Determine Ranks of ordinal text variables
Posted: Thu Jul 01, 2021 4:50 pm
Hi there,
as a lecturer on several applied universities I am teaching basic and advanced statistics courses
that are tailored for audiences without any skills in programming or even software engineering.
I really appreciate your software and I am more and more intensively using it for the
purposes of teaching.
For some tasks I have to use the "rat run" to handle text variables. For instance if they are
of an ordinal scale I am sometimes inclined to do compute the rank of such variables to determine
the Spearman rank coefficient or to executed a Wilcoxon signed-rank test, ...
but here I at first get stuck, because Jamovi is not computing the ranks of ordinal text variables
(see figure in attachement).
It is not a problem to resolve this situation (e.g. do first a transformation
as a recoding and then compute the rank of this new variable), but also a more direct solution (that
e.g. is exploiting the levels of the observations) would be more intuitive for my audience..
So I just want to check: Is this "Jamovi problem" is a bug or rather intended ?
(I am aware that from the expert perspective everything commonly starts from variables that
are coded as numbers. But for good reasons not the perspective of the novices.)
Best regards,
Michael
as a lecturer on several applied universities I am teaching basic and advanced statistics courses
that are tailored for audiences without any skills in programming or even software engineering.
I really appreciate your software and I am more and more intensively using it for the
purposes of teaching.
For some tasks I have to use the "rat run" to handle text variables. For instance if they are
of an ordinal scale I am sometimes inclined to do compute the rank of such variables to determine
the Spearman rank coefficient or to executed a Wilcoxon signed-rank test, ...
but here I at first get stuck, because Jamovi is not computing the ranks of ordinal text variables
(see figure in attachement).
It is not a problem to resolve this situation (e.g. do first a transformation
as a recoding and then compute the rank of this new variable), but also a more direct solution (that
e.g. is exploiting the levels of the observations) would be more intuitive for my audience..
So I just want to check: Is this "Jamovi problem" is a bug or rather intended ?
(I am aware that from the expert perspective everything commonly starts from variables that
are coded as numbers. But for good reasons not the perspective of the novices.)
Best regards,
Michael