Hypothesis

I hypothesized that the names of First Ladies that served in office in the last 50 years are increasing in popularity for females, and that these names signficantly increased in popularity during the first year the First Lady held their position in office.

To test this hypothesis, I selected all of the first names of the First Ladies that served in office from 1967 to 2017, which is when this dataset ends. The names are the following: Claudia, Pat, Betty, Rosalynn, Nancy, Barbara, Hillary, Laura, Michelle, and Melanie. For there to be a significant increase of popularity with the name, a large increase in the name must be reflected in the data.

Required R Packages

library(tidyverse)
library(babynames)

Analysis and Code

I began to analyze this data by filtering out the babynames library to only include the names of the chosen 10 First Ladies. I also filtered this to only include female names and data from the years of 1967 to 2017, since that is the 50 year period I chose to focus on for this analysis.

firstLadies <- babynames %>%
  filter (year %in% 1967:2017 & sex == "F" & name %in% c("Melanie", "Michelle", "Laura", "Hillary", "Barbara", "Nancy", "Rosalynn", "Betty", "Pat", "Claudia"))

I then plotted a graph to analyze the overall popularity of all 10 names from 1967 to 2017.

ggplot(data = firstLadies) + 
  geom_smooth(mapping = aes(x = year, y = prop))

This graph showed a negative corelation with year and proportion, indicating that the popularity of the names have declined since 1967. This disproved the first part of my hypthothesis, which stated that the overall popularity of the names have been on the rise.

To test the second part of my hypothesis, I analyzed each of the 10 names individually to see if there were any differences between them. To do this, I created a plot for each name using a facet wrap.

firstLadies %>% 
  ggplot(mapping = aes(x = year, y = prop)) + geom_line() + facet_wrap(~name, ncol = 2)

There does not appear to be a significant increase in any of these names in the years between 1967 to 2017. However, the following names showed little to no change in their popularity: Betty, Claudia, Pat, and Rosalynn.

I further investigated the remaining names to see if there were any significant changes in popularity during the first year of presidency of the First Lady’s husband.

Nancy

firstLadies %>% 
  filter(name %in% "Nancy") %>% 
  ggplot(aes(year, n)) + geom_point() -> nancyPlot

nancyPlot + geom_vline(xintercept = 1981, color = "red", linetype = "dotted") +
  geom_text(aes(x=1979, label="Beginning of Presidency", y=7000), colour="blue", angle=90)

There is a slight rise in popularity of the name Nancy during Nancy Reagan’s first year as First Lady. However, it is difficult to attribute this to the First Lady because it is a minor increase in popularity. After 1981, the name continues to decrease with small rises in popularity.

Barbara

firstLadies %>% 
  filter(name %in% "Barbara") %>% 
  ggplot(aes(year, n)) + geom_point() -> barbaraPlot

barbaraPlot + geom_vline(xintercept = 1989, color = "red", linetype = "dotted") +
geom_text(aes(x=1987, label="Beginning of Presidency", y=6000), colour="blue", angle=90)

There does not appear to be a rise in popularity of the name Barbara during Barbara Bush’s first year in office. The popularity of the name continues to decrease over time.

Hillary

firstLadies %>% 
  filter(name %in% "Hillary") %>% 
  ggplot(aes(year, n)) + geom_point() -> hillaryPlot

hillaryPlot + geom_vline(xintercept = 1993, color = "red", linetype = "dotted") +
  geom_text(aes(x=1994, label="Beginning of Presidency", y=1600), colour="blue", angle=90)

There is not a positive correlation between the popularity of the name Hillary and the first year that Hillary Clinton served as First Lady. The name dropped in popularity during the first year of presidency and has been on a slow decline with some jumps in popularity since 1993.

Laura

firstLadies %>% 
  filter(name %in% "Laura") %>% 
  ggplot(aes(year, n)) + geom_point() -> lauraPlot

lauraPlot + geom_vline(xintercept = 2001, color = "red", linetype = "dotted") +
  geom_text(aes(x=1999, label="Beginning of Presidency", y=10000), colour="blue", angle=90)

A positive correlation does not exist for popularity of the name Laura during Laura Bush’s first year as First Lady. The name has been on a steady decline for several years, even before the election of Bush.

Michelle

firstLadies %>% 
  filter(name %in% "Michelle") %>% 
  ggplot(aes(year, n)) + geom_point() -> michellePlot

michellePlot + geom_vline(xintercept = 2009, color = "red", linetype = "dotted") +
  geom_text(aes(x=2007, label="Beginning of Presidency", y=20000), colour="blue", angle=90)

There is not a positive correlation between the popularity of the name Michelle and Michelle Obama’s first year as First Lady. The name has been on a decline since the late 1960s.

Melanie

firstLadies %>% 
  filter(name %in% "Melanie") %>% 
  ggplot(aes(year, n)) + geom_point() -> melaniePlot

melaniePlot + geom_vline(xintercept = 2016, color = "red", linetype = "dotted") +
  geom_text(aes(x=2014, label="Beginning of Presidency", y=5000), colour="blue", angle=90)

The name Melanie has been inconsistent in popularity for several years. However, a positive correlation does not exist for popularity since Melanie Trump’s first year as First Lady.

Conclusion

My hypothesis was disproved. The data reflects that the names of First Ladies who have served in office since 1967 have decreased in popularity over time. The data also reflects that there is not a significant increase in the popularity of the names during the First Ladyies first year in office.

The names Betty, Claudia, Pat, and Rosalynn showed virtually no change in popularity since the year of 1967. The remaining of the 10 names (Nancy, Barbara, Hillary, Laura, Michelle, and Melanie) names gradually became less popular after the First Lady served in office.