In the following appendix, I present the R Code to reproduce the truth tables and analyses in Chapter 5. I do not print the truth tables, here, as they are already included in the chapter. In addition, I add some illustrations and graphs for the QCA analyses. Also, I show the results of some individual analyses that I do not present in chapter 5. They are for illustration purposes.
First, I load the required R packages1:
library(QCA)
library(QCA3)
library(wordcloud)
library(SetMethods)
library(car)
library(VennDiagram)Next, I import the raw data and create an empty data-set for the QCA data:
data <- read.csv("../data/raw-data.csv")
data$X <- NULL # clean up import artifact
data$cases <- paste(data$IO, data$Year, sep = "") # name cases
data_qca <- as.data.frame(matrix(0, ncol = 0, nrow = 70))
data_qca$cases <- data$cases
# add case names as row names
row.names(data_qca) <- data_qca$casesThe following code-blocks show the steps I took to calibrate the data-set. Again, the detailed description is included in Chapter 5.
thr1 <- 1
data_qca$TALK.Part <- c(thr1, 1, 1, thr1, thr1, rep(0, 31), rep(thr1, 5),
1,thr1, rep(1, 3), rep(thr1, 3), 1, rep(thr1, 4), 1, 0, 1, 1,
rep(thr1, 10), 1, 1)
data_qca$TALK.Trans <- c(rep(0, 36), rep(thr1, 9), 1, rep(thr1, 8), 1, 0,
rep(thr1, 2), 1, thr1, thr1, 1, rep(thr1, 7), 1)thr1 <- 0
thr2 <- 1
data_qca$DEC.Part <- c(rep(0, 15), rep(1, 4), rep(0, 36), rep(1, 7),
rep(0,8))
data_qca$DEC.Trans <- c(rep(0, 29), rep(thr1, 7), rep(thr2, 7),
rep(1, 12), 0, thr1, rep(thr2, 11), rep(1, 2))data_qca$ACT.Part.1 <- c(0, rep(1, 7), rep(0, 26), rep(1, 21), rep(1, 4),
rep(0, 8), rep(1, 3))
data_qca$ACT.Part.2 <- c(rep(0, 51), rep(1, 4), rep(0, 12), rep(1, 3))
data_qca$ACT.Part.4 <- c(rep(0, 25), rep(1, 13), rep(0, 3), rep(1, 5),
rep(0, 2), rep(1, 7), rep(0, 7), rep(1, 3),0, 0, 1, 1, 1)
data_qca$ACT.Trans <- c(rep(0, 4), rep(1, 3), rep(0, 38), rep(1, 10),
rep(0, 7), rep(1, 4), rep(0, 4))thr1 <- 1
data_qca$RB.Budget <- c(rep(0, 26), rep(thr1, 4), rep(0, 6), rep(thr1, 9),
rep(0, 10), rep(0, 2), rep(1, 4), rep(0, 6), rep(thr1, 3))
data_qca$RB.Inequality <- recode(data$RB.Inequality,
recodes = "0:0.8126000=0; 0.8126313:0.9=1")
data_qca$RB.Complexity <- c(rep(1, 55), rep(0, 15))thr1 <- 1
data_qca$NB.visibility.all <- c(rep(0, 29), rep(thr1, 5), rep(1, 4),
rep(thr1, 7), rep(1, 6), rep(thr1, 3), 1, rep(0, 15))
data_qca$NB.visibility.hl <- c(rep(0, 29), rep(1, 3), rep(0, 10),
rep(1, 5), rep(0, 5), rep(1, 3), rep(0, 15))
data_qca$NB.gov.depth <- c(rep(0, 13), rep(0, 21), rep(1, 21), rep(0, 15))
data_qca$NB.dem.mem <- c(rep(0, 36), rep(1, 19), rep(0, 7), rep(1, 8))
data_qca$NB.og.norm <- c(rep(0, 33), rep(thr1, 6), rep(1, 16), rep(1, 15))Next, I present the analysis of the individual outcome variables and of some combined outcomes. In the analysis for sufficiency, I limit the presentation to the intermediate solution for those analyses I do not present in more detail in Chapter 5.
