Poisson Regression Help

[Putelanus Barrus]

Hi all, and apologies if this has been previously addressed. Note, that I have never used stata before, so I am completely unfamiliar, but I have a solid knowledge of Excel and SPSS.

I have a data set where the first column is the year (1980 - 2013), the dependent variable is number of publications per year, three different independent variables (policy types) which take the values from 0 to 9 depending on the number of policies that were in effect that year, and a control variable.

I need to examine the effect of the IVs on the DV and since both the DVs and the IVs are count data, I think i must use a poisson regression analysis. I have checked the internet, but due to my unfamiliarity with the matter, I would really appreciate some help on the matter.

thank you!

[Carlo Lazzaro]

Putelanus:
welcome to the list.
Assuming you have cross-sectional data (please specify if you have panel or time series data, instead) you should go -poisson- (beware of overdispersion) because your depvar takes on count values, not because your indepvars do.
You may want to take a look at -help poisson- and related entry in Stata .pdf manual.

[Putelanus Barrus]

Dear Carlo,

Thank you for your immediate response.

For now, as i want to get preliminary results, it is a time series one. However, the collected dataset applies only for the UK, but this need to be applies in the all the EU member states. will have a llok at the manual, thank you, but any additional help would be really appreciated.

[Carlo Lazzaro]

Putelanus:
thanks for providing more details.
However, as nobody on the list can see your data, additional help is conditional on posting what you typed and what Stata gave you back via CODE delimiters (as per FAQ).

[Putelanus Barrus]

Carlos, thank you. Below is an example of my data collection. this is just an example; dates go back to 1980, but most values for the IVs are 0s as there were no policies introduced. I have not yet got any stata results, as I have no idea what the commands are. Again, thank you
.
.

date	Number of Publications (DV)	Policy Type 1 (IV)	Policy Type 2 (IV)	Policy Type 3 (IV)	Total Publications across all fields (CV)
x1999	33	1	0	1	31732
x2000	32	2	0	1	32747
x2001	21	3	1	1	32081
x2002	32	4	4	1	31531
x2003	21	4	5	1	31624
x2004	13	5	5	0	29954
x2005	15	7	5	0	27988
x2006	19	8	5	0	25745
x2007	18	8	5	0	24999
x2008	17	8	6	0	23379
x2009	14	9	7	0	22465
x2010	15	9	8	0	21929
x2011	16	8	10	0	22259
x2012	6	7	9	0	23235
x2013	3	7	9	0	22938

[Marcos Almeida]

Putelanus Barrus I guess "putelanus-barrus" isn't a real name and family name at all. Rather, it may well be an username selected to login in social networks. This being so, I kindly ask you to edit your register and inform your name and family name, as recommended in the FAQ. You only need to click on the "contact us" button and send the information. Thank you. Best, Marcos.

[Carlo Lazzaro]

Putelanus:
as Marcos pointed out, you are asked to re-register with your real given and family name (a FAQ explains the reason of this preference). You are obviously free to ignore that request, just like other listers are free to ignore your queries from now on (questioning whether this request is fair or not has raised pointless debates from time to time in the past: it's simply a sort of home rule; each forum has its own).
That said, you should have idea of what command to look at for your research, since you were previously pointed to -poisson-.
Please consider that posters' good willing to collaborate in fixing their own issues is expected.