IASSIST Quarterly Vol 24 No. 3


8 IASSIST Quarterly Fall 2000

The Social Science Electronic Data Library:
Serving the Needs of Data Librarians and

Users

by Michael Carley &. Josefina J.
Card *

The last decade has witnessed enormous
strides in two areas:  first, the
development of numerous social science
data sets of high quality; and, second, the
development of the computing hardware
and software capability and infrastructure
needed to locate and analyze these data
sets for minimal cost and to communicate
data findings in interesting and easy-to-
understand fashion. Hand in hand with these advances in
data development have come technological advances which
allow social science research and teaching laboratories,
with the hardware and software needed to analyze the best
data in a given field, to be set up with ease by an academic
department or even by an individual professor.
Additionally, sophisticated data analysis software packages
formerly available only for mainframe computers have
become available for microcomputers at a much reduced
cost.  Taken together, these developments make it possible
for academic departments, research institutes, and
government offices of all sizes and levels of financial
resources to access and analyze exemplary data sets for
research, teaching, and program- and policy-development
purposes.

Data archives, in both the private and public sectors, allow
easy and open access to many hundreds of the best health
and social science data sets covering a broad range of
topics, study populations, and making use of a variety of
research designs.  The data available from many of these
archives are clean and the documentation user-friendly.  For
these and other reasons, researchers and instructors who are
considering the use of secondary data welcome the
functions served by a well designed data archive.  This data
is used to:

• conduct secondary analyses of outstanding data
sets to serve the needs of policy, practice, or basic
research;
• perform meta analyses based on access to
multiple original raw data sets;
• prepare research proposals on various issues;
• write publications comparing and contrasting
results from related data sets;
• prepare masters theses and doctoral
dissertations;
• produce classroom materials for teaching

substantive, methodological, and
statistical concepts from real-world
data.

In this paper, we will discuss three areas
of major concern for data librarians and
data users wishing to obtain and use data
for secondary analysis:  data quality,
format, and dissemination.  We review

and contrast the issues and concerns of data librarians and
data users, outlining areas of similarity and difference.  We
will explore how one large data collection, the Social
Science Electronic Data Library (SSEDL), compiled over
the last 17 years by Sociometrics Corporation, has
addressed each of these issues and the conflicts and
problems that arose during that process.  Finally, we peer
into the future, assessing how data providers can bridge
knowledge gaps via recent technological advances.

The Social Science Electronic Data Library (SSEDL)
The Sociometrics Social Science Electronic Data Library is
a premium health and social science resource that consists
of seven topically focused data archives. With over 300
data sets from 200 different studies comprising seven
topically-focused collections, it is a unique source of high
quality health and social science data and documentation
for researchers, educators, students, and policy analysts.
The Electronic Data Library was made available in 1999 on
a set of CD-ROMs and includes an online membership with
free access to datasets for downloading by members.

The Collections:
The Data Archive on Adolescent Pregnancy and
Pregnancy Prevention (DAAPPP) was established by the
US Office of Population Affairs (OPA) in 1982 as the
repository for the best social science data on the incidence,
prevalence, antecedents and consequences of teenage
pregnancy and family planning. In 1994, the scope of
DAAPPP was expanded to include studies that focus more
broadly on adolescent sexual health issues, thereby
including studies examining behavioral factors related to
sexually transmitted diseases (STDs) in addition to
pregnancy.  DAAPPP currently holds data from over 150
premiere studies (many of them longitudinal) on sexuality,
health, and adolescence.

State-of-the-art research data on the American family are



IASSIST Quarterly Fall 2000 9

available through the American Family Data Archive
(AFDA).  AFDA, funded by the National Institute for Child
Health and Human Development, contains data and
documentation from 20 nationally recognized studies on
important issues relating to American family life,
demographics, and family patterns.  Among the topics
covered are educational, economic, health, social, and
psychological indicators, child welfare, family violence,
marriage, divorce, child care and child custody.

