Hello,

Is the ranking change your lonely independent variable? What is your dependent variable?

Best

CH

Thank you for reading my question.

What I want to check is the impact of the universities ranks on the type of research, employment structure, and lecture in universities.

Therefore, there are three main types of dependent variables.

ⓐ Research : the subject, language, and type (collaborative vs. single) of research articles;

ⓑ Employment : non-tenure faculty ratio & annual salary system (outcome-base vs. fixed-base);

ⓒ Education : lecture type (the rate of small scale lecture & lectures in English ratio)

Thank you again!

Hello,

I think the best way is to use R statistics packages. For example, VGAM package: https://www.stat.auckland.ac.nz/~yee/VGAM/ and https://cran.r-project.org/web/packages/VGAM/vignettes/categoricalVGAM.pdf.

But in practice, ordinal level VI are treated as continuous. You can read this paper. It explains the mistake of this treatment but also the small bias in several cases: https://www3.nd.edu/~rwilliam/stats3/OrdinalIndependent.pdf. Especially in your case, I think that the ranks are not continuous variable but not an ordinal variable as a likert one. There is not a real metric but the ranks are not simply ordered categories. If you try to verify that the effect of your VI is linear, you can use your VI as traditional metric predictor. If it is not the case, you can transform it (spline, squared, ....).

I expect it could help you.

Best

CH

Thank you again for your reply, that is very helpful.

I would like to ask two more questions.

1. In addition to R statistic package, I wonder if there is a way to solve the problem through STATA or SPSS.

2. Are there research papers that are similar to the direction I would like to analyze? Could you recommend the papers if there are?

Best

Sou Hwan

Hello,

Relating to my previous link, it seems that Pasta (2009) use SAS for this kind of analyses : http://support.sas.com/resources/papers/proceedings09/248-2009.pdf.

I found a paper which analyzes the links between continuous variable and its equivalent rank-order measure: https://www.jstor.org/stable/pdf/2578079.pdf?refreqid=excelsior%3A3adfc31e98d8d00f722e827ea1759aeb.

best

CH