation records.
• To provide a more interpretable framework for understanding research impact.

Main Methods:

• Developing a model based on citation distribution.
• Incorporating the 'rich get richer' (preferential attachment) principle for some citations.
• Including random citation assignment to account for general referencing.

Main Results:

• A novel model accurately summarizes citation records.
• The model utilizes three interpretable parameters: productivity, total impact, and luck.
• This approach offers a more comprehensive view than single-number metrics.

Conclusions:

• Scientific impact can be quantitatively described using a multi-parameter model.
• The proposed model balances preferential attachment and random citation processes.
• This framework offers a more accurate and interpretable summary of research impact.