The AIDS/STD Data Archive (AIDS) consists of original
research data and instruments from 11 premier studies on
AIDS/HIV and other sexually transmitted diseases (STDs).
The collection was established with funding from the
National Institute of Child Health and Human Development
(NICHD.  Included data sets address the following topics:
the incidence and prevalence of specific sexual behaviors
(including abstinence, vaginal and anal intercourse, oral-
genital sexual activity, masturbation); contraceptive and
STD-preventive behavior; attitudes and beliefs regarding
sexual behavior and methods of contraception and STD
prophylaxis; AIDS/HIV knowledge, attitudes, behavior,
and serostatus; current and past episodes of gonorrhea,
syphilis, chlamydia, and other STDs; and high-risk
behavior, including alcohol/drug use and prostitution.

The Maternal Drug Abuse Archive (MDA) brings
together seven state-of-the-art research databases on
maternal alcohol and drug abuse.  Funded by the National
Institute on Drug Abuse, the collection includes data on the
following topics: the prevalence of drug use among
pregnant women and women of childbearing age;
demographic characteristics of pregnant drug users; types
and patterns of illicit drug use; social, psychological and
economic antecedents of pre- and perinatal drug abuse; the
effects of pre- and perinatal substance use on pregnancy
complications and neonatal status; and the effects of fetal
alcohol and drug exposure on children’s physical,
neurobehavioral, psychological and social development.

The Data Archive of Social Research on Aging (DASRA)
was assembled with the support of a grant from the
National Institute on Aging. DASRA contains data and
documentation from three very large nationally recognized
studies. These three studies covered a variety of topics
including functional status and impairment, living
arrangements, caregiving and social support, health
attitudes, retirement income and plans, mortality, health,
financial resources and assets, expenditures, cognitive
ability, medical conditions, housing, health insurance, and
personal characteristics

The Research Archive on Disability in the United States
(RADIUS) was funded by the National Center for Medical
Rehabilitation Research (NCMRR) within the National
Institute for Child Health and Human Development
(NICHD). The purpose of the project is to facilitate access

to the best data sets on the prevalence, incidence,
correlates, and consequences of disability in the U.S.  The
heart of the archive is a collection of 19 studies that address
the topic of disability. These data sets permit analyses on
topics such as: the incidence and prevalence of specific
diseases, disorders, and impairments, including deficits of
cognition, emotion, physiology, and anatomical structure;
functional limitations across a variety of specific organ
systems; disabilities in relation to major life roles and
activities, such as work, parenting, education, and
recreation; societal limitations including physical,
attitudinal, and economical barriers that restrict full
participation in society; psychosocial and interpersonal
factors such as coping with stress, sexuality, feelings of
control and productivity, quality of life, and family
relations and support; health care and rehabilitation issues
such as medical costs, coverage, service utilization, use of
orthotic, prosthetic, assistive devices, effectiveness of
rehabilitation; as well as a variety of basic demographic
factors on respondents such as age, race, sex, income,
occupation, marital status, family size, and living
arrangements.

To facilitate access to the best contextual data,
Sociometrics has developed a Contextual Data Archive.
By contextual data we mean data that describe the
population, social, and economic characteristics of
geographic areas, from census tracts to states, in which
people reside or work. The contextual data  archive consists
of a series of files, each organized around a different
geographic unit of analysis (such as census tracts, school
districts, counties, states, etc).  Each file contains variables
drawn from various sources, but having one common
geographic unit of analysis. Support for this project was
provided by the National Institute of Child Health and
Human Development.

Data Quality
Both data librarians and data users have an abiding interest
in the availability of high quality digital data.  The librarian
must put her/his limited resources to the most efficient and
effective use possible.  The resources we mean here are not
only financial assets such as approved budgets, but also
material and human capital such as shelf space and the
person-hours of those who must purchase, assemble, and
maintain various data collections.  Because all of these
resources are limited and precious, data librarians must
make wise choices as to how to prioritize their use in order
to achieve the highest quality collection of data for the
users at their institutions.

