Frequently Asked Questions

These are AI-generated summaries of the introduction of the most recent publicly available working paper versions. In economics, introductions are typically 4-6 pages long and provide a summary of the method, findings, and contributions. We found that summarizing introductions provides better results than summarizing entire articles.

You can email us (ungatedresearch@outlook.com) if you find incorrect information. Authors can email us if they want their working paper or summary not to be included.

We tried several free and subscription-based AI models and settled on Google Gemini (which is publicly available).

Here is the prompt we used:

"I will first give you instructions. Can you provide a non-technical summary structured as follows: 2 to 5 bullet points on each of the following categories except for main findings which should have 4 to 7 bullet points? Remember that academic articles vary widely in their structure and content, so adapt your summary categories accordingly.

  1. Highlights: non-technical 1 sentence summary not exceeding 50 words that is easy to understand.
  2. Hypotheses or Research Questions: Summarize the main hypotheses or research questions the study aimed to address.
  3. Context and Sample or participants: If the introduction provides specific information about the study context, please specify the context or setting in which the research was conducted (e.g., geographical location, specific industry). If the introduction provides specific information about the sample, describe the characteristics of the study's sample or participants, including size, demographics, and any relevant criteria. Explain how data were collected, including instruments, surveys, interviews.
  4. Research Method: Describe the research method used to answer the research question and explain the rationale for their selection. This may include field experiments, lab experiments, simulations, modeling or analysis using quasi-experimental methods (e.g. Difference in difference, instrumental variables, regression discontinuity designs). Summarize the statistical or analytical methods employed to analyze the data.
  5. Key Findings: Highlight specific key findings or insights that the study finds. This can include 3 to 5 bullet points. This must include specific magnitudes of findings.
  6. Discussion: Provide a summary of practical applications or policy implications and future research directions.
  7. Contributions: Summarize the article's main contributions to the existing literature.

In terms of formatting, please make use of up to two levels of bullet points and make key words bold."

We conducted some quality checks (see “Limitations” section here) but there is no quality check for individual summaries.

We only provide links to publicly available working papers people could also find for themselves (and summarize using AI services). Our goal is to reduce transaction costs. There are other websites that provide complementary services (e.g. synthesizing literatures) - we have a list of these websites below.

We consulted with experts in copyright law and confirmed that providing information from journals’ publicly available tables of content is legal. AI-generated summaries are only based on publicly available working papers. We found that, occasionally, publicly available working papers still have a “copyright” designation. We tried to exclude these, but if you come across any copyrighted article, please email us (ungatedresearch@outlook.com). For working paper series with download limitations, we added the link to the website version of papers.