How the evidence is measured.
SPAR — Scholarly Performance Assessment Review — evaluates publication productivity and citation impact across a predefined business and economics journal corpus. Every metric is conditional on the journals, years, researchers and reference population selected by the report author.
Researcher eligibility and publications
The SPAR researcher population comprises authors with at least one substantive journal article in the SPAR journal list. Publications were collected through OpenAlex and stored in SPAR's own database. Inclusion is based on qualifying source and publication type, not institutional employment or researcher nationality.
Of the 468 entries in the AJG / ABS 2024 list, 467 are journals and all are covered. The remaining entry is a book series and is not included.
Journal selection and classification
Choose one primary list: the AJG/ABS source catalog, UTD24, or FT50. Within that list, optionally narrow by AJG rating, economics/CNRS rating, research field, specifically included journals, or excluded journals.
UTD24 and FT50 are classification tags on existing catalog journals, not separate publication collections. Memberships may overlap; selecting or filtering them never creates additional copies of a publication. The UTD24 and FT50 classifications come from the SPAR journal list, with the FT50 membership updated against the April 2026 revision.
How metrics are calculated
| Measure | Exact meaning for the selected journal/year window |
|---|---|
| Publication count | Number of qualifying unique journal articles associated with the scholar. |
| Fractional publications | For each article, credit equals 1 divided by its recorded total author count. The metric sums those credits across the scholar's qualifying articles. |
| Total citations | Sum of each qualifying article's OpenAlex citation count at the latest data update. This is not lifetime citations across all sources or an annual citation history. |
| h-index | Largest integer h such that at least h qualifying articles each have h or more citations. Recomputed for the selected publication window and journal subset. |
| i10-index | Number of qualifying articles with at least 10 citations. |
| Citations per paper | Total citations divided by publication count, or zero when no qualifying publications exist. |
| Age-adjusted citations | Sum over qualifying articles of citations divided by max(1, current calendar year minus publication year plus one). This is a simple publication-age adjustment, not field-normalized impact. |
| Distinct journals | Number of distinct qualifying journals represented by the scholar's or school's articles. |
| Active publication years | Number of distinct publication years containing qualifying articles. The profile also reports it as a share of the window (for 2010–2026, 12 active years is 12 of 17, or 70.6%). |
| Publication momentum (3 years) | Qualifying articles in the latest three complete calendar years in scope, minus the three complete years before them. Unavailable when the window is shorter than six complete years. It measures change in output, not citation velocity. |
| Unique co-authors | Distinct co-authors on the scholar's qualifying articles who are indexed SPAR scholars. Co-authors outside the SPAR population are not identified and are not counted. |
| Journal quality breakdown | Qualifying articles per AJG rating, with shares, when the selection spans several ratings. When the selection is a single rating or a list such as FT50 or UTD24, only the total is shown. |
Starting point, benchmark and percentiles
Every assessment has a starting point (one scholar, one school, or the schools of a geographic area taken together) and, optionally, one benchmark. Without a benchmark the report describes the starting point. With a benchmark, every figure is read against that benchmark and the individual profile is not repeated. Starting point and benchmark always use the same publication years and journal selection.
Benchmark populations
- Named scholars or schools: exactly the entities you pick.
- Geographic area: continents, country groups, countries, regions or a radius around a point. For scholars, the benchmark is every SPAR scholar whose latest affiliation is at a school in the area.
- Worldwide: every eligible SPAR scholar, or every school, regardless of location.
A school enters a benchmark only with at least one qualifying paper in the selected years and journals. For scholars, every scholar at the chosen schools or in the chosen area counts by default, including those with no qualifying paper, and percentiles are computed over all of them. The option Only scholars with at least one eligible publication removes those scholars before the band and every statistic. With a narrow journal list (FT50, say) most scholars may have none, so the band can keep nearly everyone at zero; the live counter and the report give how many such scholars there are and the band’s lower cutoff. Named scholars and schools always stay.
For geographic and worldwide benchmarks you can keep a percentile band of the population by one measure, for example the 95th to 100th percentile by h-index. Members outside the band are removed from the benchmark: from its size, its statistics, the takeaways and the member list. Values tied at either boundary are kept, so the band can hold slightly more than its nominal share. Named additions always stay. The live counter shows the benchmark size before the report is generated.
Statistics
The benchmark distribution reports the mean and the median first, then the sample standard deviation (n − 1 denominator) and the 1st, 5th, 25th, 75th, 95th and 99th percentiles, computed by linear interpolation between order statistics. This summary table appears when the benchmark has more than ten members. With one comparator the report shows a direct comparison and differences only; with two to ten, the members are listed without distribution statistics.
