Did Disney Princesses Distort Girl Names in USA?

Short description of the dataset

from CRAN package description

The SSA baby names data comes from social security number (SSN) applications. SSA cards were first issued in 1936, but were only needed for people with an income. In 1986, the law changed effectively requiring all children to get an SSN at birth.

The dataset is quite simple, covering US baby name records from late 1800’s until 2014. It specifies whether a name is male or female, number of respective names in a given year and what proportion they constituted.

library(babynames)
baby <- babynames
baby$sex=as.factor(baby$sex)
summary(baby)

##       year      sex             name                 n          
##  Min.   :1880   F:1100858   Length:1858689     Min.   :    5.0  
##  1st Qu.:1950   M: 757831   Class :character   1st Qu.:    7.0  
##  Median :1983               Mode  :character   Median :   12.0  
##  Mean   :1973                                  Mean   :  183.4  
##  3rd Qu.:2002                                  3rd Qu.:   32.0  
##  Max.   :2015                                  Max.   :99680.0  
##       prop          
##  Min.   :2.260e-06  
##  1st Qu.:3.900e-06  
##  Median :7.350e-06  
##  Mean   :1.391e-04  
##  3rd Qu.:2.324e-05  
##  Max.   :8.155e-02

Installing packages

library(dplyr)

## Warning: package 'dplyr' was built under R version 3.4.4

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(ggplot2)

## Warning: package 'ggplot2' was built under R version 3.4.4

Quick data pre-prep

I assign each name to a separate dataframe.

ariel <- baby %>%
  filter(name == "Ariel", sex == "F")

## Warning: package 'bindrcpp' was built under R version 3.4.4

belle <- baby %>%
  filter(name == "Belle", sex == "F")

jasmine <- baby %>%
  filter(name == "Jasmine", sex == "F")

tiana <- baby %>%
  filter(name == "Tiana", sex == "F")

merida <- baby %>%
  filter(name == "Merida", sex == "F")

elsa <- baby %>%
  filter(name == "Elsa", sex == "F")

Next, I create variables specifying the release date of a movie with character’s name.

# The Little Mermaid
ariel_release = 1989

# Beauty and the Beast
belle_release = 1991

# Alladin
jasmine_release = 1992

# The Princess and the Frog
tiana_release = 2009

# Brave
merida_release = 2012

# Frozen
elsa_release = 2013

Plots

Finally, I plot the number of names for a given year. The arrows indicate when the movie was released, so that it’s easier to compare before and after trend. Additionally, I show the number of names and their proportion for a year proceeding and following the movie release. The numbers (and graphs!) say it all :-)

Trends

Namely, the movie release indeed seems to have a positive impact on name popularity, at least short- , but often long-term. For example, Ariel, Belle showed much higher popularity that in pre-release days even a decade after the movie has been published.

  ggplot(ariel, aes(x=year, y=n)) +
  geom_line(col="red") +
  xlab("") +
  ylab("") +
  ggtitle("Ariel") +
  geom_segment(aes(x = ariel_release, y = 0, xend = ariel_release, yend = 500), arrow = arrow(length = unit(0.1, "cm"))) +
  theme_minimal()

ariel %>%
  filter(year %in% c(ariel_release - 1,  ariel_release + 1))  %>%
  mutate(when = ifelse(year == ariel_release - 1, "1 yr before",
                      "1 yr after"))

## # A tibble: 2 x 6
##    year sex   name      n     prop when       
##   <dbl> <fct> <chr> <int>    <dbl> <chr>      
## 1 1988. F     Ariel   910 0.000473 1 yr before
## 2 1990. F     Ariel  3605 0.00176  1 yr after

## # A tibble: 2 x 6
##    year sex   name      n       prop when       
##   <dbl> <fct> <chr> <int>      <dbl> <chr>      
## 1 1990. F     Belle    13 0.00000633 1 yr before
## 2 1992. F     Belle    19 0.00000948 1 yr after

## # A tibble: 2 x 6
##    year sex   name        n    prop when       
##   <dbl> <fct> <chr>   <int>   <dbl> <chr>      
## 1 1991. F     Jasmine 11522 0.00567 1 yr before
## 2 1993. F     Jasmine 12059 0.00612 1 yr after

## # A tibble: 2 x 6
##    year sex   name      n     prop when       
##   <dbl> <fct> <chr> <int>    <dbl> <chr>      
## 1 2008. F     Tiana   481 0.000231 1 yr before
## 2 2010. F     Tiana   962 0.000492 1 yr after

tail(merida)

## # A tibble: 6 x 5
##    year sex   name       n       prop
##   <dbl> <fct> <chr>  <int>      <dbl>
## 1 2001. F     Merida     5 0.00000253
## 2 2002. F     Merida     5 0.00000253
## 3 2012. F     Merida    19 0.00000982
## 4 2013. F     Merida   109 0.0000567 
## 5 2014. F     Merida   103 0.0000529 
## 6 2015. F     Merida    99 0.0000512

## # A tibble: 2 x 6
##    year sex   name      n     prop when       
##   <dbl> <fct> <chr> <int>    <dbl> <chr>      
## 1 2012. F     Elsa    540 0.000279 1 yr before
## 2 2014. F     Elsa   1138 0.000584 1 yr after