Data users are also concerned about having data of the
highest possible quality.  The users of digital data wish to
make the best possible contribution to the body of
knowledge in their field.  That contribution is placed in
jeopardy if the data used is of questionable quality.  Data
gathered via poor research design, or via a good design



10 IASSIST Quarterly Fall 2000

poorly executed are of little use in advancing knowledge.
Researchers using such data risk not only making a tainted
contribution, but also of generating criticism from
colleagues and associates who recognize the problems or
limitations of the data being used.

The research staff at Sociometrics recognized these needs
when compiling the seven data archives in the Social
Science Electronic Data Library.  It was determined that
each archive would be a ‘best of the lot’ collection,
accepting only the best data available in each of the seven
topic areas.  To accomplish this, we formed National
Advisory Panels of research scientists who were experts in
both the substantive content of the particular archive and
the research methods commonly applied in that field.  The
panel, usually consisting of six members, was asked to
evaluate candidate data sets on the following five criteria:

• Technical quality:  among the factors to be
considered are high response rates, low attrition
rates, use of reliable and valid measures, and
sound sampling and design elements.
• Substantive importance to the field:  factors
include the potential to address contemporary
issues, to break new ground, and to replicate or
confirm important findings.
• Program or policy relevance:  the ability of
the data set to answer applied questions on how to
improve public policy or shape intervention
programs;
• Potential for secondary analysis, including:
Scope of sample - The broader or more diverse the
scope of the sample, the greater the potential of
the data for generalization.  Size of sample -
Sample size is always an important consideration.
This is even more true for data intended for
secondary analysis:  sample sizes adequate to
support the originally intended analysis may be
too small to support other analyses, especially if
the new analyses focus on data cells that have a
very low proportion of cases.  Breadth of
variables and constructs covered - The potential
for secondary analysis is directly related to the
breadth of variables measured in the data set.  The
more numerous and diverse the set of variables,
the more possibilities there are for new or
expanded analyses.
• Disciplinary balance:  An archive should
attempt to be representative of the entire field of
research.  Variations in state of the art exist
between different sub-areas within any discipline.
Thus a somewhat flexible standard (as measured
by the other criteria above) should be used to
ensure that all major areas of the discipline are
represented in the archive as a whole.

In order to perform these evaluations, we provided each

panel member with briefing materials consisting of a 2-4
page description of the data source, which covered: the
purpose of the study; methods (including sampling design,
periodicity, unit of analysis, response rates, and attrition);
content (description of variables covered, number of
variables, and topics covered); limitations; sponsorship;
and a bibliography.  In addition, we provided copies of
original peer-reviewed publications for each data set, which
allowed the panel members to review issues we may have
not addressed in our briefing documents.  Panel members
were encouraged to suggest additional data sets for
consideration, and have often done so.  Panel members did
not vote yes or no on each data set, but rather rated each
with a ‘priority score’ from 1-10.  Only those data sets
receiving an average score of 7 or above were accepted,
and higher priority for archiving was given to those
receiving higher scores.

In addition to pre-screening the data sets, archivists for the
SSEDL data sets perform several other tasks designed to
ensure data quality.  We review the data thoroughly,
checking to make sure that all variable and value labels are
included and are sufficiently descriptive.  We check the
data for internal consistency and completeness, scanning in
particular for variables with an excessive number of
missing or out-of-range values.  We also perform random
checks verifying that the skip logic in the original
instrument was followed and that the variables are
consistent in relation to one another (no variables
describing a female as a father or brother or a male as a
mother or sister, etc).  Finally, we produce a user’s guide to
the machine-readable files and documentation which notes
any remaining limitations or inconsistencies.

Those archiving digital data face many challenges, not the
least of which are the limitations on their own, as well as
the users’ time and resources.  Consequently, different data
archivists take a variety of approaches to address these
challenges.  Our ‘best of the lot’ approach emphasizes
quality over quantity.  This means that our data collections,
while not as large as some of those from other sources, are
of higher overall quality and are better documented than is
the industry average.  This approach limits the size of our
collections, but contributes to their popularity among
researchers for their high quality and ease of use.

