Trackademia

Welcome to Trackademia

The idea behind Trackademia is to build on existing academic databases like Google Scholar and OpenAlex so that you can trust the data and use it for whatever you want. That comes down to two things.

More data, and more ways to use it. We aggregate publication data from every major source into a single dataset and add information pulled from faculty CVs. Then we let you slice it however you want: choose which sources to build from, write your own formulas, weight journals, adjust for department size.

More accurate, and honest about where it isn’t. We focus on a smaller set of schools so we can verify everything carefully. Faculty lists come from department websites rather than being inferred from publication records, so we know who should be in the data before we go looking. When we can’t find data on someone, we record the gap instead of dropping them. Every number comes with a coverage rate you can check against ground truth.

Everything in Trackademia is built from seven units. AI merges duplicate records across sources and tags each publication so these categories hold up:

  • Publication — citation counts, authors, journal, peer review status, and more
  • Publication group — versions of the same work combined into one entity, such as a working paper and the article it became
  • Academic
  • Department
  • University
  • Field
  • Journal

There are two modes. In Database mode you browse the underlying records and filter by any of the seven units. In Analytics mode you get graphs and comparisons, and you switch views to compare two departments, compare one department against the average of the others, or see them all at once. Tabs let you focus on a particular level of the data.

Trackademia

Every graph on this site

Publicationshere

  • Ranking Ranked table
    • Value: Citations, Weighted citations, Publications
    • Measure: Total, Mean per work, Median per work
  • Works published per year Lines over years
    • Measure: Total, Mean per academic, Median per academic
    • Calendar year or years since PhD
    • In each year or cumulative
  • Citations received per year Lines over years
    • Value: Citations, Weighted citations
    • Measure: Total, Mean per work, Median per work, Mean per academic, Median per academic
    • In each year or cumulative
  • Citation distribution Bars
    • Value: Citations, Weighted citations
  • Citations by age of work Lines over years
    • Value: Citations, Weighted citations
    • Measure: Total, Mean per work, Median per work
  • Publication type mix Ranked bars
    • Value: Citations, Weighted citations, Publications
  • Top venues Ranked bars
    • Value: Citations, Weighted citations, Publications

Academics

Departments

  • Ranking Ranked table
    • Value: Citations, Weighted citations, Publications
    • Measure: Total, Mean per work, Median per work, Mean per academic, Median per academic, Formula
  • Publications over time Lines over years
    • Measure: Total, Mean per academic, Median per academic, Formula
    • Calendar year or years since PhD
    • In each year or cumulative
  • Citations over time Lines over years
    • Value: Citations, Weighted citations
    • Measure: Total, Mean per academic, Median per academic, Formula
    • In each year or cumulative
  • Makeup by tenure track Stacked bars
  • Output by rank Grouped bars
    • Value: Citations, Weighted citations, Publications
    • Measure: Total, Mean per work, Median per work, Mean per academic, Median per academic, Formula
  • Academics ranked against academics Grouped bars
    • Value: Citations, Weighted citations, Publications
    • How many ranks (opens on 5)
  • Years to promotion Grouped bars
  • Journals used Stacked bars
    • Value: Citations, Weighted citations, Publications

Journals

Fields

  • Ranking Ranked table
    • Value: Citations, Weighted citations, Publications
    • Measure: Total, Mean per work, Median per work, Mean per academic, Median per academic
  • Departments by field Bars
    • Measure: Total, Mean per work, Median per work, Mean per academic, Median per academic
  • University rollup Ranked bars
  • Coverage by field Ranked bars

Trackademia

Search the whole site

Search a person, a department, a journal or the title of a work. A person’s name brings up their profile, their publications, their department and the journals they publish in.
Trackademia
Value Measure

