If you participate in the study, you will be asked to write a pre-analysis plan and develop Stata codes to analyze SOEP data. We will provide you with detailed guidelines and useful resources below, and via emails throughout the project.
PAP guidelines
Please detail the exact way you will test each hypothesis. In particular, please consider the following for each hypothesis:
Variables
Which variables will you include? How will these variables be coded – will you use them the way they are or recode them, and if so, how?
Inclusion/exclusion criteria
Are there any observations you will exclude? If yes, which? Which years will you include/what time period will you study?
Exact analysis
Please write down the exact specification of the regression model, including the names of the dependent and independent variables and whether you use robust standard errors or something else.
Interpretation of results
Please specify which statistic you will take into account for testing the hypothesis. How do you interpret whether your test is in line or not in line with the hypothesis?
Anonymity and incentives
RTs who complete the multi-analyst study will be invited to co-author the paper (listed by name and not in the form of a consortium) that results from the project. Additionally, RTs will receive the results from the Peer Assessments of their RT-PAPs as anonymous feedback after completion of the project.
As RTs are listed as co-authors, they are not anonymous in the final version of the paper. Moreover, RTs will be linked to their PAP, but they will not be connected to their code, their answers in the surveys of this project, or any particular analysis or results.
To be a co-author, the RT has to complete all the stages of the project in time and respond to all the surveys in time (including answering all survey questions in all the surveys). RTs who do not wish to be linked to their PAP may withdraw from the collaboration before the paper is submitted for publication. If they withdraw, they will not be listed as co-authors. Their anonymized RT-PAP, code, and survey answers will remain part of the paper.