Find & Download Datasets2 of 5
Let's browse for datasets by Topics:
- On the ICPSR homepage, at the top left click the FIND DATA dropdown menu and then click FIND DATA.
- Look for the Browse area (under the search field, middle of the page, toward the left).
- Click the Topics link in the Browse options.
Find & Download Datasets3 of 5
The ICPSR folks have grouped datasets into various topics. Each "topic classification" graphic has two options for exploring:
- Click the View More button to list sub-topic areas, and click a specific sub-topic to get results for that sub-topic, or
- Click the linked topic name to get results for the entire topic area.
Let's explore a topic area:
- Scroll to the Mass Political Behavior and Attitudes topic.
- Click the View More button to see the various sub-topics in this topic area.
- Click the main Mass Political Behavior and Attitudes link to see all datasets classified with this topic.
Find & Download Datasets4 of 5
The Search Results display the dataset records classified at this topic, sorted alphabetically.
Each dataset record has the Study Title, the Investigator name(s), and a unique ICPSR ID number for that dataset.
Down the left side of the results are several Filters for narrowing results -- of particular value is the Restriction Type filter -- click it:
- Public Use datasets are available for immediate download -- students searching for datasets for a class project should limit their searches to public-use datasets.
- Restricted Use datasets require an application process to obtain -- there'll be more on this later in the guide.
Explore the other Filter options to see how they might help you in narrowing your search results.
Find & Download Datasets5 of 5
Let's look at a specific dataset record:
- Use the Sort by: dropdown menu in the Search Results to sort by Time Period.
Find and click the result called American Identity and Representation Survey, 2012 (ICPSR 36410), Investigator Schildkraut, Deborah.
- Results are now listed in descending order, starting at the most recently-collected data.
Dataset Record -- Key Parts1 of 5
There are many parts to an ICPSR dataset record -- we'll focus on those most critical for finding datasets for your own analysis.
Click the At A Glance tab
- Brief information about the dataset (Summary, Geographic Coverage, Methodology, etc.) -- use this at-a-glance information to get a sense of whether the data is relevant for your purposes.
- Look at the Subject Terms to get keyword ideas for searching for datasets, or click them to go to datasets tagged with that term -- you can also browse the ICPSR Thesaurus for possible search terms.
- If you use the dataset for research or teaching, use the suggested Citation to document your use of the dataset and to enable others to easily find it if needed.
Dataset Record -- Key Parts2 of 5
Click the Data & Documentation tab
- Lists the data files associated with the dataset -- how many separate files are listed varies with the particular dataset.
- Click the Download button to see the options for downloading the data in different formats (SPSS, Stata, etc.) and downloading the codebook and other documentation necessary to understand and use the data -- click here for a short clip of downloading and opening an SPSS file and codebook.
Dataset Record -- Key Parts3 of 5
Click the Variables tab
- Browse or search for variables within the dataset -- a good way to shop around for specific variables of interest before committing to downloading the entire codebook/documentation and data file(s).
- The Dataset column indicates the specific data file a variable and its associated data is in -- useful when there are multiple data files in the Data & Documentation tab because it will help you decide which data files you need to download.
Dataset Record -- Key Parts4 of 5
Click the Publications tab
- Lists publications that have used the dataset -- if you are going to use the data for your own analyses, you should consult how others have used it to gain insights from their work and also explore how you might analyze the data in a different/unique way.
- All publications might not be listed -- be sure to search library databases or Google Scholar to find those not listed.
Dataset Record -- Key Parts5 of 5
Click the Analyze Online button
- Many ICPSR dataset records have the option to analyze the data online without downloading it and using software -- good for initial exploration of the data (and for teaching students data analysis without having to teach them software).
- Analyze Online has two options:
- SDA -- various analysis options plus recode/compute new variables and download customized subsets options, or
- Simple crosstab/frequency -- create table with intersecting rows/columns of two variables plus charts and summary statistics.
Let's do a simple crosstab analysis:
- Click the Analyze Online button, then click the simple crosstab/frequency button.
- Click the the Row: field, type SURRPAN (the name of a specific variable), and hit Enter.
- Click the Column: field, type PPETHM (the name of a specific variable), and hit Enter.
- Click the Build the Table button.
Browse for Datasets
Now that you're familiar with a dataset record, click FIND DATA in the purple banner to return to the Find Data page and try the other browse options for finding datasets:
- Series -- series data allows you to compare topics/variables across time (e.g., year-to-year, decade-to-decade, etc.)
- Thematic Data Collections -- ICPSR partners with several statistical agencies and foundations to create collections organized around specific topics (e.g., aging, disability, health, education, criminal justice, arts & humanities, racial & ethnic minorities).
- Global Data -- datasets that offer non-US geographic coverage or include US coverage in addition to other non-US geographies.
- Browse by Keyword -- browse and click on ICPSR Subject Terms for a list of dataset records tagged with that term.
