MARC Metadata Records
Obtaining MARC from ICPSR
ICPSR metadata is available in MARC21 format in two packages:
- One large marc file (containing all 6,678 studies)
- Discrete marc files, one for each study (6,678 files)
These files cover all additions and changes through 2007-12-31.
Obtaining MARC from OCLC
OCLC also makes available the MARC records for ICPSR in three separate collection sets based on the date when the records were submitted.
Please note that ICPSR does not control these records, and some were submitted long before ICPSR began the effort of creating its own MARC.
MARC Creation
The MARC creation process is complex.
First, ICPSR converts its DDI XML metadata records into MARC XML, using a custom-built stylesheet and oraxsl.
Then we use a series of scripts to prepare the data for OCLC. Library of Congress subject headings for the previously manually created studies by the University of Michigan University Library (UL) are extracted and appended to ICPSR metadata. In addition, we import OCLC IDs for those records that have previously been submitted to OCLC and have been updated by ICPSR.
Next, we use marc4j 
to convert the MARC XML into true MARC files.
We then submit the files to the University of Michigan library for quality control. The University Library submits the MARC records for new ICPSR studies for MARC authority control and OCLC. The records that we make available to Consortium members are the complete and up-to-date revised version of the DDI metadata. These records do not contain OCLC record numbers, however, per our licensing agreement with OCLC.
Updates
ICPSR updates its metadata with OCLC in the first quarter of every calendar year. It then updates the downloadable files on this page. When the updates are available, an announcement is sent over the OR-Announce email list. To be informed of updates, please join the ICPSR-Announce list.
To update your catalog records, please use the 035 field as the matchpoint. 035 contains the ICPSR study number, which is always unique.
