# Tagged Questions

Ranks are obtained when the sorted values in the data are replaced by their ranking.

4 views

### Testing for a statistically significant difference across multiple categories with ordinal data

I have ordinal data from a single reviewer who ranked the quality (and other factors) of 66 reports. 19 of these reports were about mining, 10 about Waste, 10 about Infrastructure, 9 about transport ...
13 views

### OLR, how to handle continuous response data?

I'm trying to find an alternative for a two way ANOVA as I'm treating heteroscedastic data, as I reported in a previous post here: Non parametric (or parametric) two way ANOVA in case of ...
10 views

### Variable Importance's reliability when model accuracy is low

I am trying to predict the termination events for a company. I also want to rank the most important variables that lead to termination. I ran a decision tree model and have the following questions: ...
52 views

### Appropriate statistical method for rank ordered/preference data where response on one variable depends on the other

I have a dataset of six variables that represent respondents' ranking to six different stimuli. Each respondent ranked this stimuli in order of preference from 1 to 6. This means that each row of a ...
23 views

354 views

### Rank data from 1 to N (N may not be equal to number of elements in the dataset)

I calculate percentile ranks based on the method at http://www.psychstat.missouristate.edu/introbook/sbk14m.htm I need to assign ranks to a dataset where ranks are ...
227 views

### Assumptions of Quade Rank-Based ANCOVA

Spelled out in plain English, what are the assumptions of the Quade (1967) rank-based ANCOVA procedure? It's not written very clearly in the paper. For instance, in regular ANCOVA, I believe the ...