Interpreting American English Dialect Maps

Mija Van Der Wege
June 13, 2013

Data Sampling

data from Bert Vaux's online dialect survey
122 questions
collected online
approximately 31,000 responses (North America)
maps created by Joshua Katz in R Studio

SurveyMapsbyState

Individual vs. Aggregate Visualizations

Key Questions:

What conclusions are revealed by the individual and the aggregate plots?
How do we interpret heat maps vs. bubbles plotted on maps?
How does this inform our data visualization choices?

DrinkNames

Multivariate Spatial Patterns

Key Questions:

How to use this data to determine dialect clusters?
What kinds of visualizations might reveal clusters?

Crayon Vowels

Working with Student-Generated Data

Figuring out the data: variables, cases, messiness

names()
levels()
summary()

require(DCF)
load("~/psyc200.RData")
names(psyc200_data)

 [1] "sex"          "race"         "neopiE"       "neopiN"      
 [5] "neopiC"       "neopiA"       "neopiO"       "weaponIAT"   
 [9] "careerIAT"    "empathy"      "system"       "visualRT"    
[13] "visualGoNoGo"

Summarizing Data

summary(psyc200_data)

 sex    race        neopiE         neopiN         neopiC    
 F:42   C :48   Min.   : 1.0   Min.   : 0.0   Min.   : 0.0  
 M:26   NC:20   1st Qu.:37.0   1st Qu.:15.0   1st Qu.:43.0  
                Median :54.0   Median :40.0   Median :75.0  
                Mean   :54.7   Mean   :39.7   Mean   :64.5  
                3rd Qu.:73.0   3rd Qu.:56.0   3rd Qu.:89.0  
                Max.   :97.0   Max.   :97.0   Max.   :99.0  
                NA's   :3      NA's   :3      NA's   :3     
     neopiA         neopiO       weaponIAT      careerIAT   
 Min.   : 1.0   Min.   : 0.0   Min.   :0.00   Min.   :0.00  
 1st Qu.:49.0   1st Qu.:29.0   1st Qu.:1.00   1st Qu.:1.00  
 Median :71.0   Median :64.0   Median :3.00   Median :3.00  
 Mean   :63.9   Mean   :54.5   Mean   :2.47   Mean   :2.29  
 3rd Qu.:82.0   3rd Qu.:77.0   3rd Qu.:3.00   3rd Qu.:3.00  
 Max.   :97.0   Max.   :99.0   Max.   :4.00   Max.   :4.00  
 NA's   :3      NA's   :3                                   
    empathy         system        visualRT    visualGoNoGo
 Min.   :23.0   Min.   :11.0   Min.   :191   Min.   :291  
 1st Qu.:43.0   1st Qu.:16.0   1st Qu.:255   1st Qu.:348  
 Median :50.0   Median :24.0   Median :277   Median :382  
 Mean   :49.1   Mean   :25.9   Mean   :288   Mean   :390  
 3rd Qu.:57.0   3rd Qu.:32.0   3rd Qu.:309   3rd Qu.:422  
 Max.   :70.0   Max.   :62.0   Max.   :449   Max.   :594  
 NA's   :1      NA's   :1      NA's   :3     NA's   :3

Transforming Data

psyc200_data <- transform(psyc200_data,weaponIATcat = factor(weaponIAT, labels=c("reverse","none","slight","moderate","strong")))
psyc200_data <- transform(psyc200_data,careerIATcat = factor(careerIAT, labels=c("reverse","none","slight","moderate","strong")))
summary(psyc200_data)

