The Spread of Scientific Information: Insights from Web Usage Statistics
Author Information
Author(s): Yan Koon-Kiu, Mark Gerstein
Primary Institution: Yale University
Hypothesis
How is the awareness of a scientific paper spread within communities of scientists?
Conclusion
The spread of scientific information follows a two-phase decay pattern, with a rapid decline in the first month after publication followed by a gradual power law decay.
Supporting Evidence
- The cumulative number of HTML views follows a lognormal distribution.
- The decay of information spread is characterized by two distinct phases.
- The patterns of information spread are similar across different journals.
- The number of HTML views is strongly correlated with PDF downloads.
- The rapid decay in views is likely influenced by the initial fame of a paper.
Takeaway
When scientists read a paper, they often share it with others, but the excitement fades quickly after the first month, and fewer people pay attention over time.
Methodology
The study analyzed web usage statistics from the PLoS Article-Level Metrics dataset, focusing on HTML views, PDF downloads, and XML downloads of published papers.
Potential Biases
The analysis may not fully capture the dynamics of publications that receive extremely high attention.
Limitations
The model does not account for inhomogeneities between different research fields or the structure of sub-communities among scientists.
Participant Demographics
The study focused on a community of scientists reading papers from six PLoS journals.
Statistical Information
P-Value
0
Statistical Significance
p=0
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website