On Tuesday, we held the webinar Data for the Non-Data Librarian, in which pro data librarians Jen Darragh and Hailey Mooney shared their vetted approach to answering data and statistics questions. Walking us through the five steps in their search strategies and citing both free and subscription resources along the way, Jen and Hailey discussed the complex data-related questions that are often posed by students and faculty members.
Fielding questions from the audience and Twitter, Jen and Hailey held a short Q&A at the end of the webinar. Though they were not able to get to all of the questions, they were happy to provide follow-up answers here on SAGE Connection. Read on for these extra insights, a Storify of the live Twitter conversation, as well as the slides from the presentation.
Intro Questions/Clarification Questions
Can you please provide definitions of data and statistics, in addition to your examples?
Statistics report facts or figures, usually in the form of numbers/percentages in a sentence, table, or chart. They are “human readable” in that they can be interpreted without any additional analysis.
Data are the raw research outputs which require analysis in order to comprehend, i.e., “machine-readable data.” In the social sciences, these are often individual responses to a survey and are called “microdata.”
When many individual responses are combined together to provide summaries about something, like a geographic area, we then have a type of statistics called aggregate data-such as demographic statistics from the Census for all of the block groups in a city. This is where the phrase “statistical datasets” applies. Aggregate statistics are usually presented as summary statistics. They alleviate the need to query an entire raw dataset.
You referred to the American Community Survey (ACS) in your presentation. When is it sent out?
I don’t know the specific month or day, but I can say that they will mail the survey up to four times to the same address to make sure it gets completed. For more information see the ACS History document here https://www.census.gov/history/pdf/ACSHistory.pdf.
I thought I heard you use the word “archives” fairly early in this presentation. This may have been in association with a resource you called “Unobtrusive Measures”. Did you say “archives”? Also, did you say “Unobtrusive Measures”? I realize I’m very confused here.
Archives are discussed in the first search strategy for data, which is to search in a data archive. Data archives may also be referred to as data repositories and are the primary publication mechanism for data sets. An example of a data archive is the Inter-University Consortium for Political and Social Research. You can find a list of data archives across many subject areas at re3data.org.
“Unobtrusive Measures” is the title of a book on archival research methods. This book includes information on publicly available data sources for secondary analysis. Consulting books on research methods is part of data search strategy #3, literature that may lead you to sources of data. These types of books are used to introduce novice researchers to major data sources within their fields.
We use usa.gov for so many statistics questions, but I don’t believe you mentioned it. It is perfect for so many of the questions we get. Any thoughts?
Good point. I mentioned searching Google using the site:.gov limiter to search across government websites. You can use usa.gov to accomplish the same thing.
Really, the main advantage of going directly to a particular agency is that it can cut down on the amount of time digging for data.
I’ve looked for water quality data using opendata.gov and found that data from all sorts of sources is available there. Do you have any good tips for helping researchers apply an information literacy lens to distinguish authority with data sources?
Researchers should consider the source, as with any other type of information. With data, consulting the methodology behind the data collection is a key consideration.
Can you recommend any African sources of data?
My thought here is to summarize the sources listed at: http://libguides.lib.msu.edu/c.php?g=95396&p=623751.
For data archives, I’d like to mention ICPSR, which archives the Afrobarometer.
Here is a list of the national statistical archives for the countries of Africa.
Interactions with Patrons
Do your institutions purchase datasets for individual users if it becomes the property of the user (as you would for an article through ILL)?
No, at MSU we only purchase datasets that will give us a license to provide use to all of our users (just as we would for purchasing any other item). I am not aware of any programs that are similar to ILL for data. The MSU collection development policy for data is available at http://libguides.lib.msu.edu/dataservicescollectiondevpolicy.
At JHU we are currently piloting a special data fund that individuals apply to use. We wouldn’t allow them to keep the CD or DVD indefinitely, but we give them unfettered access until the completion of their project (doesn’t circulate, cannot be requested). In the event that it is a digital download, we ensure that the library keeps a copy.
What would your response be to a student saying they prefer to use a popular source, such as Wikipedia? Also, is there a place or popular encyclopedia one could use to find obscure information?
For certain inquiries, Wikipedia can be a fine starting point. You can use it to lead you to the original source of the data. It is always a good idea to track the statistic or data back to the original source so that you can be sure of its authority.
Yes, Wikipedia articles can be useful for very obscure topics. For example, if I wanted some data about citation management program features there is a great Wikipedia article which has already compiled this information.
Google Scholar is something many researchers turn to at JHU because it includes basically everything in a platform they recognize. I’ve found that if I can’t find a journal article easily in a database, it’s a great place to go to still find scholarly/peer-reviewed journal articles and books. As mentioned before, as with any third party source you want to assess authority, currency, bias, quality, etc.