i want to test the correlation between the level of the inflamatory parameter (contenious independent variable) and the possibility of other test to be positive or negative (yes/no) which statistical test i have to use?
Hi, why don´t you just use a t-test to see if there´s a significant difference in the "inflamatory parameter" between the yes and no groups?
You can test for mean difference (t-test) and variance difference (F-test).
Hi, I agree with Thomas and Faye. If you still want treat the yes/no variable as DV and the inflammatory parameter as IV you can also compute a logistic regression. And if you have several yes/no tests you can compute a multiple linear regression with those tests as predictors of the inflammatory parameter.
Hi, if yours variables don't follow a normal distribution, you can also try the Mann Withney (or Wilcoxon) test.
thanks alot all of you your answers were very helpful.
If the independent variable is normal distributed then a t-test or one way analysis of variance is OK, otherwise the Wilcoxon's two sample test.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics