Airline On-Time Arrival Data

Flights categorized by destination city, airline, and whether or not the flight was on time.

Format

A data frame with 11000 observations on the following 3 variables.

airport

a factor with levels LosAngeles, Phoenix, SanDiego, SanFrancisco, Seattle

result

a factor with levels Delayed, OnTime

airline

a factor with levels Alaska, AmericaWest

Source

Barnett, Arnold. 1994. “How numbers can trick you.” Technology Review, vol. 97, no. 7, pp. 38–45.

References

These and similar data appear in many text books under the topic of Simpson's paradox.

Examples

tally(
  airline ~ result, data = AirlineArrival, 
  format = "perc", margins = TRUE)
#>              result
#> airline         Delayed    OnTime
#>   Alaska       38.89752  33.71087
#>   AmericaWest  61.10248  66.28913
#>   Total       100.00000 100.00000
tally(
  result ~ airline + airport, 
  data = AirlineArrival, format = "perc", margins = TRUE)
#> , , airport = LosAngeles
#> 
#>          airline
#> result        Alaska AmericaWest
#>   Delayed  11.091234   14.426634
#>   OnTime   88.908766   85.573366
#>   Total   100.000000  100.000000
#> 
#> , , airport = Phoenix
#> 
#>          airline
#> result        Alaska AmericaWest
#>   Delayed   5.150215    7.897241
#>   OnTime   94.849785   92.102759
#>   Total   100.000000  100.000000
#> 
#> , , airport = SanDiego
#> 
#>          airline
#> result        Alaska AmericaWest
#>   Delayed   8.620690   14.508929
#>   OnTime   91.379310   85.491071
#>   Total   100.000000  100.000000
#> 
#> , , airport = SanFrancisco
#> 
#>          airline
#> result        Alaska AmericaWest
#>   Delayed  16.859504   28.730512
#>   OnTime   83.140496   71.269488
#>   Total   100.000000  100.000000
#> 
#> , , airport = Seattle
#> 
#>          airline
#> result        Alaska AmericaWest
#>   Delayed  14.212488   23.282443
#>   OnTime   85.787512   76.717557
#>   Total   100.000000  100.000000
#> 
AirlineArrival2 <- 
  AirlineArrival %>% 
  group_by(airport, airline, result) %>% 
  summarise(count = n()) %>%
  group_by(airport, airline) %>%
  mutate(total = sum(count), percent = count/total * 100) %>% 
  filter(result == "Delayed") 
#> `summarise()` has grouped output by 'airport', 'airline'. You can override
#> using the `.groups` argument.
AirlineArrival3 <- 
  AirlineArrival %>% 
  group_by(airline, result) %>% 
  summarise(count = n()) %>%
  group_by(airline) %>%
  mutate(total = sum(count), percent = count/total * 100) %>% 
  filter(result == "Delayed") 
#> `summarise()` has grouped output by 'airline'. You can override using the
#> `.groups` argument.
  gf_line(percent ~ airport, color = ~ airline, group = ~ airline, 
          data = AirlineArrival2) %>%
    gf_point(percent ~ airport, color = ~ airline, size = ~total, 
             data = AirlineArrival2) %>%
    gf_hline(yintercept = ~ percent, color = ~airline, 
             data = AirlineArrival3, linetype = "dashed") %>%
    gf_labs(y = "percent delayed")