Factor analysis with nested data

#1
I would like to perform a factor analysis on survey data for which I have some participants (most actually) who completed multiple surveys based on perceptions of grandparents' role in their child's life. So one parent may have reported on her parents, dad's parents, her grandparents, etc. Data collection is still underway, but say for example I have 200 respondents providing responses to 400 surveys, how do I treat these data? Is it acceptable to perform a factor analysis on the 400 surveys or am I violating some assumption by doing so?
 

spunky

Doesn't actually exist
#2
Is it acceptable to perform a factor analysis on the 400 surveys or am I violating some assumption by doing so?
Yes, you are violating a more general version of the assumption of heteroskedasticity. Because you have the same people responding to multiple surveys you have measurements (surveys) nested within respondents. For this kind of data you would need to perform multilevel factor analysis:

http://www.joophox.net/papers/CFA.pdf