Investigating the effect of definitions and best practice guidelines on errors in Dublin Core metadata records

Abstract

Researchers argue that the utility of metadata records depends on the kinds of information used (i.e. definitions and best practice guidelines) when creating metadata. To verify this claim, a mixed factorial design experiment was conducted in which 120 participants were assigned to four groups on the basis of their training level and the kinds of information (definitions only or best practice guidelines) accessible to create metadata records. Participants used Dublin Core to create records for the same resources and errors were analysed and compared across groups. Although participants who used best practice guidelines made significantly fewer errors than those participants who used only definitions, the high error rates observed across all treatment groups suggest the need for additional measures to control errors in metadata records.