Data Sources
What if you need some data, either for a presentation, project or other application, but don’t
happen to have any at hand? There are many good resources for data, univariate or mulitvariate,
that may be obtained at no cost from a variety of sources. Here are a few such resources, with a
note or two about what each has to offer.
• Data and Story Library (DASL). DASL is a resource maintained at Carnegie-Mellon University
which has a variety of data sets, some oriented toward a specific data analysis method, others
could be handled in a variety of ways. Though the data sets are numerous, they tend to be small,
that is, based on a relatively small number of cases. The data sets also tend to be somewhat
dated. That doesn’t decrease their utility, however. Perhaps the biggest deficiency is that there
is often little information about the actual source of the data, explanation of how the data were
collected, and, in a few instances, the choice of how the data are quantified. Some of the data
may be fictitious–drawn from text examples rather than real data.
• Australasian Data and Story Library. OzDASL is similar to DASL, though the data sets tend to be
somewhat better documented.
• National Center for Education Statistics (NCES). NCES is the home for all of the data collected
and summarized in the “Nation’s Report Card,” which is based on the National Assessment of
Educational Progress (NAEP) test series. There is much you can find here; some of which is in
summary form, and some of which can be obtained in raw form.
• Inter-university Consortium for Political and Social Research (ICPSR). ICPSR is a treasure trove of
data sets, many of which are from very large studies, sometimes done longitudinally over many
years. Most of the data sets come from survey-type studies, but you should have no problem
finding something here.
• Mississippi Assessment and Accountability Reporting System (MAARS). MAARS is a convenient
resource for school, district, and state-level information relative to test performance of
Mississippi students from elementary school through high school. The basic retrieval system for
MAARS gives data at a specific unit level (e.g., aggregated for a school, a district, or the state),
but the Mississippi Department of Education web site also has some larger-level files that can be
downloaded as well.
• StatSci.Org list of data sets. Here’s a large list of links to data sets, some of which have been
listed above (like DASL), some from textbooks, some from elsewhere. If you can’t find something
from this site, you’re just not trying!
• HUD USER data sets. From the department of Housing and Urban Development comes a variety
of electronically-distributed data sets that can be accessed. These are data sets that have been
obtained from the Office of Policy Development and Research (PD&R) at HUD.
• Journal of Statistics Education data sets. A variety of data sets, usually larger than those in DASL,
which are submitted by authors to accompany their articles in JSE. Each data set typically has a
corresponding file with a description, though you can always consult the article.
• Need more? Just type “data set” into Google or some other internet search engine, then enjoy
the ride.
