SPEC Kit 334: Research Data Management Services · 15
responding libraries have made initial forays into data
archiving services. As funders expand data sharing
requirements and efforts at compliance,5 libraries may
see an increasing role and call from researchers to
assist in finding data archiving and dissemination
solutions.
The survey defined data archiving as providing
“longer-term retention of and access to research data
by others.” A data archive is distinguished from other
digital file storage systems by features that include
online file access, file integrity checks, and perma-
nent identifiers for locating files (Q19). Forty of the 54
respondents who provide RDM services (74%) report
their library offers data archiving. At 13 of these insti-
tutions archiving is also offered elsewhere. Five insti-
tutions (9%) only offer data archiving elsewhere than
the library. The other archiving units are most often
central IT and research centers however, in some cas-
es such services might have been considered central-
ized storage rather than archiving by our definition.
Whether or not libraries offer their own archiving
solution, nearly all the libraries with RDM services
offer assistance locating data archiving solutions for
research data, for example, at data repositories for
a specific domain of research such as neuroscience
(Q20). Twenty-six libraries (48%) help researchers de-
posit data at such repositories, and 21 (39%) have as-
sisted with depositing data at journals, which are in-
creasingly asking for supporting datasets. All but one
of the 40 libraries that have their own archives offer
direct assistance with depositing data in that archive.
Data Archive Characteristics
From prior literature and informal observations, we
expected current archiving solutions among librar-
ies to fall into two categories: repositories designed
specifically for research data, and those repositories
built for other purposes that can contain datasets. For
the latter, we expected institutional repositories (IRs)
used primarily for publications to be most common,
as well as datasets archived in digital repositories that
are used for institutional special collections such as
historical photos. We expected data-specific archives
built specifically for retaining and accessing research
data to be less prevalent because of the higher resource
requirements for setting up a system, the lack of ready-
made data repository platforms (in contrast to the
more mature enterprise of institutional repository
software), and the relative lack of demand for libraries
to initiate such platforms at this time.
Question 21 attempted to distinguish these two
categories, but after comparing Q21 responses to
those for other questions, and to links respondents
supplied to their archive websites, we found varying
interpretations of what respondents reported as a
“research data archive dedicated to data deposits and
access.” All but 5 of the 13 “data archives” appear to be
institutional or digital repositories by our definition.
The survey questions and responses section lists the
original responses, but the table below redistributes
the 40 data archiving libraries’ primary method into
three categories: institutional repositories (IRs) that
include datasets in addition to publications digital
repositories used for a variety of institutional collec-
tions besides data, such as digital photos, rather than
faculty publications and archiving solutions more
clearly dedicated to research datasets.
Archive Type (N=40) N %
IR with datasets 30 75%
Digital repository with datasets 5 13%
Data-specific repository 5 13%
Since 88% of the libraries that archive research
data use either publication- or digital collection-cen-
tered repositories, the analysis of the survey’s data ar-
chiving questions distinguishes between responses of
the five libraries we labeled as having “data archives”
from those with institutional repositories or digital
collections, which we combined into a single category
“IRs with data.” This categorization may not be exact
in all cases, so we encourage those doing further stud-
ies on library data archiving to verify directly with the
libraries their current methods.
Funding Data Archives
As is the case for funding RDM services overall (Q49),
most of the libraries with data archiving services (31
of 37, or 84%) are absorbing those costs through their
internal budgets (Q24). Nine fund archiving through
Previous Page Next Page