What is the difference between nested and crossed designs in data analysis?

[libragirl79]

Hi,

I understand the mechanics of the crossed/nested models with fixed/random effects. What I am having trouble with is initially looking at data description and identifying if it's crossed or nested. Does anyone have a good way of going about this, some points on how to distinguish between these two?

Thanks!

 

[Stephen Tashi]

You'd get better adivce if you asked about a specific example.

Glancing at sources on the web, the general scenario is that you have established several categories and each category can have various possible values ( discrete or continuous).

So an individual datum is a set of (category, value) information, such as

(datumID, 234) (state, VA) (city, Charlottesville) (age, 23) (sex, F)

If a given category can only occur within another category then those categories are "nested" with respect to each other. For example, (city,Charlotesville) (meaning the one particular city that I have in mind, not just the name) can only occur in combination with (state,VA). It wouldn't occur with (state,NM). Categories that can occur in arbitrary combinations with another categories are "crossed" with respect to each other. For example (sex,...) can occur with any given (age,...).

I see how to classify two categories as "nested" or "crossed" with respect to each other. I don't know if it makes sense to say that a single category is "crossed" or "nested" without saying "with respect to" something else.