CFA with panel data

[Farza Chow]

Hi - Can I use CFA with panel data or do I need to create an average by group? This is my data:
economy year topsubgrouplevel briberyincidencepercentoffirmsex briberydepthofpublictransactions percentoffirmsexpectedtogivegift
Afghanistan 2008 Manufacturing 44.7 29.4 39.0
Afghanistan 2008 Services 34 23.2 25.0
Afghanistan 2014 Manufacturing 48.1 32 26.1
Afghanistan 2014 Services 46 36.4 39.2
Angola 2006 Manufacturing 15.6 13.5 13.5
Angola 2006 Services 23.4 18.3 16.3
Angola 2010 Manufacturing 43.3 37.5 30.3
Angola 2010 Services 55.6 45.7 36.1


Hope my question is clear. Thank you

[Joseph Coveney]

At the command line in Stata, type

help sem_examples

and then scroll down to the hyperlink [SEM] example 30g and click on it.

[Phil Bromiley]

There are also several papers you can find with good on SEM and panel models.

[Farza Chow]

Thank you!

[Duong Vu]

At the command line in Stata, type

help sem_examples

and then scroll down to the hyperlink [SEM] example 30g and click on it.

Hi Joseph Coneney,

Sorry for digging this topic again.
I have the same problem. I have a panel data with 63 provinces from 2007 to 2015. And I want to use CFA.
I looked at the example 30g and it is Two level measurement model. I am not sure how can apply to the panel model.
Would you mind explaining more about it? Do you mean that the second level can be Year in my case?
Additionally, is it possible to treat panel data as extended cross-sectional data and use SEM as usual?

Thank you