home > data services > working with data

Working with Data

The Data Services collection resides on several digital media including data servers, CD-ROM, and online services. How you access and work with the data will depend on which study or data resource you select for your research.

Searching For a Study on Your Subject

Prior to accessing and working with a data file, you must identify a study or database that supplies you with the needed variables. General information about the major data resources available through Data Services can be accessed at Data Resources. However, you will likely need to do more in-depth searching to find a specific study on your topic. One of the primary data resources available to UGA researchers in the social sciences is the extensive ICPSR (Inter-university Consortium for Political and Social Research) data archive. You can search for studies in ICPSR's archive in three ways:

  • Search the archival holdings by words and phrases appearing in the title, study description (abstract), principal investigator, and ICPSR study number fields. Limited Boolean capability is available in the advanced mode. Browse by subject using ICPSR's outline of subject headings.

  • Browse the thesaurus for detailed subject terms, geographic terms, and personal names.

When you retrieve a list of studies using the above methods of information retrieval, you should read the study descriptions for each study of possible interest. By reading the study description you can usually get a better idea of the general applicability of the study for your research needs.

Once you have narrowed down your list of possible studies to just a few, you should consult the respective codebooks (aka documentation). Codebooks typically provide information about the methodology used, the complete variable and question lists, and the structure of the data files. Examining this detailed information carefully is the best way to make a judgement about the suitability of the study for your research needs with respect to the subjects covered, the statistical advisability of using the data, and the practicality of using the data. Codebooks are commonly available online. (See How to download ICPSR data.) Printed codebooks are located in the Main Reference room. Many of the codebooks are also available in the Libraries' circulating collection. Call numbers for many of the codebooks can be found in GIL, the UGA Libraries online catalog. If you are looking for survey or poll questionnaires without necessarily accessing the actual data, please see Finding Survey Questionnaires. This guide covers several data resources in addition to ICPSR, including the Roper Center's iPoll database. An additional method for identifying studies and databases by subject is to search for published quantitative research and statistical reports. Information provided about the data sources used in this literature can sometimes be quite useful in tracking down the underlying data. LexisNexis Statistical and many other bibliographic databases in GALILEO provide excellent subject access to the published literature of many academic disciplines. Once you have identified an interesting data source, you may be able to find it in the ICPSR archive or on the web. If you are unable to find a study, data resource, or codebook suitable for your research needs, or if you need help in assessing the suitability of a study for your research, please contact the Data librarian for assistance prechtel@uga.edu or 706-542-3472.

Accessing the Data

ICPSR now offers direct downloading of their collection of studies to users at on-campus and off-campus workstations. (See How to download ICPSR data.) It is recommended that users access their data through ICPSR if possible because only the ICPSR server is guaranteed to offer the latest version of each study. However, Data Services has also mounted many files on our local server. These include both ICPSR data and selected government data files. UGA users of our local server may fill out our Request ICPSR Data form to request that Data Services add ICPSR datasets to our local server as well as update an older study version with a newer version. Data from government sources as well as some non-governmental sources is increasingly available on the web. Try searching with your favorite web search engine. (See Finding a Web Site.) Many U.S. Government data files are available locally on CD-ROM and other portable media. In most cases users may request either to use these files in the Libraries or to circulate them. In a few cases the files may only be used in the Libraries. To access restricted online data resources such as Datastream, or Compustat/CRSP as well as SPSS and SAS software you will need to use the special access Data Services workstations in the Main Reference room. If you need assistance with accessing a study, data resource, or CD-ROM please contact the Data Librarian at prechtel@uga.edu or 706-542-3472.

Working With the Data

All of our restricted data resources as well as many government databases and a small number of ICPSR studies are provided with an interface that enables the user to retrieve and manipulate the underlying data. However, much of our collection, including most of the ICPSR archive, consists of raw data files and associated help files. To subset and analyze raw data, you need to be familiar with a statistical software package, such as SAS or SPSS. Some files can also be manipulated with spreadsheet or database management software. But, in all cases you must first bring the data into your statistical software. Depending on the format of the raw data files, this may be as simple as opening the data file with your statistical software after making sure the data file is correctly named. In some cases it may be necessary to utilize a setup ("syntax") file to read the data into the statistical software format. (See Using SPSS for useful tips.) In still other cases you may need to write your own setup file from scratch or perform other kinds of data "massaging" to get the data into a useable format. Limited individual assistance with this process is available from the Data Librarian.

EITS Statistical Software Support offers some programming and software assistance, including periodic seminars on SAS. If you need assistance with statistical methodology, the Statistical Consulting Center, a part of the UGA Department of Statistics, offers consultation services for students. They can be reached at 706-542-3286 or by e-mail at consult@stat.uga.edu

Citing your data

Staff and Services

Data Resources

Working with Data

State Data Center

Other Data Sites

News/FAQ

 
 


Comments to: web editor
© 2004 University of Georgia. All rights reserved.
The
University of Georgia Libraries