Publications

Citations · the 15 highest of 2,564 works
1 Stan: A probabilistic programming language 2017 · Journal of Statistical Software · Andrew Gelman 10,846
2 Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC 2017 · Statistics and Computing · Andrew Gelman 7,838
3 The Science of Fake News 2018 · Science · Brendan Nyhan 7,397
4 The Origins and Consequences of Affective Polarization in the United States 2019 · Annual Review of Political Science · Matthew Levendusky 5,063
5 Redefine Statistical Significance 2018 · Nature Human Behaviour · Brendan Nyhan, Donald P. Green, Dustin Tingley 3,687
6 Rank-normalization, folding, and localization: An improved R-hat for assessing convergence of MCMC 2021 · Bayesian Analysis · Andrew Gelman 3,065
7 stm: R package for Structural Topic Models 2019 · Journal of Statistical Software · Dustin Tingley 2,865
8 Why Propensity Scores Should Not Be Used for Matching 2019 · Political Analysis · Gary King 2,813
9 Why high-order polynomials should not be used in regression discontinuity designs 2019 · Journal of Business & Economic Statistics · Andrew Gelman 2,692
10 Exposure to Untrustworthy Websites in the 2016 U.S. Election 2020 · Nature Human Behaviour · Brendan Nyhan 2,369
11 Status threat, not economic hardship, explains the 2016 presidential vote. 2018 · Proceedings of the National Academy of Sciences · Diana C. Mutz 1,701
12 How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, not Engaged Argument 2017 · American Political Science Review · Gary King 1,652
13 Bayesian statistics and modelling 2021 · Nature Reviews Methods Primers · Andrew Gelman 1,628
14 The Nature and Origins of Misperceptions: Understanding False and Unsupported Beliefs about Politics 2017 · Political Psychology · Brendan Nyhan 1,608
15 Visualization in Bayesian workflow (with discussion and rejoinder) 2019 · Journal of the Royal Statistical Society Series A (Statistics in Society) · Andrew Gelman 1,519

174,526 citations over 1,664 of 2,564 works (65% carry a count). 95 of 100 academics have at least one counted work. A work with no citation count is unknown here, never a zero.

Works published per year

Publications in each year
Measure X axis Each point

Every selected work by the year it was published, each work once however many of these departments hold it. 92 carry no year and are on no year.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.

Citations received per year

Citations in each year
Value Measure Each point

The works these filters select, whatever year they were published; the year window picks the years the citations arrived in. 3,845 of 8,221 works carry a year-by-year record inside it.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.
Notes

A year-by-year record of citations to one work is OpenAlex’s. A Scholar profile counts everything one person wrote and is not about a work, so no Scholar figure is in this chart; the counts beside it still follow the one citation rule.

Citation distribution

How the citations fall across the works
Value
Series: Works.02004006008000-1 Works 1841842-10 Works 43043011-100 Works 755755100+ Works 2952950-12-1011-100100+WorksCitations

The 1,664 works that carry a count, by how many citations they have; the last bucket is everything above a hundred. The median is 21. 900 works carry no count and are in no bucket.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.

Citations by age of work

Citations by years since publication
Value Measure

The total citations of the works published that many years ago, over the works that carry a count. A work with no year is on no point, and a point at the far end is over few works: 59 at the thinnest age.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.

Publication type mix

Citations by publication type
Value
Series: value.ArticlesArticles 174,526174,526050,000100,000150,000Publication typeCitations

The 1,664 works that carry a count, by their publication type. A type none of whose works carries one is on no bar rather than on a zero. The Publication type filter decides which are here at all.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.

Top venues

Where these works earned their citations
Value
Series: value.ScienceScience 14,11314,113Journal of Statistical SoftwareJournal of Statistical Software 13,71113,711Annual Review of Political ScienceAnnual Review of Political Science 8,5368,536Statistics and ComputingStatistics and Computing 7,8467,846Nature Human BehaviourNature Human Behaviour 7,4047,404The Journal of PoliticsThe Journal of Politics 7,0367,036Proceedings of the National Academy of SciencesProceedings of the National Academy of Sciences 6,9496,949American Journal of Political ScienceAmerican Journal of Political Science 6,9416,941Political AnalysisPolitical Analysis 5,3755,375American Political Science ReviewAmerican Political Science Review 5,2485,248Bayesian AnalysisBayesian Analysis 4,3004,300International OrganizationInternational Organization 3,8963,896Comparative Political StudiesComparative Political Studies 2,7042,704Journal of Business and Economic StatisticsJournal of Business and Economic Statistics 2,6922,692Public Opinion QuarterlyPublic Opinion Quarterly 2,2602,26005,00010,000VenueCitations

The 15 venues whose works earned the most citations, of 540 holding a selected work; a venue none of whose works carries a count is on no bar rather than on a zero. 295 works are in no journal record.

What these filters select, over everything selected not what the graph divided by
Selected Academics CVScholarOpenAlex Works With a count
Everything selected 100 100/100 1.0% 75/100 0.75% 98/100 0.98% 2564 1664/2564 0.6489859594383776%
  • Over the academics these filters select, so a filtered page describes the academics it selected and not the whole department.
  • A row with nothing held draws no mark above: the graph is quiet there because the evidence is missing, not because the work is.
  • What there was to draw from, not what the graph divided by - the graph states its own denominator above.
  • A Scholar profile is attached when at least 3 of the CV titles are found on it, or 30% of them are; one at another institution or under a different name needs 10 titles, and a profile far larger than the CV that holds only a small part of it is refused. Its articles are then checked one by one: what is not a work, or not by this person, is not counted.