Abstract
On June 18, 2025, FACE received its first impact factor, IF, which is a widely accepted quantitative measure of the overall quality of an indexed journal. Developed initially by Paul Gross and Edward Gross in 1927 1 to help small college libraries select “good” journals to subscribe for the education of their students since, then as now, resources are limited. While many have contributed to its modifications, Eugene Garfield and Irving Sher, from the Institute for Scientific Information and supported by the National Science Foundation and the National Institutes of Health, significantly refined and improved the statistical methods of measurement, bringing it close to the IF as we know it today. Using a mainframe computer, they collected and analyzed 1.4 million citations in 1961 to 1962. Their key findings, still interesting and relevant today, show average article length to be 5.4 pages with 2.1 authors. An issue typically had 17.3 articles with 13.7 references per article. Importantly, there was a power law distribution (L-shaped curve) between the proportion of articles and the frequency of citations: many articles received one citation; only 25% as many received two, and 6.3% received three. Each increase in citation reduces the proportion of articles by a quarter, articles with four citations were 1.5% as many as those cited only once. 2 Power law distributions abound in both natural phenomena, such as earthquakes and viral infections, and human societies, such as income distribution and mortality from conflicts. This is because there is an invariant, non-random mechanism, often unknown, that governs the system and produces the observed findings.
The formula for calculating journal IF for any given year is: total number of citations of articles published in two prior years/total number of citable articles published in those two prior years. If n articles were cited during the two prior years and n citable articles were published in the two prior years, then IF = n/n = 1. In 2023 there were 141 citable articles in FACE, and in 2024 there were 178, totaling 319 for the 2-year period. The numerator, the sum of citations for 2023 and 2024, is 44 + 77, or 121. Our first IF for FACE is therefore 121/319, or about 40%. 3 , 4 this means on average 40% of all articles published in 2023, 2024 received one citation in the same period. More precisely this is the 2-year IF for 2025. Why two years? It allows for a larger sample size and evens out potential annual fluctuations. There are also 3-year and 4-year IFs for journals, and they are usually higher than the 2-year IFs, as citations continue to increase with time.
IF belongs to a large number of measurements collectively known as bibliometrics. H-index, for example, is a very common bibliometric. The “H” in H-index stands for Hirsch, the person who proposed its use to quantify otherwise qualitative impressions. Unlike IF, H-index does not count citations over a certain time period, rather it continuously counts all citations of all citable articles and reports the number of articles, x, for example, that have received x citations. The H-index for Albert Einstein, as of 2021, was 67, meaning that he has 67 articles which have been cited 67 times. 5 There are many other “factors,” “scores,” and “indices,” trying to answer how important, productive, or how good a particular publication actually is, a journal, an author, a university, or even a country. However, there are several critical questions we must ask, as we would and should any test or measurement.
What exactly are we measuring?
How do we validate the results?
Why are we measuring it?
Who will use the results and how?
We love tests and measurements, from head circumference, to IQ, to inflation and unemployment rates, to name just a few. Are IF and H-index really measuring how impactful a journal is? Many authors today have IF much higher than Albert Einstein, what does that mean? The great difficulty, inherent by nature, is the lack of a “gold standard,” Because of this deficiency, we cannot answer question 2 by validating these measurements effectively and accurately. Regarding the telos for bibliometrics, they allow for fair comparison, similar to what we do in meta-analysis by converting different outcomes to changes in average as a percentage of pooled standard deviation. 6 The last question, with its two parts, are both important and potentially dangerous. Results of measurements can and have been misused to devastating consequences. In his book The Mismeasurement of Man, Stephen J. Gould describes in detail the horrific saga of how Alfred Binet’s IQ test, initially designed to identify children in need of additional, specialized instructions, morphed by Henry H. Goddard as the basis for institutionalization and custodial care, or worse still: forced sterilization and the later eugenic movement in the United States in the early part of the last century. 7
Finally, looking at the common feature which underpins many bibliometric measurements: citation. We cite peer-reviewed, published articles to support our views and arguments. As a result of this practice, a paper, once published in a peer-reviewed journal, can have a very long life. It can be cited, and the paper citing it can itself be cited, and on and on, until another published paper or papers refute the argument. Such refutation may take years or decades. This highlights the extreme importance of editorial control at the front end and continuous, independent challenge. Science must be objective and subject to verification. 8