Implementasi Wilcoxon Signed-Rank Test Univariat dan Multivariat Untuk Menguji Perbedaan Derajat Nyeri Pasien Endometriosis Sebelum dan Sesudah Tindakan Operasi

Abstract

In general, to test a difference in the mean of one paired sample, that is by using the paired T-test statistic, while to test an average difference in multivariate cases with paired samples, you can use the   test where the data is assumed normally distributed for the paired T test and multivariate normal for the   test. This study will compare the degree of pain of endometriosis symptoms in the form of menstrual pain that is felt a few days (a week) close to menstruation, menstrual pain in the first 1-2 days, menstrual pain on the 3rd day and until the last day of menstruation, menstrual pain outside of menstruation that feels like menstrual pain, pain when defecating during menstruation, pain when urinating during menstruation, lower back pain during menstruation, nausea and dizziness before and during menstruation, lower abdominal pain during menstruation, pain during sexual intercourse (dyspareunia) in patients before and after surgery which is paired data. Cases like this can be solved univariately using paired T-test statistics and multivariately using the   test however, after the normality assumption test is performed it is found that the assumptions of normally distributed data are not met. So for the data in this study used a univariate Wilcoxon signed ranking test and a multivariate Wilcoxon signed ranking test to test simultaneously. The results of the tests show that surgery is effective in reducing the degree of pain and symptoms in endometriosis patients.