Formats
The needs of data users and librarians diverge somewhat
when it comes to format preferences for digital data
collections.  Users look for data in the most easily
accessible form, while librarians must be concerned with
the big picture, and look for collections that serve the needs
of as many users as possible, both in the present and the
future.  The format(s) in which data are provided also have
an impact on the role the librarian will take in the data
distribution process, which could vary from that of a
facilitator to an active gatekeeper.  The challenge for data



IASSIST Quarterly Fall 2000 11

providers is to address both the preservation needs of the
librarian and the ease of use needs of data users.

Distribution Media.   Librarians must conserve their many
precious resources, including both financial resources and
shelf space.  However, they also desire that data collections
be in a form that is not easily lost, damaged, or misused by
careless users.  Therefore, a data collection should be
durable, and in a format that is not likely to change rapidly
with evolving technology.  CD-ROM technology meets
these requirements.  CDs are more durable than diskettes
and do not have the associated (at least perceived) transient
qualities of internet sites.  Data made available on a CD-
ROM are not likely to be easily lost as library staff can,
should they choose to, maintain tight control over them or
copy their contents to a central repository for safekeeping.
Diskettes are more likely to become corrupted and internet
sites often are revised and require more constant updating
on the part of both the data provider and the librarian to
keep all links accurate and up to date.

Data users, on the other hand, prefer that the data are made
available in the simplest, easiest to access format possible.
However data are made available, it must be transferred to
the computer where the user will actually be working.
With desktop computer speed and hard drive space
increasing exponentially, users often prefer that data be
available for copying to their own system, rather than
residing at a central repository. Internet downloads may be
preferred to CD-ROMs as the data can be copied to the
user’s own computer, then manipulated and transformed as
necessary.

To best accommodate these divergent needs, we found it
necessary to make our data sets available in both CD-ROM
format and via our internet web site.  SSEDL Volume I is
distributed via 17 CD-ROMs, along with accompanying
support material.  Additionally, each purchasing institution
is given free web access to all of the data sets, as well as to
new data that have not yet been added to the CD collection.
By allowing the users to download the data sets or use the
CD-ROMs, we were able to provide both the user and the
data librarian with flexibility in both data format and data
access.

Analytic Software.  A user who wishes to use data on a
particular topic would be best served by data that can be
retrieved quickly and effortlessly with a variety of
software.  Given the wide variety of statistical software
available to users in different fields, this can be a challenge.
Users should have the capacity to use the software of their
choice, and the ability to access the data quickly with that
software.  At the same time, data providers must
understand that software currently popular may change or
become outdated, making files created from these programs
unusable or at the least cumbersome and inefficient to use.

Most data collections have taken one of two approaches to
this problem.  First, the data provider may distribute raw
data with a codebook.  The raw data is typically stored in
an ascii file which is simply useless text (numbers) without
the codebook.  The codebook provides the user with the
location of specific variables and cases within the raw data
file.  The advantage to this approach is that it addresses the
issue of durability well.  Users can access the data by
writing a program using the statistical software package of
their choice, inserting the variable and case locations given
in the codebook to access the data.  Changes in software
applications do not affect data distributed in this method, as
users write their own program with the language in which
they have expertise.  However, writing the programs to
read the raw data can be a time consuming process, causing
users to waste much of their resources on mundane tasks.
In addition, such writing is prone to error; one misplaced
character can cause much of the data to be written
incorrectly.

Other data providers address these issues by distributing the
data in a pre-packaged format using one of the most
popular statistical software packages (typically SPSS or
SAS).  By distributing these formatted files (usually either
complete system files or portable files), the user can access
the data directly simply by opening the files in the
appropriate software.  When portable files are used, the
data can be used with different versions of the same
software package (either earlier versus later versions or
versions for different operating systems) or in some limited
cases, in other popular statistical packages.  The advantage
to this method is clear:  quick, easy access to data.  The
disadvantage is that this approach cannot possibly be
flexible enough to address all user needs.  Some users wish
to access the data with a software package that is not
among the most popular.  Also, data distributed in this
method can become unusable when software packages
radically change their formats.  Data made available in the
most recent format could be inaccessible within just a few
years.

