README File – 4 surveys dataset

merged survey results from Porto Alegre (2005), Nairobi (2007), Atlanta (2007) and Detroit (2010).

Modifications performed by Joel Herrera (jsherrerx@gmail.com) 02 March 2015

Only 2010 variables with equivalents in the other surveys were transferred. All the excluded 2010 variables can still be found in the original 2010 dataset. Transferred variables were named according to their counterparts in the other surveys and some were recoded as necessary.

The “strongly identified” variables from all four datasets were transferred to the merged dataset. Many respondents that are “actively involved” for any given movement did not select “strongly identify”. We assume that a respondent that falls in the “actively involved” category should also “strongly identify” with their movement. Therefore, these variables were modified so that all “actively involved” respondents are also labeled as “strongly identified”.

The original variables were kept; the modified ones are named “mod” (e.g. SIqueer à SIqueermod).

RECODED VARIABLES –

USSF 2010 Detroit var30àagesimplified: birth years recoded to match age cohorts var36àcustodyofkids18: number of children recoded to no/yes children var44àeducation: level of schooling recoded to years of schooling var37àHowreligious: categories of religiousness shortened from six to three var38àReligiousbeliefs: categories recoded to match other surveys var31àracewother: categories recoded to match other surveys; South Asian, East Asian, and Pacific Islander collapsed into Asian

var33bàregion: countries collapsed into regions

var2ànumberWSF: number of WSF attended collapsed into ranges that match other surveys

computed SUM(var4ab, var4bb, var4cb, var4db, var4eb, var4fb, v4argb, var4hb, var4ib, var4jb, var4kb)àldrshippaid: all paid staff/leadership categories collapsed into a single paid staff variable computed SUM(var4ac, var4bc, var4cc, var4dc, var4ec, var4fc, var4gc, var4hc, var4ic, var4jc, var4kc)à

ldrmod: all leadership categories collapsed into a single leadership variable var9aàprotests: number of protests attended collapsed into ranges that match other surveys

var24aàIMF: “reform” response changed to “negotiate” to match other surveys var24càWTO: “reform” response changed to “negotiate” to match other surveys var22a, var22b, var22c, var22dàlevelprobsolve These variables originally ranked the importance of the levels of solving contemporary problems. The rankings were dropped and only the responses originally ranked as first were kept. For example, if a respondent ranked communities as most important, their responses to the other levels were dropped so communities could be considered their preferred level of solving contemporary problems. After this, the four different variables (one for each level) were merged into a single variable to match the other surveys.

var23àdemgov: “bad idea/possible” and “bad idea/not possible” responses were collapsed to match other surveys

var44àEducationDichotomou: level of schooling recoded to years of schooling according to binary categories