Regression analysis of ranked data

[alova]

The depended variable is Binary data and independent variables are ranks from 1 to 5. What methods i suppose to use in order to see relationship between these variables? If yes, then is there any preliminary analysis needed? Thanks,

 

[Karabiner]

The depended variable is Binary data and independent variables are ranks from 1 to 5.

I suppose that you have more than 1 independent variable?
For binary dependent variables a binary logistic regression
could be used. Predictors which are ranks should be transformed
into binary variables (4 dummy-variables, one for each rank,
except for one rank which serves as reference) .

If you just want to check bivariate relationships, then you can
use the Mann-hitney U-test with your binary dependent variable
as grouping variable.

With kind regards

K.

 

[alova]

The problem is that each predictor has value from 1 to 5 which i think is not appropriate to use, no?
i was thinking to median them as it is described here: ftp://metron.sta.uniroma1.it/RePEc/articoli/2009-1-5.pdf
but i don't know how to apply it in SPSS or R

 

[Karabiner]

The problem is that each predictor has value from 1 to 5 which i think is not appropriate to use, no?

Predictors which are ranks should be transformed
into binary variables (4 dummy-variables, one for each rank,
except for one rank which serves as reference) .

Or maybe I don't understand your data structure.

With kind regards

K.