#truth table w/o IAEA1957:1961, OPCW 1997
data_qca$TALKPART <- data_qca$TALK.Part # for QCA package conventions
tt <- truthTable(data_qca[c(6:55, 57:70), c(18, 10:17)], outcome =
c("TALKPART"), incl.cut1=1, sort.by="incl",
show.cases = TRUE)
# necessary conditions for participation talk
superSubset(data_qca[c(6:55, 57:70), c(18, 10:17)], outcome =
"TALKPART", relation="necessity", incl.cut=1, cov.cut=0.8)##
## incl cov.r
## ------------------------------------------------
## 1 NB.OG.NORM 1.000 0.917
## 2 rb.complexity+NB.DEM.MEM 1.000 1.000
## 3 rb.complexity+NB.GOV.DEPTH 1.000 0.943
## 4 rb.complexity+NB.VISIBILITY.ALL 1.000 0.825
## 5 RB.INEQUALITY+NB.DEM.MEM 1.000 1.000
## 6 RB.INEQUALITY+NB.GOV.DEPTH 1.000 0.943
## 7 RB.INEQUALITY+NB.VISIBILITY.ALL 1.000 0.825
## ------------------------------------------------# neg. outcome, to check for contradictions
superSubset(data_qca[c(6:55, 57:70), c(18, 10:17)], outcome =
"TALKPART", relation="necessity", incl.cut=1, cov.cut=1,
neg.out = T)##
## incl cov.r
## -------------------------------------------------------
## 1 RB.COMPLEXITY*nb.dem.mem 1.000 1.000
## 2 rb.inequality*nb.dem.mem 1.000 1.000
## 3 rb.inequality*RB.COMPLEXITY*nb.dem.mem 1.000 1.000
## -------------------------------------------------------## sufficient conditions without nc og.norm
tt <- truthTable(data_qca[c(6:55, 57:70), c(18, 10:16)],
outcome = c("TALKPART"), incl.cut1=1, sort.by="incl",
show.cases = TRUE)
eqmcc(tt, include="?", show.cases = TRUE, details = TRUE,
dir.exp = c(rep(1, 7)))##
## n OUT = 1/0/C: 33/31/0
## Total : 64
##
## p.sol: NB.DEM.MEM + rb.complexity
##
## Number of multiple-covered cases: 0
##
## M1: RB.INEQUALITY*rb.complexity +
## RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM <=> TALKPART
##
## incl cov.r cov.u
## -------------------------------------------------------------------------------
## 1 RB.INEQUALITY*rb.complexity 1.000 0.424 0.424
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM 1.000 0.576 0.576
## -------------------------------------------------------------------------------
## M1 1.000 1.000
##
## cases
## -----------------------------------------------------------------
## 1 RB.INEQUALITY*rb.complexity OPCW1998,OPCW2003; OPCW2004,OPCW2005,OPCW2006,OPCW2007,OPCW2008;
## OPCW1999,OPCW2000,OPCW2001,OPCW2002;
## OPCW2009,OPCW2010,OPCW2011
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM IAEA2004,IAEA2005,IAEA2006,IAEA2007,IAEA2008;
## IAEA2002,IAEA2003; IAEA2009,IAEA2010,IAEA2011;
## IAEA1993,IAEA1994,IAEA1995,IAEA1996,IAEA1997,IAEA1998;
## IAEA1999,IAEA2000,IAEA2001
## -----------------------------------------------------------------
##
##
## p.sol: NB.DEM.MEM + RB.INEQUALITY
##
## Number of multiple-covered cases: 3
##
## M1: RB.INEQUALITY + RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM
## <=> TALKPART
##
## incl cov.r cov.u
## -------------------------------------------------------------------------------
## 1 RB.INEQUALITY 1.000 0.515 0.424
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM 1.000 0.576 0.485
