How Pew Research Center is – and is not – using large language models in our work
In this post, we’ll share our current guidelines for the internal use of large language models and potential areas of experimentation.
A behind-the-scenes blog about research methods at Pew Research Center.
For our latest findings, visit pewresearch.org.
Data Science
-
-
Reproducibility as part of code quality control
In this post, we discuss reproducibility as a part of Pew Research Center’s code review process.
-
Redacting identifying information with computational methods in large text data
In this post, we discuss three methods to identify and remove specific words and phrases in unstructured text data.
-
How we adopted Kubernetes for our data science infrastructure
In this post, we delve into Kubernetes – the back-end tool that powers the systems our research team interacts with.
-
What Twitter users say versus what they do: Comparing survey responses with observed behaviors
We explore the connection between Americans’ survey responses and their digital activity using data from our past Twitter research.
-
Analyzing text for distinctive terms using pointwise mutual information
PMI is a quick and easy way to identify words that distinguish one group of documents from another.
-
How we built our data science infrastructure at Pew Research Center
After venturing into the world of computational social science in 2015, the Center needed to develop new tools and workflows.
-
Are topic models reliable or useful?
The final post in our series examines how topic models can and can’t help when classifying large amounts of text.
-
How keyword oversampling can help with text analysis
Keyword oversampling can be a powerful way to analyze uncommon subsets of text data.
-
Contending with the 80/20 rule when studying online behavior
The Pareto principle, or “80/20 rule,” holds that in many systems, a minority of cases produce the majority of outcomes.
