For all samples that use Civil War veterans, data are
collected for each individual from original sources in three datasets:
Data are input into
collection screens developed by CPE
programmers.
Following is a description of
the collection from each source.
Military, Medical, and Pension Records
The Military, Medical, and Pension Records are collected by
research assistants at the National Archives in Washington, DC. Information from three different sources is
input into the Military, Medical, and Pension Records dataset: the Pension, Military Service Record (MSR), and Carded Medical Record (CMR).
The collection
software consists of 8 screens.
Research assistants go through each record and input the information
asked for on the screens. For the Pension, this means reading through 50
to over 500 pages of information to extract the necessary information.
Surgeons’ Certificates
Surgeons’
Certificates are part of the Pension
Record, but are collected into a different dataset. The Surgeons’ Certificates are copied from
the pension and collected by separate research assistants.
The collection
software consists of 42 screens, separated into 21 disease groups. All information from each certificate is
collected into these screens. This
collection differs from the Military, Medical, and Pension Records
collection, in that all information from each certificate is captured in the
screens. Inputters
endeavor to capture the exact language from the certificates as far as
possible.
Census Records
The U.S. Federal
Census Records are searched for each census year that an individual is
known to be alive, beginning with 1850 (before 1850, only the head of household
is listed on the census). Note that the
1890 census is not available, as it was destroyed. For further information on
the 1890 census, see http://www.archives.gov/publications/prologue/1996/spring/1890-census-1.html.
Union Army Census Collection
When the Union Army Sample
was collected, the latest Census publicly available was the 1910 Federal
Census. Due to budget constraints, the
1870 and 1880 censuses were not collected (the 1880 census was later collected
on updated software). So, for each
individual, the 1850, 1860, 1900, and 1910 censuses were searched and the
information collected by research assistants at Brigham Young University in
Provo, UT.
The collection
software consisted of 4 screens, one for each census year collected. These screens follow the headings on the
original census manuscripts and information was input verbatim from the
manuscripts. The census manuscripts were
microfilmed by NARA
and are available at many locations. Each
individual was searched from paper indexes which led to where that person was
located on the microfilm. For a more
detailed description of how the censuses were searched, click here.
All Samples after the Union Army
All samples collected since the Union Army Sample (U.S.
Colored Troops Sample, Examination
Sample, IN/WI Sample, Oldest Old Sample, Urban Sample, etc.) have the benefit of much better access to
census records. Each individual is
searched for in all census years from 1850-1930 (except 1890, which was
destroyed) that he is known to be alive.
Soldiers who lack death information are searched through the 1880
census. All U.S. census manuscripts are now
available online, which greatly facilitates searching and has considerably
raised the find rates. The most useful
internet site for searching the censuses is Ancestry.com.
The collection
software was rewritten as a Windows application and additional screens for
1870, 1880, 1920, and 1930 were created, allowing greater flexibility in
inputting. The collection screens still
follow the census manuscripts. Research
assistants search the online resources for each individual and enter the
information for his household into the screens.