## -------------------------------------------------------------------------------
## M1 1.000 1.000
##
## cases
## -----------------------------------------------------------------
## 1 RB.INEQUALITY OPCW1998,OPCW2003; OPCW2004,OPCW2005,OPCW2006,OPCW2007,OPCW2008;
## IAEA2009,IAEA2010,IAEA2011;
## OPCW1999,OPCW2000,OPCW2001,OPCW2002;
## OPCW2009,OPCW2010,OPCW2011
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM IAEA2004,IAEA2005,IAEA2006,IAEA2007,IAEA2008;
## IAEA2002,IAEA2003; IAEA2009,IAEA2010,IAEA2011;
## IAEA1993,IAEA1994,IAEA1995,IAEA1996,IAEA1997,IAEA1998;
## IAEA1999,IAEA2000,IAEA2001
## -----------------------------------------------------------------# neg. outcome, to check for contradictions
tt <- truthTable(data_qca[c(6:55, 57:70), c(18, 10:16)],
outcome = c("TALKPART"), incl.cut1=1, sort.by="incl",
show.cases = TRUE, neg.out = T)
eqmcc(tt, include="?", show.cases = TRUE, details = TRUE)##
## n OUT = 1/0/C: 31/33/0
## Total : 64
##
## Number of multiple-covered cases: 0
##
## M1: RB.COMPLEXITY*nb.dem.mem <=> talkpart
## M2: rb.inequality*nb.dem.mem <=> talkpart
##
## -------------------
## incl cov.r cov.u (M1) (M2)
## --------------------------------------------------------------
## 1 RB.COMPLEXITY*nb.dem.mem 1.000 1.000 0.000 -
## 2 rb.inequality*nb.dem.mem 1.000 1.000 0.000 -
## --------------------------------------------------------------
## M1 1.000 1.000
## M2 1.000 1.000
##
## cases
## ----------------------------------
## 1 RB.COMPLEXITY*nb.dem.mem IAEA1962,IAEA1963,IAEA1964,IAEA1965,IAEA1966,IAEA1967,IAEA1968,IAEA1969,IAEA1970,IAEA1971,IAEA1972,IAEA1973,IAEA1974,IAEA1975,IAEA1976,IAEA1977,IAEA1978,IAEA1979,IAEA1980,IAEA1981,IAEA1982;
## IAEA1989,IAEA1990; IAEA1991,IAEA1992; IAEA1987,IAEA1988; IAEA1983,IAEA1984,IAEA1985;
## IAEA1986
## 2 rb.inequality*nb.dem.mem IAEA1962,IAEA1963,IAEA1964,IAEA1965,IAEA1966,IAEA1967,IAEA1968,IAEA1969,IAEA1970,IAEA1971,IAEA1972,IAEA1973,IAEA1974,IAEA1975,IAEA1976,IAEA1977,IAEA1978,IAEA1979,IAEA1980,IAEA1981,IAEA1982;
## IAEA1989,IAEA1990; IAEA1991,IAEA1992; IAEA1987,IAEA1988; IAEA1983,IAEA1984,IAEA1985;
## IAEA1986
## ----------------------------------data_qca$TALKTRANS <- data_qca$TALK.Trans # for QCA package conventions
tt <- truthTable(data_qca[c(1:55, 57:70), c(19, 10:17)],
outcome = c("TALKTRANS"), incl.cut1=1, sort.by="incl",
show.cases = TRUE)
# necessary conditions
superSubset(data_qca[c(1:55, 57:60), c(19, 10:17)], outcome = "TALKTRANS",
relation="nec", incl.cut=0.9, cov.cut=0.76)##
## incl cov.r
## ------------------------------------------------
## 1 NB.OG.NORM 1.000 0.885
## 2 rb.complexity+NB.DEM.MEM 1.000 1.000
## 3 rb.complexity+NB.GOV.DEPTH 1.000 0.920
## 4 rb.complexity+NB.VISIBILITY.ALL 1.000 0.767
## 5 RB.INEQUALITY+NB.DEM.MEM 1.000 1.000
## 6 RB.INEQUALITY+NB.GOV.DEPTH 1.000 0.920
## 7 RB.INEQUALITY+NB.VISIBILITY.ALL 1.000 0.767
## 8 RB.BUDGET+NB.DEM.MEM 0.957 0.846
## 9 RB.BUDGET+NB.GOV.DEPTH 0.957 0.786
## ------------------------------------------------# neg. outcome, to check for contradictions
superSubset(data_qca[c(1:55, 57:60), c(19, 10:17)], outcome = "TALKTRANS",
relation="nec", incl.cut=0.9, cov.cut=0.76, neg.out = T)##
