In my study i have two groups of newborns: one group with Hemodinamically significant PDA(HSPDA)labeled 1 and one group with non-hemodinamically significant PDA(nHSPDA) labeled 0.

Each of these groups has a set of tests(like blood Ph, sugar blood levels etc) and the data is ordinal(ph of 7.27, sugar blood of 100 etc). The tests are dependent on wether the PDA is significant or not.

Which test do i apply to see if there is a correlation between the resuslts of a certain test and the presence of a significant PDA? Do i just compare the test results between the groups?

  • $\begingroup$ Although you can compute a correlation, it seems to me that comparing the mean difference and reflecting this as a Cohen's d effect size would be more meaningful. This is a more natural way of thinking about the difference between these groups. $\endgroup$ – dbwilson Jun 24 '18 at 12:41
  • $\begingroup$ is there a statistical test that can give me a p? And could you be a bit more detailed, please, i am quite illiterate when it comes with statistics in general, i'd be much grateful to you, professor. $\endgroup$ – Leonard Jun 24 '18 at 14:01