New answers tagged categorical-data
0
votes
Cohen's Kappa for repeated-measures binary variables
I'm not sure the best option but I'm pretty sure this would be a valid approach:
Intraclass correlation coefficient (ICC) is common statistic used to assess the agreement between 2 numeric variables, ...
- 419
0
votes
Categorical Dependent Variable
How to approach this? I'd say "very, very cautiously, especially if you don't have a huge sample."
First, I'd think very carefully about whether and why you need 60 levels of the dependent ...
- 128k
1
vote
Are There Techniques for Analyzing Data That Have Predictor Variables That Contain Ranges of Numbers Instead of Single Values?
You could use indicator variables for those depth categories. Then the data could be used to tell what relationships can be found. For example, some characteristics could increase or decrease ...
- 59.5k
2
votes
Are There Techniques for Analyzing Data That Have Predictor Variables That Contain Ranges of Numbers Instead of Single Values?
One thing you could do is to make each the midpoint of its category, so 0 to 5 becomes 2.5. This is rather unrealistic, but may be good enough.
To get a little more sophisticated, you could add random ...
- 128k
2
votes
Can I use raw scores of MMPI 2 in logistic regression?
Lots of issues here.
First "prove" is too strong a word. You can find evidence of something, but proving things with statistics in what will have to be an observational study is not possible ...
- 128k
2
votes
Accepted
Converting a Continuous Variable Into a Larger Ordinal Categories for Logistic Regression?
Do not discretize your data. This loses information for no gain at all. For one, it treats values in the same bucket as identical, and values that are close together but in different buckets as ...
- 131k
0
votes
McNemar test as z test or chi squared test
I would strongly NOT recommend to use McNemar test, since commonly it is impossible to reasonably interpret its results. An example in R:
...
- 1,045
6
votes
Accepted
McNemar test as z test or chi squared test
Strictly the test statistic is discrete (in either form) since it's based on counts; indeed under the assumptions (conditional on $n=b+c$), $b-c$ is a scaled and shifted binomial.
The normal and chi-...
- 290k
Top 50 recent answers are included