The starting point's position is a midrank percentile: 100 × (benchmark values strictly below + one half of values equal) / benchmark size. Its rank on a measure is its place among the benchmark members plus itself. The starting point is excluded from benchmark statistics unless you include it explicitly.
Profiles without a benchmark
A profile still places its figures in context. A scholar is placed among SPAR scholars with at least one qualifying paper in the same years and journals, read through a reproducible random sample of up to 2,000 of them (all of them when fewer qualify); each tile shows the midrank percentile. A school is ranked among every school worldwide with qualifying output under the same filters, with its rank and percentile on each measure. Takeaways then summarize each dimension: output, impact, journal quality, trend (the three-year windows of publication momentum and normalized citation momentum), consistency, collaboration and concentration. A dimension that does not apply, or has no reliable reference, is left out rather than shown empty.
How concentrated a body of work is
For research fields, journals and co-authors, the profile reports the number of distinct entries and the effective number: 1 / Σ s², where s is each entry's share of fractional credit (fields, journals) or of joint papers (co-authors). It is the number of equally weighted entries that would give the same concentration. Ten journals with one dominant outlet may have an effective number close to two; ten journals used equally have an effective number of ten.
Effective numbers grow with output, so they are read against SPAR scholars with similar output: those with 0.75 to 1.33 times as many qualifying papers (at least three), under the same journal and year selection. A reproducible random sample of up to 1,000 such scholars gives the typical range (25th to 75th percentile). Below the range reads as Concentrated, above it as Broad, inside it as Typical.
For schools, the effective number of contributing scholars, fields and journals is reported without a reading range.
Publications and citations by year
The yearly chart stacks qualifying publications by journal rating. Two labelled rows under the year axis give the distinct co-authors and distinct journals of each year. The line is a centred three-year average (the year itself and one year either side); it is dashed where that window is incomplete, at the edges of the range or where it reaches the current year. The current year is included to date and hatched.
Citations by year come from OpenAlex's record of the citations each work received in each calendar year. The chart shows, for each year, the citations received in that year and the two before it (a rolling three-year total), so a single unusual year does not dominate; citations per paper divides that total by the qualifying papers published by that year. Years whose three-year window reaches before the years covered by these records are marked as not covered rather than shown as zero, and citations recorded before a scholar's first qualifying paper (for example to preprints) are left out.
Comparisons use the same charts for everyone in the report, on one shared scale: each starting point, the benchmark as a whole (its average member: every member's papers or citations in a year averaged, a member without a paper counting as zero) and, for a benchmark of ten or fewer, each member in its own colour, the same colour in every chart and table of the report. Area against area compares the average school of each side.
Researchers, affiliations and countries
Researchers are indexed using their OpenAlex author identifiers. Identity disambiguation is inherited from OpenAlex and may be imperfect; an author may have split profiles or homonyms. Country and institution filters use OpenAlex last-known institutional affiliations, including multiple institutions where reported, rather than nationality or verified current employment.
How school rankings are constructed
School rankings use researchers' most recently reported OpenAlex last-known institutions, restricted to institutions OpenAlex categorizes as educational. These are the latest available affiliations, not where articles were written and not independently confirmed current employment. Each school report gives the date of the affiliation data.
An eligible publication is counted once for a school when at least one author has that school among their last-known OpenAlex institutions. Two authors affiliated with the same school do not count the paper twice for that school; articles coauthored across schools may count once for each school. Citations follow this same unique-paper-per-school rule, so school totals are not additive worldwide.
For fractional publications, an author's 1/N author contribution is split equally between that author's last-known multiple institutions. This avoids multiplying that fractional credit across all their institutional affiliations. School rankings can be narrowed by country, research field, journal membership list, rating, and publication window. School benchmarks follow the same journal and year filters as the starting point.
Data and reproducibility
Publications, citations and affiliations come from OpenAlex, the open catalogue of scholarly work. SPAR covers the journals of the Chartered ABS Academic Journal Guide 2024 (AJG), with the UTD24 and FT50 lists marked within it. A researcher is included through at least one qualifying article in these journals; schools and countries follow from researchers' latest known affiliations.
The data is refreshed at the start of each month. Each report records its years, journals and benchmark, and the date of the data it uses. Publications and citations are point-in-time values: new articles, corrections, retractions and citation totals may change at later updates. The online report and the PDF use the same calculations.
Errors in an author profile, a paper or an affiliation are corrected at the source: report them to OpenAlex (fixing data errors). The correction appears in SPAR after the next monthly update.
The measures are descriptive: they show output and attention, not the originality or rigour of the work, and they are no substitute for reading it. Reports should not be used to make consequential academic or employment decisions without additional qualitative evidence, source verification, and awareness of bibliometric limitations.