## incl cov.r
## --------------------------------------------------------------------------------
## 1 nb.og.norm 0.917 1.000
## 2 nb.dem.mem 1.000 0.900
## 3 nb.gov.depth 0.944 0.895
## 4 nb.dem.mem*nb.og.norm 0.917 1.000
## 5 nb.gov.depth*nb.og.norm 0.917 1.000
## 6 nb.gov.depth*nb.dem.mem 0.944 0.895
## 7 nb.visibility.hl*nb.dem.mem 0.917 0.892
## 8 RB.COMPLEXITY*nb.og.norm 0.917 1.000
## 9 RB.COMPLEXITY*nb.dem.mem 1.000 1.000
## 10 RB.COMPLEXITY*nb.gov.depth 0.944 1.000
## 11 rb.inequality*nb.og.norm 0.917 1.000
## 12 rb.inequality*nb.dem.mem 1.000 1.000
## 13 rb.inequality*nb.gov.depth 0.944 1.000
## 14 nb.gov.depth*nb.dem.mem*nb.og.norm 0.917 1.000
## 15 RB.COMPLEXITY*nb.dem.mem*nb.og.norm 0.917 1.000
## 16 RB.COMPLEXITY*nb.gov.depth*nb.og.norm 0.917 1.000
## 17 RB.COMPLEXITY*nb.gov.depth*nb.dem.mem 0.944 1.000
## 18 RB.COMPLEXITY*nb.visibility.hl*nb.dem.mem 0.917 1.000
## 19 rb.inequality*nb.dem.mem*nb.og.norm 0.917 1.000
## 20 rb.inequality*nb.gov.depth*nb.og.norm 0.917 1.000
## 21 rb.inequality*nb.gov.depth*nb.dem.mem 0.944 1.000
## 22 rb.inequality*nb.visibility.hl*nb.dem.mem 0.917 1.000
## 23 rb.inequality*RB.COMPLEXITY*nb.og.norm 0.917 1.000
## 24 rb.inequality*RB.COMPLEXITY*nb.dem.mem 1.000 1.000
## 25 rb.inequality*RB.COMPLEXITY*nb.gov.depth 0.944 1.000
## 26 RB.COMPLEXITY*nb.gov.depth*nb.dem.mem*nb.og.norm 0.917 1.000
## 27 rb.inequality*nb.gov.depth*nb.dem.mem*nb.og.norm 0.917 1.000
## 28 rb.inequality*RB.COMPLEXITY*nb.dem.mem*nb.og.norm 0.917 1.000
## 29 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.og.norm 0.917 1.000
## 30 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.dem.mem 0.944 1.000
## 31 rb.inequality*RB.COMPLEXITY*nb.visibility.hl*nb.dem.mem 0.917 1.000
## 32 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.dem.mem*nb.og.norm 0.917 1.000
## --------------------------------------------------------------------------------# sufficient conditions, w/o og. norm
tt <- truthTable(data_qca[c(1:55, 57:70), c(19, 10:16)],
outcome = c("TALKTRANS"), incl.cut1=1, sort.by="incl",
show.cases = TRUE)
eqmcc(tt, include="?", show.cases = TRUE, details = TRUE,
dir.exp = c(rep(1, 7)))##
## n OUT = 1/0/C: 33/36/0
## Total : 69
##
## p.sol: NB.DEM.MEM + rb.complexity
##
## Number of multiple-covered cases: 0
##
## M1: RB.INEQUALITY*rb.complexity +
## RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM <=> TALKTRANS
##
## incl cov.r cov.u
## -------------------------------------------------------------------------------
## 1 RB.INEQUALITY*rb.complexity 1.000 0.424 0.424
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM 1.000 0.576 0.576
## -------------------------------------------------------------------------------
## M1 1.000 1.000
##
## cases
## -----------------------------------------------------------------
## 1 RB.INEQUALITY*rb.complexity OPCW1998,OPCW2003; OPCW2004,OPCW2005,OPCW2006,OPCW2007,OPCW2008;
## OPCW1999,OPCW2000,OPCW2001,OPCW2002;
## OPCW2009,OPCW2010,OPCW2011
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM IAEA2004,IAEA2005,IAEA2006,IAEA2007,IAEA2008;
