Summer webinar recap and link: Exploring and Utilizing the TSHS Resources Portal
Search, Apply, and Contribute to the TSHS Resources Portal
Where did these data come from? This application is not relevant. What do these variables represent? These data are too messy. Where is the data dictionary? These data are too soft. These data are too hard. Are these some of the thoughts and questions going through your mind as you spend an afternoon exploring the web in search of the right datasets to use in your courses? Then the TSHS Resources Portal may be just right for you. Dr. Amy Nowacki of the Cleveland Clinic and Dr. Carol Bigelow of the University of Massachusetts guide us through the TSHS Resources Portal in this webinar: “Exploring and Utilizing the TSHS Resources Portal.”
What is it?
Amy and Carol explain that the TSHS Resources Portal is a repository of ready to use, clean datasets based on health science applications that have been reviewed by an editorial board and ad hoc reviewers. Each dataset is provided in multiple formats and is accompanied by an abstract, a data dictionary, a publication citation, and (for some) companion teaching resources such as projects or test questions. All datasets are open to the public, and the teaching resources and solutions are available to teachers. Amy emphasizes that the mission is to promote excellence in the teaching of statistics in the health sciences through the dissemination of peer-reviewed teaching materials archived in a public domain.
Where can you find it?
The TSHS Resources Portal is located here: https://www.causeweb.org/tshs. Amy and Carol extend their gratitude to Dr. Dennis Pearl, who is the Director of CAUSE and Professor of Biostatistics at Penn State University, for hosting this portal on CAUSE.
Search it!
Just as we observe during Amy’s tour through the portal, you will be greeted with a welcome page that gives you the option to browse or search through the list of available datasets. You can search by:
statistical topic (e.g., survival analysis or logistic regression)
study design (e.g., randomized controlled trial or longitudinal study)
medical topic (e.g., heart disease or surgery).
Once you click on the dataset of your choice, you will be taken to the dataset’s homepage. There you will find the corresponding:
abstract
contributor
background
objective
subjects and variables
data dictionary (including variable names with descriptive labels, units, and codes for categorical variables)
data downloads in six formats (e.g., SAS, R, Excel)
teaching resources (if available).
Note for Teachers: Teachers (and teaching assistants) must register and then log in to access the teaching resources that are available.
To register, click on the Membership tab at the top of the page. From there, click the Join button and enter your name, school email, school affiliation, and a new password. To reduce the likelihood of students accessing teaching resources, someone confirms the teaching status of each applicant. Providing your affiliation and related email expedites this process.
To log in, simply click login at the top right of the page.
Apply it!
These datasets can be applied in multiple settings. Amy suggests having your students use these datasets to reproduce published tables provided in the cited article. She has also used them when giving database-related talks at conferences to identify best practices for data collection and discuss the importance of secure, quality data. Amy does an excellent job walking us through an example using the surgery timing dataset and a related research paper. Amy reminds us that using demos for such presentations makes them much more captivating for the audience.
Contribute to it!
Amy and Carol encourage you to become a contributor. They review three types of submissions:
dataset
dataset with teaching resource
teaching resource associated with an existing dataset.
For details on how to contribute, click on the For Contributors tab at the top of the page.
Enjoy these resources, and please consider contributing your own! We welcome your thoughts in the comments below.