Code for Merging Cross-National Surveys

In my empirical work, I make regular use of the ever-growing number of cross-national surveys available to researchers. These surveys allow us to get a better sense of trends in political preferences and behavior across both space and time.

Along with the World Values Survey (WVS), which now covers a total of 95 countries over its six waves, a number of regional survey projects have further expanded the geographic coverage of these surveys to 142 different countries. Still, its difficult to use these different resources together because of their unique coding schemes. Although there are instructions for how to merge the WVS with the four waves of the European Values Survey (EVS), no such instructions are currently available for merging the WVS with the other regional surveys.

Below, please find STATA .do files for recoding a number of regional surveys so that they may be merged with the aggregate WVS dataset. Each .do file renames (and sometimes recodes) the variables that appear in both the regional survey and the WVS, discarding those variables that are not found in the WVS (even if they appear in other regional surveys). The one exception is the code for the Latinobarometer, which also includes a recoding of all other variables to allow the different surveys to be merged together.

             Afrobarometer:     Wave 1  Wave 3  Wave 5

             Arab Barometer:   Wave 1  Wave 2  Wave 3

             Asia Barometer:    2003 2004 2005 2006 2007

             Asian Barometer:    Wave 1  Wave 2  Wave 3

             Latinobarometer1995  1996  1997  1998  2000  2001  2002  2003

                                                  2004  2005  2006  2007  2008  2009  2010  2011

             World Values     Wave 6                            

In the few cases where the regional survey questions are similarly worded to those in WVS but the answers are coded differently, I created a new, recoded variable (listed here). To make the best use of these recoded variables when using the merged surveys, it is important to recode the WVS variables to match (.do file) before merging.

If you use any of these files and find an error or have suggestions for how they could be improved, please do not hesitate to get in touch.

Panel Data from Turkey

To be added shortly.

Home    Research     Data    Teaching    About Me    Contact