Data and AI policies

Data policy

Replicability and reproducibility of research denote a cornerstone in the scientific system. They enable the validation of published results, and therethrough creates confidence in the reliability, robustness and generalisability of scientific findings. Replicability and reproducibility also promote scientific progress by fostering the discovery of new evidence, expanding our understanding of economic problems and challenging existing theories or findings.Replicability and reproducibility of research denote a cornerstone in the scientific system. They enable the validation of published results, and therethrough creates confidence in the reliability, robustness and generalisability of scientific findings. Replicability and reproducibility also promote scientific progress by fostering the discovery of new evidence, expanding our understanding of economic problems and challenging existing theories or findings.

For these reasons, GJAE is committed to open science, replication research and the FAIR (findable, accessible, interoperable, and reusable) principles for research data. This means that (i) a data availability statement is a requirement for papers published in the GJAE, (ii) GJAE expects authors to make data available in GJAE Data Repository at the ZBW – Leibniz Information Centre for Economics’ Journal Data Archive whenever legally and ethically feasible, and (iii) expects authors to make program code, software, etc. openly available using replication packages that are archived in the GJAE Data Repository at the ZBW – Leibniz Information Centre for Economics Journal Data Archive.

Data Availability Statement

Data availability statements confirm the presence or absence of shared data, are short and standardized to inform readers about the availability of the data used in the research process. For this purpose, the GJAE is offering a template.

The data availability statement should mention any restrictions on accessing the data. If the data is restricted (third-party data, legal or ethical constraints), this must be explained in the data availability statement. In addition, authors are asked to provide detailed information on how other researchers can obtain the restricted data in their readme-file (see below). In this case, we encourage authors to provide synthetic data sets that mimic the original data for testing the code.

Authors that have employed empirical methods and data will receive an invitation email by the GJAE Data Repository at the Journal Data Archive (send to [journaldata at zbw dot eu]) to deposit their data if legally and ethically feasible, and if these data not stored prior to the submission in a comparable repository. This will happen in the final step of the publication process.

It is generally not acceptable that data be provided "upon request" if the request must be approved by the authors themselves.

Data Citation

In recognition of the relevance of data as an output of research effort, GJAE endorses the FORCE11 Data Citation Principles. Therefore, authors are asked to cite the data in the same way as article, book and web citations and include data citations as part of their reference list.

When citing or making claims based on data, authors should refer to the data at the relevant place in the manuscript text and in addition provide a formal citation in the reference list. We recommend the format proposed by the Joint Declaration of Data Citation Principles:

[dataset] Authors; Year; Dataset title; Data repository or archive; Version (if any); Persistent identifier (e.g. DOI)

Replication Packages

We expect all our authors of accepted papers that contain empirical work, simulations, or experimental work to provide information about the data, programs, and other details of the computations sufficient to permit computational reproduction. Data, programs and other relevant material should be archived in the GJAE Data Repository at the ZBW – Leibniz Information Centre for Economics’ Journal Data Archive, if legally and ethically possible.

Data and Variables

Data should be provided as part of the replication package unless the exceptions for restricted data apply (see above) or unless they can be fully reproduced from other accessible data within a reasonable time frame and with reasonable resources.

Each variable in the provided datasets should have a meaningful name or description (label), or authors may provide separate codebooks or similar metadata that describe the allowed values and their meaning. It is also possible to refer to publicly accessible documents that fulfil this purpose.

Readme-file

All replication packages must include a “Read me” file (clearly labelled) containing a list of all files included and guiding a user on the types of files and how to use them to do replication.
If necessary, you will find detailed instructions on how to create a readme file here.

Program Code

Programs that produce computational results such as estimation, simulation, model solution must be included. Ideally, these programs reproduce all the computational exhibits in the paper with minimal human intervention.

A master script is encouraged. When no master script is included, please provide sufficient and precise step-by-step instructions in your readme file, allowing users to exactly reproduce the generated outputs with the least amount of effort.

Experimental Instructions and Surveys

Details regarding experimental procedures and instructions, and questionnaires are necessary to evaluate submissions of experimental research and research based on primary data collection. This information is also important to facilitate replicability and related work by subsequent researchers. Authors of experimental and survey-based papers must include detailed experimental instructions, along with screen shots for computerized experiments or record sheets for non-computerized ones, or the questionnaire. These instructions, questionnaires and related materials should be provided in the replication package. In cases where the instructions and materials were not presented to participants in English, both the original version and an English translation of materials shall be provided.

Further information

AI policy

This policy refers to generative AI tools, i.e generative models for the creation of new content in the form of text, images, audio, video, software code, or data sets. Currently known examples include text-generating chatbots such as ChatGPT, Copilot, Gemini, or LLaMA as well as image-generating AI tools such as DALL-E, Stable Diffusion, or Midjourney.

  • In accordance with the COPE position statement on Authorship and AI tools, AI tools cannot be listed as authors of a paper. These tools cannot take responsibility for the submitted work and hence do not meet the requirements for authorship such as the ability to declare competing interests or to agree to the license agreement.
  • Any use of AI tools in the writing of a manuscript, image/graphic generation, or in the collection and analysis of data, must be disclosed transparently in the paper by describing for what purpose and how AI tools were used. Moreover, it has to be stated which AI tool(s) were used and when they were used.
  • When authors use AI for literature research, they always need to verify the sources as AI-generated references may be inaccurate or non-existent. Submission containing non-existent references will be rejected.
  • The use of AI tools in the review process is not permitted, as the confidentiality of the submissions cannot be guaranteed. Furthermore, the reviewers are fully responsible for the content of their reviews.