The average length of time that a research paper continues to be cited in the scientific literature depends on the length of the paper and the speed of growth in that area of science, according to a survey by Helmut Abt of the Kitt Peak National Observatory in the US (Nature 395 757). Abt, who is also editor of the Astrophysical Journal, studied papers published in leading astrophysics, chemistry, geophysics, physics and general science journals. He found that papers published in geophysics had the shortest "citation half-life", 6.4 years, while astrophysics papers had the longest, 29.3 years.
The survey was carried out by using the Science Citation Index to count the number of citations received by papers published in 1954 volume of the Astrophysical Journal in each of the 40 following years. The same was done for papers published in the January 1959 issue of Physics Review over a 30-year period. Abt then repeated this procedure for papers published in the Journal of the American Chemical Society, the Journal of Geophysical Research, and the Reports section of Science in 1959.
Abt found that papers published in the Physical Review had a half-life of just 10.6 years, significantly shorter than the Astrophysical Journal. He suggests that main reason for this is that astronomy has been growing more rapidly than physics in the past few decades. For example, the number of papers published in the Astrophysical Journal increased from 165 papers in 1954 to 1812 in 1994 – a growth factor of 11. When Abt added a correction factor to account for the growth rates in different subjects he found that the half-lifes became much shorter – 5.9 years in physics and 8.0 years in astronomy. Within astronomy he found that observational papers had much longer half-lives than theoretical papers. Abt also found that long papers tend to be cited more than shorter one. In physics the average number of citations received in 30 years is 7.4 + 6.7x(the number of pages).
