Code4Lib Article on EAD Tag Usage Analysis

The contents of this post relate to EAD 2002.

In the newest issue (22) of the Code4Lib included a fascinating article analyzing EAD tag usage across repositories. The team analyzed EAD-encoded finding aids, using the beta ArchivesGrid discovery system to document usage and more. The article includes both raw data about how often things were used, or repeated, in finding aids. The authors note that “not all elements are created equally,” so we needn’t expect, say, 50%+ usage of <frontmatter> (which is actually down at 37%) because it can be entirely generated using information from <eadheader>.

It’s definitely worth reading to see what current (perhaps not best) practices are across 120,000+ documents. The authors address not only % of usage of such heavily-used tags as <unitdate>—which was not used as often as I thought it would have been within <did> (73%)—but also the variance in descriptive practices which may make it less useful for sorting. For example, if <unittitle> is “Records” or “Reports,” searching that field along is not going to produce useful results unless, say, the <persname> or <corpname> or other relevant field is included as well.

Also, thanks M. Bron, M. Proffitt and B. Washburn for the EADiva recommendation.