0

How can I find out if there is a relationship between these two variables: x1 (ordinal) and x2 (continuous), this is a portion of my data:

x1  x2
severo  0,688
severo  0,879
leve    0,75
leve    0,775
severo  0,659
severo  0,801
severo  0,964
severo  0,803
moderado    0,843
leve    0,759
severo  0,865
leve    0,753
leve    0,873
leve    0,806
severo  0,293
leve    0,835
leve    0,781
leve    0,896
moderado    0,381
severo  0,767
leve    0,66
leve    0,697
severo  0,813
leve    0,92
leve    0,835
moderado    0,666
severo  0,611
severo  0,645
leve    0,856
severo  0,999
severo  0,128
severo  0,539
moderado    0,889
severo  0,94
leve    0,578
severo  0,506
severo  0,489
leve    0,644
severo  0,476
severo  0,351

Thanks in advance

nilrem
  • 39
  • 6
  • 2
    Ordinal variables are tricky, because everything in statistics must be converted to a number. Even categorical variables are converted to numbers-- 1's for belonging to a category and 0 for not belonging to a category. So with ordinal variables you generally have a choice-- treat them as unordered, which would lead you to run an ANOVA, or assign them numeric intervals, such as 1,2,and 3, which would lead you to potentially towards a simple regression. Your question is a little vague though. "Relationship" doesn't tell me much about what you're trying to figure out. Difference in group means? – Tanner Phillips Jun 26 '20 at 04:09
  • I would like to know if x2 influences the category changes of x1. Can an ordinal regression help (x1 is the dependent variable and x2 is the independent variable)? I work with the R software – nilrem Jun 26 '20 at 10:28
  • Each row is a patient, x2 is the result of a complementary examination that was carried out on a patient and x1 is the stage of a disease that the patient presents, I have other variables (age, sex) that can influence x1. – nilrem Jun 26 '20 at 10:36

1 Answers1

2

There are many similar posts on site, like Ordinal dependent variable with continuous independent variables which is almost a duplicate, and search the site!

  1. Plot the data.
  2. Group means/contrasts?
  3. Look into ordinal logistic regression.
kjetil b halvorsen
  • 77,844