Are these variables ordinal? I am analysing some survey data, but before running some tests I'd like to be sure about the level of measurement of some variables:
income group:
Below £10,000, £10,000-19,999, £20,000-39,999, £40,000 or above
Age group:
Below 18 years old, 18-25 years old, 26-35 years old etc.
Number of travels by train:
Less than 2 times, 2-4 times, 5-6 times, More than 6 times
I would treat them as ordinal, but I've found discording info on the web.
Thanks a lot!
 A: The first two are definitely ordinal: the next category differs from the previous one on an underlying quantitative trait. The last one is a frequency of some kind, so while some nitpickers would see otherwise, I would treat it as ordinal for most analyses, too.
A: Stevens' typology of level of measurement is not really adequate for describing the whole range of variable types we deal with in practice (and analysis recommendations based only on it are certainly less than adequate in many situations).
Just calling those ordinal throws out the specific information in the bin-labels (bin-boundaries).
Strictly speaking you're dealing with interval-censored data, and in some cases it may be valuable to make use of that, rather than just shoehorn this into the more general but lower-information "data are ordinal" category. For example, by adding a distributional model to the interval-censored data you'd be able to estimate moments (means, and variances, for example), which you can't do if you ignore the information in the bin-boundaries. 
For example, a common model for income is lognormal. If you add that assumption, you can estimate average income (via maximum likelihood, for example), or use it in various other forms of analysis that may be difficult or hard to justify otherwise.
For some kinds of things, treating them just as ordinal may be fine. On the other hand, don't toss out any information you could possibly use later.
(Does this make me a nitpicker? Perhaps it does -- but often that's an essential characteristic for a statistician so I'm not too bothered by it.)