Search for Datasets
Click FIND DATA in the purple banner to return to the Find Data page.
Try a keyword search
-- click the search tips link under the search field to learn how to best construct your search and to browse the ICPSR Subject Thesaurus
for possible keywords.
Keyword searching finds your keyword(s) in the documentation for datasets, including descriptions of the individual variables.
Speaking of variables...
Search/Compare Variables1 of 4
Researchers often want datasets that contain specific variables they are interested in -- the dependent variable often the most critical, followed by independent variables that the researcher believes impact what's happening in the dependent variable.
ICPSR's Social Science Variables Database (SSVD) enables users to examine and compare variables and questions across studies or series. The SSVD includes over 5 million variables from about 76% of ICPSR's datasets.
Search/Compare Variables2 of 4To search the SSVD for variables:
- Click SEARCH/COMPARE VARIABLES in the purple banner (or at top left click FIND DATA dropdown menu and then click SEARCH/COMPARE VARIABLES).
- Let's find variables that measure how liberal or conservative a person is -- type liberal conservative in the search field and click GO.
- Let's limit our search to studies from year 2000 forward -- in Filters at the left, click Time Period button, set from date to 2000-01-01 and to date to 2018-12-31, then click Submit button.
The results list variables from different studies that have both liberal and conservative in their variable questions -- click a variable name (e.g., K_38) to see a frequency distribution of response categories, summary statistics, and a link to the study's dataset record.
Search/Compare Variables3 of 4
Let's compare two variables:
- Click the browser's back arrow to return to the Search Results list.
- In the Search Results list, find and checkbox the following variables: K_38 and F1Q1
- Click the Actions button, then click Compare.
We now have a top-to-bottom comparison of the liberal-conservative variables from the two datasets -- allowing us to easily explore similarities/differences in question wording, response categories, study time period, universe of the sample population, etc.
Search/Compare Variables4 of 4
Comparing variables is useful for various purposes:
- You may choose one dataset over the other due to the wording of the variable question or response categories.
- The number of variable response categories and their level of measurement impacts what analyses you can do, which may influence your decision to use a dataset.
- The study universe may influence your decision to use a dataset (e.g., the K_38 variable study includes only Detroit and Chicago residents, whereas the F1Q1 variable study is a US population sample).
- If you're creating your own survey instrument for original data collection, you can explore how others have worded variable questions and response categories to inform how you construct your variable.
Restricted Use Data1 of 5
Thus far you've gained familiarity with what's in an ICPSR dataset record and how to find and download "public use" data.
Let's look at a restricted-use dataset:
- Click FIND DATA in the purple banner to return to the Find Data page.
- Click the Topics link in the Browse options.
- Scroll to and click the Mass Political Behavior and Attitudes topic link.
- Find and click the result called 2012 Latino Immigrant National Election Study (LINES) (ICPSR 36680), Investigators McCann, James A.; Jones-Correa, Michael.
Restricted Use Data2 of 5
In the dataset record you'll see a Notes box (right side of screen) that says:
One or more files in this data collection have special restrictions. Restricted data files are not available for direct download from the website; click on the Restricted Data button to learn more.
Restricted Use Data3 of 5
ICPSR restricts direct download of entire datasets or specific data files and/or variables within a dataset for various reasons -- typically because they contain sensitive information about research participants, such as:
- Data that could permit the identification of individuals -- e.g., geographic location down to the Census block-level, Protected Health Information (PHI), etc.
- Data about participants that would put them at risk if not used ethically -- e.g., the restricted-use dataset you're currently looking at surveyed "Latino immigrants, the majority of whom were not U.S. citizens."
- Data from recognized "vulnerable populations" -- e.g., children, prisoners, patients, adults who are cognitively or decisionally impaired, etc.
Restricted Use Data4 of 5
Each restricted-use dataset has different processes for obtaining access to it -- click the Access Restricted Data button to see what a specific dataset requires.
Obtaining access to ICPSR restricted-use data typically requires:
- An application process requiring GSU IRB approval and a legal contract regarding ethical and responsible use of the data.
- The primary researcher to have an appointment at a research institution (like GSU) and a doctorate degree -- but faculty can apply on behalf of students, so restricted data is not necessarily off limits for graduate students' theses or dissertations.
Restricted Use Data5 of 5
Once access is granted, depending on the individual dataset, you may:
- Download the dataset and work with it on your own computer, typically with special security restrictions (e.g., a computer in a locked office, a non-networked computer), OR
- Access the data via an ICPSR virtual data enclave (VDE), which allows secure access to the data through a virtual private network connection to your own computer, OR
- Be required to travel to ICPSR (at University of Michigan in Ann Arbor) to analyze the data via a physical data enclave.
The ICPSR "Official Representative" at GSU has no authority to bypass the restricted-data requirements. They also are not the "Representative of the Institution" mentioned in the Restricted Data Use Agreement (contact GSU Legal Affairs Office
to find out who is).