To address the limitations in each of the above approaches,
data sets in SSEDL are distributed with raw data files and
machine-readable set-up statements for use with both SPSS
and SAS statistical software.  These set-up statements
provide for the best of both worlds:  ease of use, combined
with flexibility.  Users use these syntax files to create the
system or portable files in whichever software they are
using.  For those users who are use software other than
SPSS or SAS, the set up statements serve essentially the
same purpose as the codebook described above.  Because
the set-up files are machine-readable, they can often be
converted for use with other software with a minimal
investment of time.  Should the syntax requirements of
SPSS or SAS change radically, these files would serve also
accomplish this goal.



12 IASSIST Quarterly Fall 2000

In addition to the above files, SSEDL data sets also are
distributed with a machine-readable SPSS data dictionary
file and an SPSS frequency and statistics file.  These aid
the user in making sure that they have created their system
or portable files correctly.  Users can compare the statistics
in the frequency file to their own, thereby preventing
mistakes in data analysis.  Both the frequency and
dictionary files are also useful in reviewing the contents of
the data set.  Each data set is also accompanied by a
printed User’s Guide (provided in machine-readable form,
in addition to printed form, for the more recent archives)
comprised of a standard set of sections and subsections.
The provision of standard machine-readable and printed
documentation assists users in familiarizing themselves
with the Sociometrics data sets.  Once a user has worked
with one Sociometrics-packaged data set, it is easy for him
or her to work with any of the others. The original
instrument and codebook are offered as optional,
supplementary documentation for each data set, when
available.  For the more recent archives, the original
instrument is distributed in machine-readable form along
with the data, as a set of graphics files (page images).

Search and Retrieval Software.  As data sets get larger,
both in the number and scope of variables covered and in
the number of cases, users are faced with an increasingly
overwhelming task of reviewing which parts of a study are
necessary and appropriate to their needs.  Often, users will
begin work with a data set containing over 5,000 variables
(and often several thousand cases) only to find that their
interests only require 30-40 of those variables.  It is
important for users to be able to quickly sort though the
variable list and find those of interest.  Given current
(though perhaps temporary) limitations in speed and disk
space, users also need to be able to reduce the large data
set into one with only those variables needed for analysis.

To address this need, Sociometrics staff developed
powerful search & retrieval software which now
accompanies each data archive.  This software allows a
user to search an entire topically-focused collection, a
customized group of data sets created explicitly for a given
user, or a single data set; to identify variables of interest
across this designated search space and to save located
variables as a search set.  Users can conduct: (1) full-text
keyword searches, including variable names, words in
variable labels (question descriptors), and words in value
labels (response descriptors); (2) searches by assigned
topic and type codes; and (3) searches by study name or
assigned data set number.  Standard Boolean operators
(i.e., “and,” “or,” “not”) can be used to combine search
sets.  Alongside this software, we provide data extract
software which allows users of CD-ROM versions of
archived data sets to create customized SPSS or SAS
program files containing only those variables of interest to
them.  This capability permits analyses of subsets of large
data sets to be conducted quickly (with rapid turn-around)

on most microcomputers.  It also saves users significant
program development time writing and re-writing SPSS
and SAS program statements to define variables used in a
given analysis.

Technical  Support
The Role of the Data Librarian. The role of the data
librarian in this process varies a great deal among
institutions.  In some cases, the librarian serves as an expert
gatekeeper to the data, allowing access to users as s/he
deems appropriate and answering a wide range of questions
users may have.  Others may serve a minimal role, simply
providing access to the data and support materials and little
else.  Librarians also vary in their level of statistical
knowledge, as well as their expertise in the various topics
that may be covered by the data in their collections.

