z- and z̃-Index Calculation Guide

Preparation

Download the latest retraction_watch.csv file from the Retraction Watch GitLab:

https://gitlab.com/crossref/retraction-watch-data

Requirements

• Python 3.9 or later
• pandas

Install the required package using:

pip install pandas

Notes

Only records with RetractionNature = Retraction (column Q in retraction_watch.csv) are included in the calculation; records classified as Expression of concern, Correction, or any other category are excluded.

Calculating Author z- and z̃-Indices

Step 1: Create an Author-Specific Retraction List

Run:

python ret-author.py retraction_watch.csv "John Smith" > "John Smith.csv"

This script will create a file named John Smith.csv containing all retracted papers associated with the matched author found in the Retraction Watch Database.

Notes

• Check John Smith.csv to exclude authors with the same name. Delete entire lines, not just text, as blank lines may cause errors.
• Special care is required for names containing umlauts. For example, Schön and Schon may return different search results. In such cases, running the script with a partial name, e.g., Jan Hendrik, and checking the resulting CSV file may be helpful.

Step 2: Calculate the z- and z̃-Indices

Run:

python calc-za.py ifdb.csv "John Smith.csv"

Notes

• If WARNING: journal not found: xxx is displayed, the journal is not yet included in ifdb.csv. In that case, search for the journal in Clarivate Journal Citation Reports:

  https://jcr.clarivate.com/jcr/home

  and add its impact factor data to ifdb.csv.

• If the journal is included in ifdb.csv but still not detected, add a journal name mapping to dict.csv.

Optional: stricter weighting

By default, calc-za.py uses the author-specific adjusted function. To apply the unadjusted function, shown as the solid curve in Fig. 1 of the manuscript, run:

python calc-zj.py ifdb.csv "John Smith.csv"

Calculating Journal z- and z̃-Indices

Step 1: Create a Journal-Specific Retraction List

Run:

python ret-journal.py retraction_watch.csv "Journal Name" > "Journal Name.csv"

This will create a file named Journal Name.csv containing all retracted papers associated with that journal.

Notes

• Use the full journal name whenever possible.
• Check Journal Name.csv to exclude entries from similarly named journals. Delete entire lines, not just text, as blank lines may cause errors.
• If no records are returned, the journal may be stored under a different name in the Retraction Watch Database. Search the journal name (column E in retraction_watch.csv) and check for alternative names, e.g., The Journal of ....
• If necessary, add the mapping to dict.csv. Leave the retraction_watch entry in column A unchanged. In column B, add abbreviations or alternative journal names separated by semicolons.

Step 2: Calculate the z- and z̃-Indices

Run:

python calc-zj.py ifdb.csv "Journal Name.csv"

Notes

• If WARNING: year not found is displayed, this is not an error. It simply indicates that ifdb.csv does not contain impact factor data for years earlier than 1997.
• This warning may disappear if additional historical Journal Impact Factor data are added to ifdb.csv.

Code Availability

The Python scripts used to calculate the z- and z̃-indices, together with the processed datasets, are available in this repository for research and educational purposes.

License

This repository is released under a modified BSD 3-Clause license with an additional non-commercial restriction.

Redistribution and use in source and binary forms, with or without modification, are permitted for non-commercial purposes only, subject to the terms described in the LICENSE file.