SPSS Statistics

Expand all | Collapse all

Data set german_credit is faulty

  • 1.  Data set german_credit is faulty

    Posted 17 days ago
    Dear SPSS Statistics community,

    the data set german_credit contains a (widely used) faulty version of the German credit data. What initially made me suspicious: 963 out of 1000 credit users at a bank in southern germany are allegedly foreign workers. On further inspection, I found many various further coding errors.

    I wrote a report on how to correct the data set (Grömping 2019: South German Credit Data: Correcting a Widely Used Data Set

    ) and uploaded a corrected version also to the UCI Machine Learning Repository. Some R packages have already adopted the corrected data. It would be good if that would also happen in SPSS. I hope that this post may bring the correction to the attention of someone who can do something about it.

    Thanks and regards,

    Ulrike Grömping



    ------------------------------
    Ulrike Grömping
    ------------------------------


  • 2.  RE: Data set german_credit is faulty

    Posted 17 days ago
    Hi, and thank you for bringing this to our attention! We will definitely look at it.