Our approach to this issue was again to provide the greatest
amount of flexibility in the collection as possible.  While
some data librarians do take on something close to a
gatekeeper role, we chose to allow for those who had
neither the time nor the expertise to do so.  Librarians need
to be fully informed, not on all of the topics included in
their data collections, but on the process by which they can
aid users in finding data of interest to them.  Given the
wide variety of topics covered in SSEDL, one could never
expect librarians to provide users with all of the help they
may require.  To this end, the Social Science Electronic
Data Library includes a variety of tools to facilitate this
process.  A user’s manual and contents manual detail the
data sets included in the collection, and a quick start guide
offers advice on how to implement the software and the
knowledge needed to use the data sets.  Most importantly, a
‘Guide to the Social Science Electronic Data Library’ CD
is provided with each collection.  This CD takes the
librarian through the process of using the data library using
a brief step by step tutorial.

SSEDL’s Research Support Group.  In addition to the
support given to the data librarian, users have direct access
to help from Sociometrics’ archiving and scientific staff
through our Research Support Group (RSG). The Research
Support Group consists of Ph.D. and Masters level social
scientists who provide free technical assistance for users
who have questions about accessing or using our data sets.
In addition, the RSG occasionally performs consultant
work such as the creation of customized data set extracts;
user-defined statistical tables and analyses; data archiving,
management and analysis services; customized CD-ROMs;
and training workshops.  These services greatly aid users
with limited expertise or resources with which to conduct
their own analyses.

Dissemination
The manner and methods by which digital data are
disseminated by data providers and eventually by data
librarians are crucial to the usability of such data.  Users



IASSIST Quarterly Fall 2000 13

must be made aware of the availability of data that meets
their needs.  However, with today’s rapidly expanding
technologies, the problem of ‘information overload’ is a
crucial one.  Even within our own collections, users can
become overwhelmed with the sheer amount of information
available to them.  If these issues are not handled properly,
they can inhibit the users’ ability to locate and make use of
the most appropriate data.  Data providers must do what
they can to ensure that users have the capability to quickly
and easily locate the data sets and even the particular
variables their topic of interest requires.

While the search and retrieval software made available for
each individual data archive helps users find variables
within a study they have already chosen, it does not help
when users have not yet selected the study that meets their
needs.  To address this issue, we created a search
mechanism for use on our internet site which is cross
archive.  This allows for users to search by keyword(s), or
designated variable topic or type, for variables of interest in
all of the SSEDL data sets.  Users can search for words in
variable labels (question descriptors) or in value labels
(response descriptors).  In addition, we include a brief
abstract of each study, and users can search for keywords
within those abstracts.  We chose to make this software
available to the general public as well as users on our
internet site, in order to allow researchers at non-
purchasing institutions the opportunity to find data sets of
interest and order them individually.

Data librarians must also make an effort to make help users
become aware of available data collections.  In order to
facilitate this process, we provided not only the above
mentioned guide to SSEDL on CD-ROM, we provided
informational flyers and brochures to help the librarian
make potential users aware of the availability of our data
collections.  In addition, the Research Support Group
provides both librarians and users ongoing advice as to how
to find data sets of interest in our collections.

Looking to the Future:  New Technologies, New
Audiences
The value of any data collection is in part predicated upon
its ability to address issues of the day.  Therefore, any
collection will inherently be of greater value the newer the
data are that are contained within it.  The preservation of
historic data is clearly important, but any collection that
aspires to be useful must also be kept up to date with the
addition of more recent data.  We will continue enlarging
the content and capabilities of our data set collections.  We
will be adding to our current data archives as funds permit,
and expanding our efforts by adding new topic areas to the
collection.  We expect to begin the establishment of a data
archive on child well-being shortly. A feasibility study on
the formation os a complementary and alternative medicine
data archive has just been successfully completed.

