Estimating publication bias using meta-studies
This app allows you to estimate models of selective publication using meta-studies. The estimated models give (1) the probability of publication given the z-statistic, and (2) the distribution of true effects among all studies (published and unpublished).
To use this app:
- Upload a .csv file of your meta-study, with estimates in the first column and standard errors in the second column.
Omit column headers and use comma-separated format.
- Choose a step-function model for p: Symmetric in Z or not, and cutoffs corresponding to critical values for two-sided tests at the 10%, 5%, and 1% levels.
- Choose a model for the distribution true effects. For now, this app allows normal distributions and t-distributions.
- Estimate the model by pressing the button.
To read more about the theoretical background, go to PublicationBias.pdf.
The source code for the app can be downloaded at https://github.com/maxkasy/MetaStudiesApp.
An example dataset can be found here: cleaned_minwage_data.csv. These are data from Wolfson and Belman (2015), who conducted a meta-analysis of studies on the elasticity of employment with respect to the minimum wage.