Statistical Validation of Citation Distribution Models
Author Information
Author(s): Michael J Stringer, Marta Sales-Pardo, Luís A Nunes Amaral
Primary Institution: Northwestern University
Hypothesis
Is the citation distribution of scientific papers consistent with a discrete lognormal distribution?
Conclusion
The study concludes that the discrete lognormal distribution is a globally accurate model for the eventual impact of scientific papers published in single-discipline journals.
Supporting Evidence
- 30 out of 2184 journals showed citation distributions inconsistent with a discrete lognormal distribution.
- Large, multidisciplinary journals were over-represented among those inconsistent journals.
- The study suggests that citation distributions are lognormal within a discipline.
Takeaway
This study looks at how often scientific papers get cited and finds that most papers follow a predictable pattern, like a lognormal distribution.
Methodology
The study performed a large-scale empirical analysis of citation data from the Web of Science database, focusing on journals with a steady-state citation distribution.
Potential Biases
Potential biases may arise from the exclusion of certain journals and the reliance on citation data that can change over time.
Limitations
The study may not account for all types of journals, particularly those with fewer than 50 articles per year or those that are not primary research literature.
Participant Demographics
The analysis included journals from various scientific fields represented in the Web of Science database.
Statistical Information
P-Value
p<0.001
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website