MH and I are exploring GSS data in R. I will post the syntax here so others can explore it as well if they are interested.

Dataset: http://openpsych.net/datasets/GSSsubset.7z

Dataset: http://openpsych.net/datasets/GSSsubset.7z

Code:

`library(foreign) #needed to load SPSS`

library(plyr) #for easy recode

library(Hmisc) #for rcorr

library(psych) #for stuff

library(gplots) #for plotmeans

data = read.spss("GSSsubset.sav") #read SPSS file

DF = as.data.frame(data) #convert to DF

#fix wordsum

DF$wordsum[DF$wordsum==-1] = NA #recodes -1's as NA

DF$wordsum[DF$wordsum==99] = NA #recodes 99's as NA

#sex and race

describeBy(DF$wordsum,DF$sex) # descrip. stats by sex

describeBy(DF$wordsum,DF$race) # descrip. stats by race

#yearly changes

hist(DF$year) #histogram of years to see distribution

rcorr(DF$year, DF$wordsum) #year x wordsum cor

cor(DF$year, DF$wordsum, use="pairwise.complete.obs") #using cor fun with pairwise complete

year.mean = by(DF$wordsum, INDICES = DF$year, FUN = mean, na.rm=TRUE) #get mean by year, remove missing values

year.mean.matrix = as.matrix(year.mean) #convert to matrix for plotting

plot(year.mean.matrix) #plots the matrix, but it doesnt show the year properly

plotmeans(formula = wordsum ~ year, data=DF, n.label=F) #much easier way of plotting means by year,

#remove labels for sample size. No clear FLynn effect.

**Attached Files**

**Thumbnail(s)**