Psi Chi R Contest - Dec 2023

Source: Psi Chi R Contest

#Load packages and import data

#install.packages(ggplot2,dplyr,lubridate,readr)

library(ggplot2)
library(dplyr)
library(readr)

theme_set(theme_bw())

december=read.csv('C:/Users/alanh/Documents/R/Flight Data December 2023 Contest.csv')

Level 1-

Provide descriptive statistics for the actual_flown_miles variable. Indicate the minimum number of miles flow, the maximum number of miles flow, the average number of miles flown, and the median number of miles flown.

sum(is.na(december$actual_flown_miles))

## [1] 0

min(december$actual_flown_miles,na.rm = T) #Minimum = 45

## [1] 45

max(december$actual_flown_miles,na.rm=T) #Maximum = 8440

## [1] 8440

mean(december$actual_flown_miles,na.rm=T) #Average = 2477.738

## [1] 2477.738

median(december$actual_flown_miles,na.rm=T) #Median = 1846

## [1] 1846

Level 2-

Describe the relationship between satisfaction (satisfaction_score) and delay time (arrival_delay_minutes). Note that for satisfaction_score a value of 1 is the lowest (dissatisfied) and 5 is the highest (satisfied). Note that negative values in the arrival delay minutes mean the flight was early, while positive values indicate the flight was late.

summary(december$satisfaction_score) #transform from char into numeric

##    Length     Class      Mode 
##     47074 character character

summary(december$arrival_delay_minutes)

##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## -104.000  -19.000   -9.000   -1.702    4.000  907.000

december$satisfaction_score=as.numeric(december$satisfaction_score)

december$arrival_delay_minutes=as.numeric(december$arrival_delay_minutes)

december1=december %>% 
  filter(complete.cases(satisfaction_score))

december2=december1 %>% 
  filter(complete.cases(arrival_delay_minutes))

cor.test(december2$satisfaction_score,december2$arrival_delay_minutes) #p-value < 2.2e-16, cor -0.07935616

## 
##  Pearson's product-moment correlation
## 
## data:  december2$satisfaction_score and december2$arrival_delay_minutes
## t = -14.885, df = 34961, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.08976354 -0.06893144
## sample estimates:
##         cor 
## -0.07935616

Level 3-

Does satisfaction level (satisfaction_score) vary by the cabin (cabin_code) that they’re in?

summary(december2$satisfaction_score)

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##   1.000   2.000   3.000   2.876   4.000   5.000

december2 %>% 
  ggplot(aes(x=cabin_code,y=satisfaction_score,fill=satisfaction_type))+
  geom_col()+
  labs(title='Satisfaction Score by Cabin Type',y='Satisfaction Score',x='Cabin Type',fill='Satisfaction Type')+
  theme(plot.title = element_text(hjust=.5))

#Yes. Per the column chart, there are varying levels of satisfaction by cabin type.

Level 4-

Describe the relationship between loyalty program membership (loyalty_program_level) and satisfaction (satisfaction_score). Hint: you’ll have to do some recoding for folks who don’t belong to a program

december3=december2 %>% 
  filter(complete.cases(loyalty_program_level))

#Rename blanks as NA
december3$loyalty_program_level[december3$loyalty_program_level==""]="Non-Membership"

december3 %>% 
  ggplot(aes(y=loyalty_program_level,x=satisfaction_score,fill=satisfaction_type))+
  geom_col()+
  labs(title='Loyalty and Satisfaction',y='Program Membership',x='Satisfaction Score',fill='Satisfaction Type')+
  theme(plot.title = element_text(hjust=.5))