[EB] weight problem

Fri Oct 17 11:30:41 EDT 2003

Dear Antonio, 

the SAMPLE WEIGHT (wsample) in the Mannheim Eurobarometer Trend File
includes system missing cases (sysmis) for the European Communities Study 
1973 (115 cases for Ireland are excluded from analyses when weighting by 
wsample), Eurobarometer 5 (17 cases for Italy are excluded from analyses 
when weighting by wsample), and Eurobarometer 41.0. In the case of 41.0 in 
total 373 cases from all countries are excluded from analysis when weighting 
by wsample. These are cases of a non-national European Union citizens 
oversample which (for test purposes) was included in Eurobarometer 41.0 
(starting with Eurobarometer 41.1 the EB target population is the population 
of any nationality of an European Union member country). Hence no problem 
with Eurobarometer 41.0. 

I will have to contact the producers of the trend file about weighting ECS 1973 
and EB 5. The original European Communities Study 1973 only included a 
weight for the Netherlands; the original Eurobarometer 5 data only included a 
European Weight (no wsample kind of weighting). The weight variables in the 
trend file have been harmonized backwards by the producers in accordance 
with INRA weighting procedures from EB 32 onwards (see codebook page 
37). 

The next update of the Mannheim Trend File will inlude a note on these 
observations, especially explaining the 41.0 case. 

Thank you for the information and with best regards
Meinhard         

On 13 Oct 2003 at 14:38, Antonio Alaminos wrote:

> 
> Dear Meinhard,
> I have a problem when weighting the Eurobarometer Trend File.
> If a weight by sample I get the following message
> 
>  >Warning # 3211
>  >On at least one case, the value of the weight variable was zero,
>  negative, >or missing.  Such cases are invisible to statistical
>  procedures and graphs >which need positively weighted cases, but
>  remain on the file and are >processed by non-statistical facilities
>  such as LIST and SAVE.
> 
> If you weight and ask for the intervals  in the weight for each
> barometer
> 
> Informe
> SAMPLE WEIGHT
> EUROBAROMETER NUMBER    Mínimo    Máximo
> ECS70    ,42    4,89
> ECS71    ,09    12,73
> ECS73    ,05    15,76
> EB2    ,19    6,41
> EB3    ,07    6,67
> EB4    ,15    5,10
> EB5    ,09    7,21
> EB6    ,22    7,16
> EB7    ,23    4,70
> EB8    ,18    9,13
> EB9    ,18    4,25
> EB10    ,20    11,38
> EB10A    ,19    8,46
> EB11    ,11    6,34
> EB12    ,16    13,09
> EB13    ,10    7,66
> EB14    ,10    5,56
> EB15    ,23    7,18
> EB16    ,12    8,95
> EB17    ,11    8,56
> EB18    ,07    7,94
> EB19    ,10    14,62
> EB20    ,14    6,59
> EB21    ,05    6,61
> EB22    ,10    7,18
> EB23    ,05    8,64
> EB24    ,04    13,66
> EB25    ,04    13,57
> EB26    ,09    6,90
> EB27    ,06    8,42
> EB28    ,06    6,84
> EB29    ,09    8,43
> EB30    ,09    6,91
> EB31    ,09    8,05
> EB31A    ,04    7,52
> EB32A    ,10    5,16
> EB32B    ,12    5,27
> EB33    ,17    4,52
> EB340    ,13    4,42
> EB341    ,15    5,57
> EB350    ,18    4,23
> EB351    ,19    4,48
> EB36    ,11    4,79
> EB370    ,10    5,52
> EB371    ,08    4,72
> EB380    ,14    4,69
> EB381    ,17    4,43
> EB390    ,17    4,00
> EB391    ,05    4,75
> EB40    ,15    3,91
> EB410    ,12    4,93
> EB411    ,14    4,74
> EB42    ,11    4,18
> EB430    ,12    4,03
> EB431    ,15    4,31
> EB431B    ,14    4,55
> EB440    ,06    4,38
> EB441    ,11    5,48
> EB451    ,08    3,48
> EB460    ,10    4,21
> EB461    ,01    3,88
> EB47    ,28    4,03
> EB471    ,08    3,76
> EB472    ,06    4,58
> EB48    ,11    4,52
> EB49    ,08    5,01
> EB500    ,05    6,71
> EB501    ,04    5,80
> EB510    ,06    4,36
> EB511    ,13    6,42
> Total    ,01    15,76
> 
> You get the message I put earlier.
> I think the problem is 1997. But not sure.
> All the best,
> Antonio Alaminos
> University of Alicante
> 
> 
> 
> 

------------------------------------------------------------
 Zentralarchiv fuer Empirische Sozialforschung
 Meinhard Moschner
 Bachemer Str. 40
 D-50931 Koeln  

 Tel.:  (++49)-(0)221-47694-21                              
 Fax.:  (++49)-(0)221-47694-44                          
 e-mail: moschner at za.uni-koeln.de
 URL:    http://www.gesis.org/

 Eurobarometer information: 
 http://www.gesis.org/eurobarometer/                                     
------------------------------------------------------------