## IAEA2002,IAEA2003; IAEA2009,IAEA2010,IAEA2011;
## IAEA1993,IAEA1994,IAEA1995,IAEA1996,IAEA1997,IAEA1998;
## IAEA1999,IAEA2000,IAEA2001
## -----------------------------------------------------------------
##
##
## p.sol: NB.DEM.MEM + RB.INEQUALITY
##
## Number of multiple-covered cases: 3
##
## M1: RB.INEQUALITY + RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM
## <=> TALKTRANS
##
## incl cov.r cov.u
## -------------------------------------------------------------------------------
## 1 RB.INEQUALITY 1.000 0.515 0.424
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM 1.000 0.576 0.485
## -------------------------------------------------------------------------------
## M1 1.000 1.000
##
## cases
## -----------------------------------------------------------------
## 1 RB.INEQUALITY OPCW1998,OPCW2003; OPCW2004,OPCW2005,OPCW2006,OPCW2007,OPCW2008;
## IAEA2009,IAEA2010,IAEA2011;
## OPCW1999,OPCW2000,OPCW2001,OPCW2002;
## OPCW2009,OPCW2010,OPCW2011
## 2 RB.COMPLEXITY*NB.VISIBILITY.ALL*NB.GOV.DEPTH*NB.DEM.MEM IAEA2004,IAEA2005,IAEA2006,IAEA2007,IAEA2008;
## IAEA2002,IAEA2003; IAEA2009,IAEA2010,IAEA2011;
## IAEA1993,IAEA1994,IAEA1995,IAEA1996,IAEA1997,IAEA1998;
## IAEA1999,IAEA2000,IAEA2001
## -----------------------------------------------------------------# neg. outcome, to check for contradictions
tt <- truthTable(data_qca[c(1:55, 57:70), c(19, 10:16)],
outcome = c("TALKTRANS"), incl.cut1=0.8, sort.by="incl",
show.cases = TRUE, neg.out = T)
eqmcc(tt, include="?", show.cases = TRUE, details = TRUE)##
## n OUT = 1/0/C: 36/33/0
## Total : 69
##
## Number of multiple-covered cases: 0
##
## M1: RB.COMPLEXITY*nb.dem.mem <=> talktrans
## M2: rb.inequality*nb.dem.mem <=> talktrans
##
## -------------------
## incl cov.r cov.u (M1) (M2)
## --------------------------------------------------------------
## 1 RB.COMPLEXITY*nb.dem.mem 1.000 1.000 0.000 -
## 2 rb.inequality*nb.dem.mem 1.000 1.000 0.000 -
## --------------------------------------------------------------
## M1 1.000 1.000
## M2 1.000 1.000
##
## cases
## ----------------------------------
## 1 RB.COMPLEXITY*nb.dem.mem IAEA1957,IAEA1958,IAEA1959,IAEA1960,IAEA1961,IAEA1962,IAEA1963,IAEA1964,IAEA1965,IAEA1966,IAEA1967,IAEA1968,IAEA1969,IAEA1970,IAEA1971,IAEA1972,IAEA1973,IAEA1974,IAEA1975,IAEA1976,IAEA1977,IAEA1978,IAEA1979,IAEA1980,IAEA1981,IAEA1982;
## IAEA1989,IAEA1990; IAEA1991,IAEA1992; IAEA1987,IAEA1988; IAEA1983,IAEA1984,IAEA1985;
## IAEA1986
## 2 rb.inequality*nb.dem.mem IAEA1957,IAEA1958,IAEA1959,IAEA1960,IAEA1961,IAEA1962,IAEA1963,IAEA1964,IAEA1965,IAEA1966,IAEA1967,IAEA1968,IAEA1969,IAEA1970,IAEA1971,IAEA1972,IAEA1973,IAEA1974,IAEA1975,IAEA1976,IAEA1977,IAEA1978,IAEA1979,IAEA1980,IAEA1981,IAEA1982;
## IAEA1989,IAEA1990; IAEA1991,IAEA1992; IAEA1987,IAEA1988; IAEA1983,IAEA1984,IAEA1985;
## IAEA1986
## ----------------------------------#truth table
data_qca$TALK <- fsor(data_qca$TALK.Part,
data_qca$TALK.Trans) # for QCA package conventions
tt <- truthTable(data_qca[c(20, 10:11, 13:17)], outcome =
c("TALK"), incl.cut1=1, sort.by="incl",
show.cases = TRUE)
# necessary conditions for participation talk
superSubset(data_qca[c(20, 10:17)], outcome =
