The Rasch model. Measuring the impact of scientific journals: Analytical chemistry

Abstract

This article focuses on a fresh way to determine a ranking of science journals according to the “number of citations‐to and items‐published,” data used by SCI JCR of ISI to determine journal ranking by “impact factor.” This is the first application of latent traits theory to bibliometrics. “Impact” is considered a latent variable defined by a set of items (or indicators), citations, and sources. The theoretical background is Item Response Theory, which suggests that, if we can understand how each item in a set of items operates with an object, then we can estimate a measure for the object. The Rasch model is the commonest formulation of that theory. This technique is here applied to the citations and sources of 43 Analytical Chemistry journals (objects) to provide a Rasch measure for these journals which is compared with the current “impact factor” computation. © 1996 John Wiley & Sons, Inc.

What is wrong with obsolescence?

“…Like any other variable, “topicality” is conceived of as a directed line along which the items (years) and journals are situated (Alvarez & Pulgarín, 1996, 1999). One assumes there is a single direction that implies “more” of the variable.…”

Measuring information through topical subheadings of the Medline database: a case study

“…If there are two or more subheadings, differences in information will be revealed by their locations relative to a number of years. Thus, the latent trait 'information' is conceived of as a continuum along which parameters ␦ i for years and ␤ n for subheadings are located [3,4,5,6,7]: --------------------------------------------…”

The diffusion of scientific journals analyzed through citations

“…Diffusion may then be the SCI Journal Citation Reports of ISI, which gives the thought of as a latent variable defined by ''items'' that citations received by journals over time, we have defined are the citations made in different periods of time, and ''diffusion'' in a scientific field as referring to the dissemithe Rasch model is a useful instrument for measuring that nation of knowledge that is channeled through citations and distributed over different periods of time. We choose variable (Alvarez & PulgarıB n, 1996a;Andrich, 1988; the word ''diffusion'' because it seems freer of connota- Rasch, 1980;Wright & Stone, 1979). The analysis protions that might lead to misunderstandings or confusion posed here is not the same as that described in the previthan other possibilities such as ''popularity.''…”

“…The analysis protions that might lead to misunderstandings or confusion posed here is not the same as that described in the previthan other possibilities such as ''popularity.'' What is ous articles (Alvarez & PulgarıB n, 1996a, 1996b, alrelevant in a journal is not the number of copies of an though the methodology is identical. The difference beissue distributed, but the number of citations to its articles tween the two analyses does not lie in the differences over a period of time.…”

“…How are we to make use of the citations? (years) and journals are situated (Alvarez & PulgarıB n, 1996a). One assumes there is a single direction which…”