Week 5 Assignment 5A - Airline Delays

Author

Supriya P.

Introduction

For this assignment, I recreated an airline arrival table as a CSV in its original wide layout, with the five cities as columns and on-time and delayed counts listed under each airline. The original table leaves the airline name blank on each “delayed” row and includes an empty row between the two airlines, so those gaps are kept in the CSV. After loading the data into R, I filled in the missing values, reshaped the data into long format, and compared the delay rates of Alaska and AM West, both overall and city by city.

Load the Data

library(tidyverse)

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.2.1     ✔ readr     2.2.0
✔ forcats   1.0.1     ✔ stringr   1.6.0
✔ ggplot2   4.0.3     ✔ tibble    3.3.1
✔ lubridate 1.9.5     ✔ tidyr     1.3.2
✔ purrr     1.2.2     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

raw <- read_csv(
  "https://raw.githubusercontent.com/0pree/symmetrical-robot/refs/heads/main/airline_delays.csv",
  show_col_types = FALSE
)

New names:
• `` -> `...1`
• `` -> `...2`

knitr::kable(raw)

…1	…2	Los Angeles	Phoenix	San Diego	San Francisco	Seattle
ALASKA	on time	497	221	212	503	1841
NA	delayed	62	12	20	102	305
NA	NA	NA	NA	NA	NA	NA
AM WEST	on time	694	4840	383	320	201
NA	delayed	117	415	65	129	61

The first two columns have no headers, the airline name is missing on the “delayed” rows, and there is a fully empty row separating the two airlines.

Clean and Fill Missing Data

flights <- raw %>%
  rename(airline = 1, status = 2) %>%
  filter(!is.na(status)) %>%
  fill(airline)

knitr::kable(flights)

airline	status	Los Angeles	Phoenix	San Diego	San Francisco	Seattle
ALASKA	on time	497	221	212	503	1841
ALASKA	delayed	62	12	20	102	305
AM WEST	on time	694	4840	383	320	201
AM WEST	delayed	117	415	65	129	61

Reshape from Wide to Long

flights_long <- flights %>%
  pivot_longer(
    cols = -c(airline, status),
    names_to = "city",
    values_to = "count"
  )

knitr::kable(head(flights_long, 10))

airline	status	city	count
ALASKA	on time	Los Angeles	497
ALASKA	on time	Phoenix	221
ALASKA	on time	San Diego	212
ALASKA	on time	San Francisco	503
ALASKA	on time	Seattle	1841
ALASKA	delayed	Los Angeles	62
ALASKA	delayed	Phoenix	12
ALASKA	delayed	San Diego	20
ALASKA	delayed	San Francisco	102
ALASKA	delayed	Seattle	305

Each row now represents one airline, one city, and one flight status, which makes it straightforward to group and summarize.

Overall Delay Rates

overall <- flights_long %>%
  group_by(airline, status) %>%
  summarize(count = sum(count), .groups = "drop") %>%
  pivot_wider(names_from = status, values_from = count) %>%
  mutate(
    total = `on time` + delayed,
    pct_delayed = round(delayed / total * 100, 1)
  )

knitr::kable(overall)

airline	delayed	on time	total	pct_delayed
ALASKA	501	3274	3775	13.3
AM WEST	787	6438	7225	10.9

ggplot(overall, aes(x = airline, y = pct_delayed, fill = airline)) +
  geom_col() +
  geom_text(aes(label = paste0(pct_delayed, "%")), vjust = -0.5) +
  labs(title = "Overall Percentage of Delayed Flights", x = NULL, y = "% Delayed") +
  theme_minimal() +
  theme(legend.position = "none")

Across all five cities combined, AM West has the better record, with 10.9% of flights delayed compared to 13.3% for Alaska.

Delay Rates by City

by_city <- flights_long %>%
  pivot_wider(names_from = status, values_from = count) %>%
  mutate(
    total = `on time` + delayed,
    pct_delayed = round(delayed / total * 100, 1)
  )

by_city %>%
  select(city, airline, pct_delayed) %>%
  pivot_wider(names_from = airline, values_from = pct_delayed) %>%
  knitr::kable()

city	ALASKA	AM WEST
Los Angeles	11.1	14.4
Phoenix	5.2	7.9
San Diego	8.6	14.5
San Francisco	16.9	28.7
Seattle	14.2	23.3

ggplot(by_city, aes(x = city, y = pct_delayed, fill = airline)) +
  geom_col(position = "dodge") +
  labs(title = "Percentage of Delayed Flights by City", x = NULL, y = "% Delayed", fill = "Airline") +
  theme_minimal()

City by city, the result reverses. Alaska has a lower delay rate than AM West in all five cities. The largest gap is in San Francisco, where 28.7% of AM West flights were delayed compared to 16.9% for Alaska.

Explaining the Discrepancy

by_city %>%
  group_by(airline) %>%
  mutate(pct_of_flights = round(total / sum(total) * 100, 1)) %>%
  select(airline, city, total, pct_of_flights, pct_delayed) %>%
  knitr::kable()

airline	city	total	pct_of_flights	pct_delayed
ALASKA	Los Angeles	559	14.8	11.1
ALASKA	Phoenix	233	6.2	5.2
ALASKA	San Diego	232	6.1	8.6
ALASKA	San Francisco	605	16.0	16.9
ALASKA	Seattle	2146	56.8	14.2
AM WEST	Los Angeles	811	11.2	14.4
AM WEST	Phoenix	5255	72.7	7.9
AM WEST	San Diego	448	6.2	14.5
AM WEST	San Francisco	449	6.2	28.7
AM WEST	Seattle	262	3.6	23.3

AM West looks better overall even though Alaska is better in every city. This is an example of Simpson’s paradox, and it comes from where each airline’s flights are concentrated. About 73% of AM West’s flights go to Phoenix, which has the lowest delay rate of any city for both airlines. About 57% of Alaska’s flights go to Seattle, where delays are much more common. AM West’s overall rate is pulled down by its large volume of Phoenix flights, while Alaska’s overall rate is pulled up by its large volume of Seattle flights. The overall numbers mostly reflect which cities each airline flies to, not how well each airline performs.

Conclusions

Comparing the airlines only on overall delay rates would lead to the wrong conclusion. Alaska performs better in every city, and AM West’s overall advantage comes from the mix of cities it serves. Looking at the data by group gives a more accurate comparison. A next step would be to add more cities or more time periods to see whether this pattern holds, or to look at other factors, like time of day or weather, that might explain why some cities have higher delay rates than others.

AI Citation

Anthropic. (2026). Claude Sonnet 5 [Large language model]. https://claude.ai. Accessed October 2026.