Little help for a beginner?

#1
Hi all,

I'm new to the forum!

I have recently completed an introductory course in statistics and research methods using STATA.

I am nearing the end of collecting data for a research project, but I'm not all that sure as to how to go about using it in STATA.

From my reading I understand that I am looking at time-series cross-sectional data (9 different companies over the space of 10 years). I am looking to see how the change in the frequency of language used by the media effects the language used by these companies when reporting.

What is the best way of going about looking at this? I'm only looking for a possible correlation between the variables; but I have read that I need to introduce dummy variables and time lags? (something I am unsure about).

Any help would be much appreciated!

Regards,
Lucas
 
#3
That's a very broad methodological question rather than a technical one. Only your supervisor can help you with this because we don't know what kind of variables you have (in order to generate new dummies). For the time series analysis I would do this:
1. Set datevar: tsset datevar
2. Create lags, leads and differences
3 Explore autocorrelation (AC, PAC, PAC graph)
4 Explore the relationship between two time series
5. Select lag(Schwarz's Baysian, Akaike's, Hann & Quinn)
6. Dickey-Fuller test for stationarity
7. Test for cointegration
8.Granger causality
9.Chow-test
10.Test for unknown breaks (theoretically yes. I am not that advanced so I would skip this)
11. White noise test