Hi,
I am new to Stata and this forum. I am trying to find the effect of changes in fuel price on road traffic crashes in a panel of 28 EU countries over 14 years of monthly data (168 months for each country). My dependent variable is a monthly road traffic accident and my independent variables are GDP per capita, unemployment, fuel prices (petrol prices and diesel prices, used in two different models due to multicollinearity). My data has large T=168 and small N=28 (countries), also some non-stationary series. So far most of the literature in accident research is pointing towards the negative binomial model and after running the fixed and random negative binomial models, according to Hausman test, FE negative binomial model is a better fit for my data. My data is an unbalanced panel.
However, test for autocorrelation, xtserial rejects the null and there is serial correlation in the series. Unit root test xtunitroot using {xtunitroot ips varname, lags(1)}, however, providing independent variable with I(0) and some independent series I(1).
I am confused about whether to consider dynamic panel models for large T like ARDL PMG,MG, DFE and xtdcce2. Or I should I go ahead with negative binomial models?
Your help will be much appreciated. Thanks
Related Posts with Accident Count Panel Data Model
lclogit2: Problem with [if] conditionDear Prof. Yoo (and all others), I am currently trying to use lclogit2 to conduct a Latent Class An…
Fixed effect model with more than two categorical variablesHi I am studying how company specific features are impacting reinsurance purchasing behavior in Kor…
Generating Dummy variable for two consecutive years with the condition that the dummy will be equal to a percentage of another variableHello everyone, I am trying to construct an overconfidence measure, namely Holder 67. The director w…
egen function xtile() not found even with egenmore installedHello everyone, I am not a pro on Stata but need to use it for a research paper in Finance for my d…
How to overcome problems in fuzzy match via matchit and reclink?Hi Statalisters, I try to use fuzzy match commands matchit and reclink to merge two datasets. Here…
Subscribe to:
Post Comments (Atom)
0 Response to Accident Count Panel Data Model
Post a Comment