Replication Guidelines

 "The replication standard holds that sufficient information exists with which to understand, evaluate, and build upon a prior work if a third party can replicate the results without any additional information from the author."  - King, Gary. 1995. Replication, Replication. PS: Political Science and Politics 28: 443–499.

According to Science Exchange who have recently put together a Reproducability Initiative: "Many of the world's top media outlets, including The New York Times and The Wall Street Journal, have reported on the issue of reproducibility in scientific research. Currently researchers lack easy avenues to validate and publish reproduced results."

However, this issue can be remedied by following the Replication Guidelines below in conjunction with publishing your datasets on the Dataverse Network.

1. Preparing data for replication:

What should be included and considered to ensure proper data replication?

    a)  Add all relevant Cataloging Information in a replication study in your Dataverse.

      • Publication citation: Including a permanent link to the original publication based on the data, since most likely the article will be in a journal that requires subscription. If no subscription required, a text file of the article can be included in the Dataverse study as part of the Documentation.
      • Data Citation Details
        • Title of the study (will be added to data citation)
        • Author(s) (will be added to data citation)
        • Production date
        • Contact information
      • Description and Scope
        • Description/abstract taken directly from your publication.
        • Keywords
      • Data Collection / Methodology
      • Data Set Availability
        • Rights of first publication: When will my data be available for others to access (embargo)? If there is an embargo, include a specific date for when the research data will be available for others to replicate.
      • Terms of Use: Who will have access to my data? More information on Terms of Use here.

    b)  Add all relevant files (Research Data, Documentation and Analysis Files) in a replication study in your Dataverse.

      • List of code, scripts, documents and data files that are needed in order to make replication possible. 
      • Create a replication data set: 
        • Deposit preferred or commonly used file formats in your discipline to ensure that others will be able to more easily replicate your research. Please remember to remove information from your datasets that must remain confidential (ex. names of survey respondents).
          • For Astrophysics: FITS files with image metadata stored in a human-readable ASCII header.
          • For the Social Sciences: Original SPSS, STATA, R files, etc with variable names and description.
        • Sets of computer program recodes (if needed).
        • Program commands, code or script for analysis (if needed).
        • Extracts of existing publicly available data (or very clear directions for how to obtain exactly the same ones you used).
        • Documentation files (full set of supporting documentation)
          • readmefile (explanatory document on how to use the files to replicate the study)
          • Text/pdf file of the article (if no subscription required).
          • Include a list of links to software or newly generated software used to replicate the data (see Gary King's website for examples).
          • Codebook
          • Data collection instruments
          • Summary statistics
          • Project summaries
          • Bibliographies of publications pertaining to the data

2. Citing for replication

Use the data citation generated by your Dataverse and add it as the first footnote or in the References within your publication to indicate where researchers can access your replication data. The persistent URL (ex. http://hdl.handle.net...) will link to a specific study in your Dataverse, and the study will contain everything needed for replication.

Visit this page for more general information on Data Citation.