On Tuesday Nov. 12 I attended a webinar sponsored by Swets and Library Journal, on the topic of measuring library value using data. The webinar has been archived if you are interested in watching it. It had three speakers and a moderator, and the middle session impressed me the most; it was about determining journal value within a disciplinary context.
The speaker, Chan Li of the California Digital Library (CDL), described the “weighted value metric” that the CDL is developing to evaluate the journal subscriptions it supplies to the University of California (UC) system. This metric combines multiple perspectives on the value of a journal: utility, quality, and cost effectiveness.
Each of these perspectives is composed of two measures and weighted to create a value score.
Utility = Usage by UC patrons, Citations by UC faculty
Quality = Impact Factor, Source Normalized Impact Per Paper (SNIP*)
Cost Effectiveness = Cost Per Use, Cost Per SNIP
*SNIP — Journal-level metric developed at Leiden University to”correct for differences in citation practices between scientific fields.” Defined as the ratio between impact factor and the citation potential in a subject field. Broader title coverage than journal impact factor. Freely available at the JournalMetrics.com website.
Each of these six measures is then weighted. Usage gets 30%, and the remaining 5 measures each get 14%. Li didn’t explain how exactly, but the resulting score is then converted into quartiles on a 1-7 scale, in which 0-1.5 is “Lowest” value, 2-3 is “Low,” 3.5-5.5 is “Medium” and 5.5-7.0 is “High.”
Within a discipline, each journal is compared to the discipline’s median score which becomes the discipline benchmark.
Possible future changes to the metric algorithm: weighting PDF vs. HTML views/downloads; social media saving/sharing. Chan and her colleagues plan to publish an article about this metric so stay tuned!
— Heidi Senior