Authors who publish in the American Journal of Physics (AJP) or The Physics Teacher (TPT) will no doubt be happy that their article was accepted for publication, and even happier if it is widely read. Since December 2016, both journals have tracked online views. (In addition to these online views, the journals are read in libraries and by print subscribers.) Presumably, a view might include anything from a quick glance to a thorough study. What do these views tell us about the useful lifetime of an article?

An indication of the fate of an article is given in Fig. 1, which shows the average number of views from December 2016 to April 2022 of articles going back to the first issue (February 1933 for AJP and April 1963 for TPT). One issue was selected every 10 years, then at shorter intervals in recent years, and the average number of views of the articles in that issue is plotted vs the year of publication. Views before December 2016 are not included in the statistics. Some individual articles had more than 7000 views and were excluded from Fig. 1 since they were not representative. The most viewed article in AJP (February 2018) had 60 953 views and the most viewed article in TPT (October 2019) had 33 027 views.

Fig. 1.

Average views during the period from December 2016 to April 2022 of all papers in a selected issue, vs year of publication.

Fig. 1.

Average views during the period from December 2016 to April 2022 of all papers in a selected issue, vs year of publication.

Close modal

The data show that modern readers are far more likely to read recent publications than old publications, as one would expect. In that sense, most publications have a useful life of about 20 years in The Physics Teacher and about 40 years in American Journal of Physics, then fade into relative obscurity. The number of views for papers published in the February 2021 issue of AJP and the April 2021 issue of TPT is relatively low, since views were counted only one year after publication. A larger number of views is indicated for papers published from 2018 to 2020 since the views were counted over a greater number of years. Fewer views are observed for papers published before 2018 since (a) views before December 2016 were not counted, (b) readers tend not to read old papers, (c) more readers appear to be accessing the journal in more recent years, and (d) open access articles were not available before 2018.

A surprising result is that the number of views of most of my own papers has been increasing at a roughly constant rate since 2016, even if the paper was published around the year 2000. For example, a paper on bouncing balls published in 1999 had 406 views in December 2018, 628 views in February 2020, and 1054 views in April 2022. That result indicates that the number of readers has increased rapidly in recent years, in both journals, presumably because the internet has made the journals more accessible in many more countries throughout the world. Other authors may have observed a similar result, but progressive views are not published in the journals.

Additional data on many different journals are available in the Clarivate Journal Citation Reports database, describing citations rather than views or downloads. An interesting statistic is that the Cited Half-Life for American Journal of Physics is 22.4 years. That is, in 2021, half of all cited items in the journal were published in the previous 22.4 years, and the other half were published before then. That is roughly consistent with the half-life of views. By comparison, the Cited Half-Life for Physical Review B is 12.1 years, and for Nature is 10.6 years.