Statistical Analysis
X
00:00
/
00:00
CC
Welcome to the Odum Institute’s Dataverse
Network tutorial video on Statistical Analysis of
Tabular Data.
The Dataverse Network provides a
comprehensive set of statistical analysis tools
through a user interface to Zelig, a powerful, R-
based statistical computing tool.
First a user will need to identify a study's tabular
data they would like to analyze.
For this example, I will use a Harris Poll from
the 1970s dealing with attitudes towards racial
and religious minorities and towards women.
After clicking on the study title, click the
Documentation, Data, and Analysis tab,
then locate the data file and click on the link to
Access Subset/Analysis.
So I am going to expand the view to show all the
variables.
If you click on the summary icon to the right of
the variables, you can see the summary
statistics for each variable,
and for discrete variables it is displayed as a
frequency table.
From this page you can also click on any
variables and download the subset of the data in
your preferred format.
You can also apply Descriptive Statistics to
selected variables by clicking on the boxes
and choosing the type of summaries you would
like displayed and then hitting the Run Statistics
button.
From this screen you can view the summary
statistics,
or you can download your R workspace for
replication purposes, or view citation information.
The Network also provides the functionality for
users to recode subsets of data.
So I’ll click on Recode & Case-Subset, then you
pick the variable you would like to recode, select
that variable, and hit start.
Then you can check values to drop them from
the subset, enter new values, labels or ranges
as needed.
Or you can create more entries by clicking the
Add Value/Range button. And when finished you
would click Apply Recodes.
You can also preform Advanced Statistical
Analysis.
First you would choose a statistical model from
the drop-down list and select your variables for
analysis if not already selected.
For this example I will use a Cross-Tabulation
Model.
Then select the variables to be analyzed, adjust
your output options as needed, and hit Run
Model.
So here you see a Cross-Tabulation of two
different variables,
and like with the descriptive statistics, a Zip file
is provided for replication purposes as well as a
complete citation for the data set and subset.
So this tutorial showed how the Dataverse
Network provides powerful statistical analysis
tools allowing users to view descriptive statistics
recode subsets, or run advanced statistical
analysis models making it simple to analyze
data stored in the Dataverse Network.
Thanks for watching!