The goal of camerondata is to provide graduate students and applied researchers with a quick and easy access to the data sets from the empirical examples found in Cameron and Trivedi’s Microeconometrics Methods and Applications.
You can install the development version of camerondata from GitHub with:
Let’s do the data summary from Section 14.2 Binary Outcome Example: Fishing Mode Choice. You’ll need the fishing
data set.
The choice is between two alternatives: fishing from a charter boat or fishing from a pier. The relative price of charter fishing to fishing from the pier is used as explanatory variable.
The sample has data on four fishing mode alternatives and different possible explanatory variables. So we start by taking the subset of the data that we need for this example.
# Load package
library(camerondata)
# Subsample: only two modes of fishing
subsample <- fishing %>%
filter(mode == 2 | mode == 4) %>%
select(mode, dcharter, pcharter, ppier)
subsample
#> # A tibble: 630 x 4
#> mode dcharter pcharter ppier
#> <int> <int> <dbl> <dbl>
#> 1 4 1 183. 158.
#> 2 4 1 34.5 15.1
#> 3 2 0 84.9 15.1
#> 4 4 1 63.9 192.
#> 5 4 1 56.7 15.1
#> 6 2 0 48.3 34.9
#> 7 2 0 28.1 17.9
#> 8 2 0 84.9 15.1
#> 9 2 0 153. 33.5
#> 10 2 0 31.7 33.5
#> # ... with 620 more rows
Table 14.1 in the book (page 464) shows the data summary for the sample of 630 individuals. Overall average prices for the charter boat and the pier are 85 and 95 usd.
# Overall Data Summary
subsample %>%
summarize(meanPriceCharter = mean(pcharter),
meanPricePier = mean(ppier),
n = n())
#> # A tibble: 1 x 3
#> meanPriceCharter meanPricePier n
#> <dbl> <dbl> <int>
#> 1 84.9 95.2 630
Averages prices are also calculated by grouping those who choose to fish from the charter boat and those fishing from the pier. We can see that price matters in the way we would expect. People selected the alternative that was on average less expensive for them.
# Data Summary by charter boat (y=1) and pier (y=0)
subsample %>%
group_by(dcharter) %>%
summarize(MeanPriceCharter = mean(pcharter),
MeanPricePier = mean(ppier),
n = n())
#> # A tibble: 2 x 4
#> dcharter MeanPriceCharter MeanPricePier n
#> <int> <dbl> <dbl> <int>
#> 1 0 110. 30.6 178
#> 2 1 75.1 121. 452
Finally, we can calculate the sample probabilities of each choice. These are simply how frequently each choice appears in our sample. You can see that 71.7% of the sample chose charter boat fishing.