Causal Claims Research Assistant (CClaRA)
Welcome to the Causal Claims Research Assistant (CClaRA) Built on Causal Graph of Economics.
A literature review tool grounded on the knowledge graph.
Search for papers by causal claims (where X causes Y)
Search for paper by concept nodes (e.g., effects of X, causes of X)
Search by claims made in specific journal (e.g. the top five)
Search by authors working on specific fields, and so on.
About CClaRA
CClaRA is a customised GPT-4o model fine-tuned with access to our knowledge graph dataset.
This means each response is based on a combination of ChatGPT -4o 's training data and the corpus.
CClaRA searches across multiple files of structured data, ensuring that every query you make provides accurate and comprehensive results.
How to Use CClaRA
Go to https://chatgpt.com/g/g-QE4aPgEVQ-cclara and sign in.
Simply type in your query.
Here are some common use cases
Exact Cause and Effect Match
Example Prompt: "Find papers where fiscal policy causes economic growth."
This search will look for papers that contain the specified cause-effect relationship and return exact matches, if available.
Cause-only Search
Example Prompt: "Find papers with 'job mobility' as a cause."
When only a cause is specified, the assistant will return all papers that mention this cause, along with a list of effects associated with it.
Effect-only Search
Example Prompt: "Find papers with 'earnings growth' as an effect."
If only an effect is provided, the assistant will list papers where this effect is documented, identifying the variety of causes related to this effect.
Semantic and Related Suggestions
Example Prompt: "Find papers where 'government spending' influences 'GDP growth'."
If an exact match isn’t found, the assistant will suggest papers with related terms or semantically similar causal claims, helping to broaden the scope when needed.
Combining Searches with Broadening Queries
Example Prompt: "Find papers where 'education' impacts 'wage growth'."
If too few results are found, the assistant may prompt you to expand the query to include related terms, such as including other measures of education or wage outcomes.
Search Scenarios and Example Prompts
"What papers discuss how monetary policy affects inflation?"
"Find studies where technological advancements lead to productivity increases."
"Show papers where carbon taxes are linked to emission reductions."
"List papers examining the impact of trade openness on economic inequality."
Example: Acemoglu, Johnson and Robinson (2001)
What to Expect in the Results
For each search, the assistant will return:
Paper ID and Title
Authors and Year of Publication
The Causal Graph of cause-effect pairs found within the paper
Journal name, whether it is a "top 5" journal and Citations count (As of November 2024)
If an exact match is unavailable, the assistant will provide a brief explanation on how the returned results relate to your search, such as showing relevant or semantically close pairs.
Feedback and Early Access to Data
If you’re interested in the underlying raw data or have specific feedback, please visit our contact form to get in touch. We welcome researchers and authors to provide corrections or enhancements to the causal claims database, ensuring accuracy and utility for the broader research community.