"TALK", relation="necessity", incl.cut=0.8, cov.cut=0.8)##
## incl cov.r
## ------------------------------------------------
## 1 NB.OG.NORM 0.868 0.892
## 2 rb.complexity+NB.DEM.MEM 0.868 0.971
## 3 rb.complexity+NB.GOV.DEPTH 0.868 0.917
## 4 rb.complexity+NB.VISIBILITY.ALL 0.868 0.805
## 5 RB.INEQUALITY+NB.DEM.MEM 0.868 0.971
## 6 RB.INEQUALITY+NB.GOV.DEPTH 0.868 0.917
## 7 RB.INEQUALITY+NB.VISIBILITY.ALL 0.868 0.805
## 8 RB.BUDGET+NB.DEM.MEM 0.816 0.886
## ------------------------------------------------# neg. outcome, to check for contradictions
superSubset(data_qca[c(20, 10:17)], outcome =
"TALK", relation="necessity", incl.cut=0.8, cov.cut=0.8,
neg.out = T)##
## incl cov.r
## --------------------------------------------------------------------------------------
## 1 nb.og.norm 0.875 0.848
## 2 nb.dem.mem*nb.og.norm 0.875 0.848
## 3 nb.gov.depth*nb.og.norm 0.875 0.848
## 4 RB.COMPLEXITY*nb.og.norm 0.875 0.848
## 5 RB.COMPLEXITY*nb.dem.mem 0.969 0.861
## 6 RB.COMPLEXITY*nb.gov.depth 0.906 0.853
## 7 rb.inequality*nb.og.norm 0.875 0.848
## 8 rb.inequality*nb.dem.mem 0.969 0.861
## 9 rb.inequality*nb.gov.depth 0.906 0.853
## 10 rb.budget*nb.dem.mem 0.875 0.800
## 11 nb.gov.depth*nb.dem.mem*nb.og.norm 0.875 0.848
## 12 RB.COMPLEXITY*nb.dem.mem*nb.og.norm 0.875 0.848
## 13 RB.COMPLEXITY*nb.gov.depth*nb.og.norm 0.875 0.848
## 14 RB.COMPLEXITY*nb.gov.depth*nb.dem.mem 0.906 0.853
## 15 RB.COMPLEXITY*nb.visibility.hl*nb.dem.mem 0.875 0.848
## 16 RB.COMPLEXITY*nb.visibility.hl*nb.gov.depth 0.812 0.839
## 17 rb.inequality*nb.dem.mem*nb.og.norm 0.875 0.848
## 18 rb.inequality*nb.gov.depth*nb.og.norm 0.875 0.848
## 19 rb.inequality*nb.gov.depth*nb.dem.mem 0.906 0.853
## 20 rb.inequality*nb.visibility.hl*nb.dem.mem 0.875 0.848
## 21 rb.inequality*nb.visibility.hl*nb.gov.depth 0.812 0.839
## 22 rb.inequality*RB.COMPLEXITY*nb.og.norm 0.875 0.848
## 23 rb.inequality*RB.COMPLEXITY*nb.dem.mem 0.969 0.861
## 24 rb.inequality*RB.COMPLEXITY*nb.gov.depth 0.906 0.853
## 25 rb.budget*RB.COMPLEXITY*nb.dem.mem 0.844 0.844
## 26 rb.budget*rb.inequality*nb.dem.mem 0.844 0.844
## 27 RB.COMPLEXITY*nb.gov.depth*nb.dem.mem*nb.og.norm 0.875 0.848
## 28 RB.COMPLEXITY*nb.visibility.hl*nb.gov.depth*nb.dem.mem 0.812 0.839
## 29 rb.inequality*nb.gov.depth*nb.dem.mem*nb.og.norm 0.875 0.848
## 30 rb.inequality*nb.visibility.hl*nb.gov.depth*nb.dem.mem 0.812 0.839
## 31 rb.inequality*RB.COMPLEXITY*nb.dem.mem*nb.og.norm 0.875 0.848
## 32 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.og.norm 0.875 0.848
## 33 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.dem.mem 0.906 0.853
## 34 rb.inequality*RB.COMPLEXITY*nb.visibility.hl*nb.dem.mem 0.875 0.848
## 35 rb.inequality*RB.COMPLEXITY*nb.visibility.hl*nb.gov.depth 0.812 0.839
## 36 rb.budget*rb.inequality*RB.COMPLEXITY*nb.dem.mem 0.844 0.844
## 37 rb.inequality*RB.COMPLEXITY*nb.gov.depth*nb.dem.mem*nb.og.norm 0.875 0.848
## 38 rb.inequality*RB.COMPLEXITY*nb.visibility.hl*nb.gov.depth*nb.dem.mem 0.812 0.839
## --------------------------------------------------------------------------------------