In putting together the SSEDL, Sociometrics’ staff have
learned a great deal about data collection methods and
ways to improve efficiency and reduce costs to researchers.
Currently, we are developing a software product that will
aid researchers on this aspect of the process.  Sociometrics’
Automated Dataset Development Software (ADDS) is an
integrated software program that, when completed, will
develop and document social science research studies. The
program will perform the following functions: 1)
Instrument generation—generate a fully formatted research
instrument in print, ASCII, and other machine-readable
formats.  2) Codebook generation—generate the data set
documentation in a printed codebook (also in ASCII and
other formats), flow chart (skip map), and data file map.  3)
Data entry—provide for data entry from completed
questionnaires, with simultaneous error checking.  4)
Program file generation—produce a raw data file in ASCII
format, and build the program statement files needed to
transform the raw data file into SPSS and/or SAS system
files.  The software will automate tasks best done by
computer, improve instrumentation and documentation by
providing a complete, high-quality structure and format,
and reduce the post data-collection effort of documenting a
public-use data set.

Additionally, we are also building an item bank of high
quality, commonly used questions, scales, and interviewing
tools from the SSEDL collection.  This bank will be
accessible within the ADDS program to permit users to
select questions to develop their own research instruments.
The item bank will be filled with several thousand
questionnaire items drawn from some of the leading studies
in research on the American family.  Using questions or
scales that have been previously tested will not only
improve the choice of questions, but will also lead to
greater comparability between studies and over time.

In addition to keeping the Social Science Electronic Data
Library current and helping researchers imrpove their
methods, we hope to reach new audiences through new
technologically innovative products.  Bridging the
‘knowledge gap’ is of prime importance to those who wish
to make practical contributions through social science
research.  Rather than limiting our efforts to trained
researchers, we must reach out to other professionals, and,
when possible, the lay public as well.  We are beginning
our efforts to reach the ‘paraprofessional’ audience with
two new products related to the SSEDL.  The U.S. Social
Surveys: A Sampler of Questions and Responses, will
contain searchable, edit-ready, and print-ready machine-
readable versions of  the demographic, behavioral, and
health science instrumentsæquestionnaires, medical forms,
interview protocolsæused to collect the data in SSEDL.
Questionnaire items will be linked to crosstabulations with
age, race/ethnicity, and gender, obtained from the linked
SSEDL data archives.



14 IASSIST Quarterly Fall 2000

Secondly, the Multivariate Interactive Data Analysis
System (MIDAS), will allow online analysis of the data in
SSEDL. Online data analytic procedures will include
weighted and unweighted frequencies, percentiles, and
measures of dispersion and central tendency, as well as
two-way and n-way tables with measures of association,
comparison of means (2-group and ANOVA) and
correlations, and the calculation of complex variance
estimations. Users will be able to define case subsets,
recodes, or aggregations for analysis, and then produce
output which can be downloaded or printed. Custom
dataset downloads will also be available.

The goal of ADDS is to aid expert researchers in handling
the ‘front end’ of the research process.  Through the use of
the ADDS software, researchers will be able to reduce their
costs and improve the accuracy and efficiency of
instrument development, data collection, input,
management, and analysis.  The goal of Social Surveys and
MIDAS is to increase the accessibility of the data to those
who are not competent in the sophisticated statistical
software packages such as SPSS or SAS.  These products
will help the ‘paraprofessional’--people with college
degrees who are not necessarily trained in complex data
analysis— avail themselves of exemplary social science
data.  They will provide a basic introduction to social
science methodologies as well, and will be linked to the
SSEDL data sets for those who wish to progress to the next
stage of data analysis (e.g., advanced undergraduate
students).

In sum, the historic progression of SSEDL has been to
expand the definition and purpose of a data archive.
SSEDL staff have worked to enhance archiving methods to
make data easily accessible to researchers.  Our advances in
this field have helped to make the research process more
efficient, especially for those conducting secondary
analysis.  ADDS will improve the process for primary
research as well.  Non-researchers will be introduced to
data analysis through the new products, Social Surveys and
MIDAS.  Together, these products will allow us to extract
the greatest possible value from our research dollars as data
will be used in as many ways as are feasible and by a much
wider audience.

* Michael Carley and Josefina J. Card, Sociometrics
Corporation, Contact Name and Address: Josefina J. Card
Sociometrics Corporation, 170 State St. Suite 260, Los
Altos CA 94022, (650) 949-3282, ext. 211, FAX (650)
949-3299, jjcard@socio.com

mailtoi:jjcard@socio.com