 sex    race        neopiE         neopiN         neopiC    
 F:42   C :48   Min.   : 1.0   Min.   : 0.0   Min.   : 0.0  
 M:26   NC:20   1st Qu.:37.0   1st Qu.:15.0   1st Qu.:43.0  
                Median :54.0   Median :40.0   Median :75.0  
                Mean   :54.7   Mean   :39.7   Mean   :64.5  
                3rd Qu.:73.0   3rd Qu.:56.0   3rd Qu.:89.0  
                Max.   :97.0   Max.   :97.0   Max.   :99.0  
                NA's   :3      NA's   :3      NA's   :3     
     neopiA         neopiO       weaponIAT      careerIAT   
 Min.   : 1.0   Min.   : 0.0   Min.   :0.00   Min.   :0.00  
 1st Qu.:49.0   1st Qu.:29.0   1st Qu.:1.00   1st Qu.:1.00  
 Median :71.0   Median :64.0   Median :3.00   Median :3.00  
 Mean   :63.9   Mean   :54.5   Mean   :2.47   Mean   :2.29  
 3rd Qu.:82.0   3rd Qu.:77.0   3rd Qu.:3.00   3rd Qu.:3.00  
 Max.   :97.0   Max.   :99.0   Max.   :4.00   Max.   :4.00  
 NA's   :3      NA's   :3                                   
    empathy         system        visualRT    visualGoNoGo   weaponIATcat
 Min.   :23.0   Min.   :11.0   Min.   :191   Min.   :291   reverse : 2   
 1st Qu.:43.0   1st Qu.:16.0   1st Qu.:255   1st Qu.:348   none    :16   
 Median :50.0   Median :24.0   Median :277   Median :382   slight  :14   
 Mean   :49.1   Mean   :25.9   Mean   :288   Mean   :390   moderate:20   
 3rd Qu.:57.0   3rd Qu.:32.0   3rd Qu.:309   3rd Qu.:422   strong  :16   
 Max.   :70.0   Max.   :62.0   Max.   :449   Max.   :594                 
 NA's   :1      NA's   :1      NA's   :3     NA's   :3                   
   careerIATcat
 reverse : 5   
 none    :17   
 slight  :11   
 moderate:23   
 strong  :12

Visualizing Patterns

table(psyc200_data$weaponIATcat,psyc200_data$careerIATcat)


           reverse none slight moderate strong
  reverse        1    0      0        1      0
  none           3    4      1        6      2
  slight         1    5      2        3      3
  moderate       0    6      4        6      4
  strong         0    2      4        7      3

Correlating Variables

NEOpi <- subset(psyc200_data, select = c(neopiE,neopiN,neopiA,neopiC,neopiO))
cor(NEOpi,use="complete.obs")

        neopiE   neopiN  neopiA  neopiC   neopiO
neopiE  1.0000 -0.49173  0.2382  0.1663  0.35560
neopiN -0.4917  1.00000 -0.5355 -0.3897 -0.02323
neopiA  0.2382 -0.53546  1.0000  0.3565  0.01650
neopiC  0.1663 -0.38970  0.3565  1.0000 -0.35206
neopiO  0.3556 -0.02323  0.0165 -0.3521  1.00000

Grouping Data

SexPersonality <- groupBy(psyc200_data,by=sex,c(meanE=mean(neopiE,na.rm=TRUE),meanN=mean(neopiN,na.rm=TRUE),meanC=mean(neopiC,na.rm=TRUE),meanA=mean(neopiA,na.rm=TRUE),meanO=mean(neopiO,na.rm=TRUE)))
RacePersonality <- groupBy(psyc200_data,by=race,c(meanE=mean(neopiE,na.rm=TRUE),meanN=mean(neopiN,na.rm=TRUE),meanC=mean(neopiC,na.rm=TRUE),meanA=mean(neopiA,na.rm=TRUE),meanO=mean(neopiO,na.rm=TRUE)))
print(SexPersonality)

  sex meanE meanN meanC meanA meanO
1   F 53.95 41.74 65.23 61.62 53.05
2   M 55.88 36.73 63.38 67.19 56.58

print(RacePersonality)

  race meanE meanN meanC meanA meanO
1    C 54.72 43.21 64.45 65.45 52.13
2   NC 54.72 30.67 64.61 59.67 60.56

Visualizing Patterns

Bar chart?

bwplot(neopiN ~ race, data=psyc200_data)

plot of chunk unnamed-chunk-7

Visualizing Patterns

GoNoGo <- subset(psyc200_data, select = c(visualRT, visualGoNoGo))
ggplot(data=GoNoGo,aes(x=visualRT,y=visualGoNoGo))+geom_point() + stat_smooth(method=lm)

plot of chunk unnamed-chunk-8

summary(psyc200_data$visualRT)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
    191     255     277     288     309     449       3

summary(psyc200_data$visualGoNoGo)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
    291     348     382     390     423     594       3