Percentile-Based Citation Impact (PPtop10%)
Also known as: Percentile Rank Citation Indicators, Top 10% Highly Cited Papers Indicator, PPtop10%, Integrated Impact Indicator (I3)
Percentile-based citation impact replaces the average citation count with a paper's rank position within a properly defined reference set. Instead of asking how many citations a paper received, it asks where the paper falls in the citation distribution of comparable papers from the same field, year, and document type. Because citation distributions are extremely skewed, a single highly cited paper can inflate a mean, so Lutz Bornmann and Loet Leydesdorff argued that impact should be measured non-parametrically through percentile ranks and the share of papers reaching the top of their field. The most widely used summary is PPtop10%, the proportion of a unit's papers that belong to the most-cited 10% of their reference set; Leydesdorff and Bornmann's Integrated Impact Indicator (I3) generalizes this idea by integrating the full percentile curve. Ludo Waltman and Michael Schreiber clarified how percentile ranks should be computed when many papers share the same citation count.
Key highlights
- Robust to the extreme skew of citation distributions because it uses rank position rather than the citation mean.
- Field-normalized by construction through reference sets defined by field, year, and document type.
- Yields an interpretable scale where the world average is fixed (10% of papers in the top 10%), enabling cross-field comparison.
- The Integrated Impact Indicator unifies size and quality in a single non-parametric statistic that generalizes the top-share family.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use percentile-based citation impact when you need a field-normalized measure of research performance that is robust to the extreme skew of citation distributions and to a few outlier papers. It is the indicator of choice for comparing universities, departments, countries, or journals across disciplines, where mean-based measures are distorted by differing citation densities and by blockbuster papers. PPtop10% and I3 are appropriate when you have well-defined reference sets by field, year, and document type, and when the population of papers is large enough that decile membership is meaningful. The approach is less useful for very small sets where a single paper can swing the top-share, for very recent papers whose citations have not accumulated, and in fields with poor database coverage where reference sets are unreliable. It complements rather than replaces journal-level indicators, since it operates on the paper distribution directly.
Strengths & limitations
- Robust to the extreme skew of citation distributions because it uses rank position rather than the citation mean.
- Field-normalized by construction through reference sets defined by field, year, and document type.
- Yields an interpretable scale where the world average is fixed (10% of papers in the top 10%), enabling cross-field comparison.
- The Integrated Impact Indicator unifies size and quality in a single non-parametric statistic that generalizes the top-share family.
- Results depend heavily on how the reference set is delineated, and field classification schemes are imperfect and contested.
- Percentile ranks are unstable for very small units where one paper can shift the top-share substantially.
- Handling of tied citation counts is non-trivial and different conventions can produce different indicator values.
- Like all citation measures it lags publication, so recent output is undercounted until citations accumulate.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
Why use percentiles instead of average citations?
Citation distributions are extremely skewed, so a single highly cited paper can dominate a mean and make averaging statistically misleading. Percentile ranks score each paper by its position within a comparable reference set, which is robust to outliers and naturally field-normalized. Bornmann and Leydesdorff argued that for skewed data, non-parametric rank-based statistics are more appropriate than central-tendency measures, which is why PPtop10% and the Integrated Impact Indicator have largely displaced mean-based normalization in serious research evaluation.
What does PPtop10% of 0.10 mean?
It means the unit performs exactly at the world average. Because the top 10% is defined to contain 10% of the reference set, a unit whose papers are randomly distributed would expect 10% of them in the top decile. A PPtop10% above 0.10 indicates more high-impact papers than expected, and below 0.10 indicates fewer. Expressing the result as a ratio to the expected 0.10 gives a scale-free score where one is average, which is what makes the indicator comparable across fields of very different citation density.
How are tied citation counts handled?
Ties are common at low citation counts, where many papers have the same small number of citations, and they require care. Waltman and Schreiber showed that naive ranking rules can produce percentile ranks that do not average to the expected value across the reference set, biasing the aggregate indicators. Their recommended approach distributes tied papers fractionally across the rank positions they jointly occupy, so the percentile ranks remain consistent and the top-share and integrated indicators built on them stay unbiased.
Sources
- 1.Leydesdorff, L., & Bornmann, L. (2011). Integrated impact indicators compared with impact factors: An alternative research design with policy implications. Journal of the American Society for Information Science and Technology, 62(11), 2133-2146.
- 2.Waltman, L., & Schreiber, M. (2013). On the calculation of percentile-based bibliometric indicators. Journal of the American Society for Information Science and Technology, 64(2), 372-379.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Percentile-Based Citation Impact (PPtop10%). ScholarGate. https://scholargate.app/bibliometrics/percentile-based-citation-impact