SPSS Statistics

SPSS Statistics

Your hub for statistical analysis, data management, and data documentation. Connect, learn, and share with your peers! 


#Analytics
#SPSSStatistics
#Analyticstools
 View Only
Back to discussions
Expand all | Collapse all

p-value and CI not fitting together in Bootrapping the "t-test for 2 independent samples"

  • 1.  p-value and CI not fitting together in Bootrapping the "t-test for 2 independent samples"

    Posted 06/29/22 04:18 PM
      |   view attached
    I am performing a bootstrapped "t-test for 2 independent samples" in SPSS28 on my attached dataset.
    The confidence interval seems to be okay.
    However, the p-value is far to big. The p-value estimated as .101 while the 95%CI of the mean difference does not include zero [.014, .053].
    I also tried to build bootstrap scripts from scratch in Python according to the SPSS28 documentation (https://www.ibm.com/docs/SSLVMB_28.0.0/pdf/IBM_SPSS_Statistics_Algorithms.pdf) to compare the values. I can replicate the estimated CI well, but I get far smaller p-values <.01. This problem occurs with the BCa as well as with the percentile method.

    Any ideas for clarification would be really appreciated.


    ------------------------------
    Johannes Steffen
    ------------------------------

    #SPSSStatistics

    Attachment(s)

    sav
    dataset.sav   1 KB 1 version


Related Content